Geobotany Studies

Basics, Methods and Case Studies

Franco Pedrotti
Elgene Owen Box   Editors

Tools
for Landscape-Scale
Geobotany and
Conservation
Geobotany Studies

Basics, Methods and Case Studies

Series Editor
Franco Pedrotti, Department of Botany and Ecology, University of Camerino,
Camerino, Italy

Editorial Board Members

S. Bartha, Vácrátót, Vácrátót, Hungary
F. Bioret, University of Brest, Brest, France
E. O. Box, University of Georgia, Athens, GA, USA
A. Čarni, Slovenian Academy of Sciences, Ljubljana, Slovenia
K. Fujiwara, Yokohama City University, Yokohama, Japan
D. Gafta, “Babes-Bolyai” University, Cluj Napoca, Romania
J. Loidi, University of Bilbao, Bilbao, Spain
L. Mucina, University of Perth, Perth, Australia
S. Pignatti, Università degli Studi di Roma “La Sapienza”, Roma, Italy
R. Pott, University of Hannover, Hannover, Germany
D. Sánchez Mata, Universidad Complutense Madrid, Madrid, Spain
A. Velázquez, Centro de Investigación en Sciéncias Ambientales, Morelia, Mexico
R. Venanzoni, University of Perugia, Perugia, Italy
The series includes outstanding monographs and collections of papers on a range of
topics in the following fields: Phytogeography, Phytosociology, Plant Community
Ecology, Biocoenology, Vegetation Science, Eco-informatics, Landscape Ecology,
Vegetation Mapping, Plant Conservation Biology, and Plant Diversity.
Contributions should reflect the latest theoretical and methodological developments
or present new applications on large spatial or temporal scales that will reinforce
our understanding of ecological processes acting at the phytocoenosis and
vegetation landscape level. Case studies based on large data sets are also
considered, provided they support habitat classification refinement, plant diversity
conservation or vegetation change prediction. Geobotany Studies: Basics, Methods
and Case Studies is the successor to Braun-Blanquetia, a journal published by the
University of Camerino from 1984 to 2011 in cooperation with the Station
Internationale de Phytosociologie (Bailleul, France) and the Dipartimento di
Botanica ed Ecologia (Università di Camerino, Italy) and under the aegis of the
Société Amicale Francophone de Phytosociologie, the Société Française de
Phytosociologie, the Rheinold-Tüxen-Gesellschaft and the Eastern Alpine and
Dinaric Society for Vegetation Ecology. This series promotes the expansion,
evolution, and application of the invaluable scientific legacy of the Braun-Blanquet
school.

More information about this series at http://www.springer.com/series/10526

Franco Pedrotti Elgene Owen Box
•

Editors

Tools for Landscape-Scale

Geobotany and Conservation

123
Editors
Franco Pedrotti Elgene Owen Box
Department of Botany and Ecology Department of Geography
University of Camerino University of Georgia
Camerino, Italy Athens, GA, USA

ISSN 2198-2562 ISSN 2198-2570 (electronic)

Geobotany Studies
ISBN 978-3-030-74949-1 ISBN 978-3-030-74950-7 (eBook)
https://doi.org/10.1007/978-3-030-74950-7
© The Editor(s) (if applicable) and The Author(s), under exclusive license
to Springer Nature Switzerland AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Tools for Landscape-Scale Geobotany and Conservation

This book presents a variety of methodologies, tools (s.l.), and other approaches
for ecological analysis and evaluation of sites and landscapes, especially for
geobotanical study and conservation purposes. Most of the chapters represent talks
from special sessions at two annual meetings of the International Association for
Vegetation Science (IAVS):
1) Pirenopolis (Brazil, 2016), on “Applied Mapping for Conservation and
Management: from Plant to Landscape”
2) Palermo (Sicily, 2017), on “Vegetation Patterns in relation to multi-scale levels
of ecological complexity: from associations to geoseries”
The individual papers are grouped by methodology: mapping, phytosociology,
basic ecological analysis, and conservation strategy.

Mapping

Mapping was a main theme of both special sessions and is an often indispensable
tool in vegetation study at all scales. Javier Loidi presents two chapters on various
aspects of vegetation mapping. The first represents a good introduction to major
concepts of biogeography in general and then considers how to map biogeo-
graphical territories at different ranks, from kingdom to down to district, using
floristic, vegetation, and landscape criteria. The second chapter explains and
develops further the concepts of vegetation catenas, series, and geosigmeta.
A catalog of different types of sigmeta is presented, along with a methodology for
mapping dynamic vegetation catenas, as a tool for ecological restoration and
conservation.
The science of symbols and symbolism is called semiology, and some basic
principles of semiology are explored by Perrin and Bioret, especially the use of
hues and luminosity as well as colors per se to represent different phytosociological

v
vi Preface

units. Another mapping consideration involves bias due to differential sampling

effort. How to map this bias explicitly, in Europe, is treated by Garzon-Lopez and
Rocchini. The concept of ecosystem services is rather anthropocentric but is also
well established in various aspects of environmental policy. A pilot project using
detailed data of many kinds from many sources is described in detail by Mesquita
et al. to assess and map ecosystem services in southern Portugal. In addition to
ecosystem services, this project defines criteria for assessing the value, threat level,
rarity, naturalness, phytodiversity, and replaceability of the ecosystems treated.
Mountains pose particular problems for both field study of vegetation and
mapping. In the Bolivian Andes, at elevations over 3000m, wetland mapping for
sustainable management is described by Liberman Cruz, who also explains in detail
the local traditional system for managing the domesticated native camelids that
depend on the wetlands. Also in Bolivia, Pedrotti applies the concepts of sigmeta
and geosigmeta to geobotanical mapping in three quite different environments,
presenting maps for evaluation of the conservation status of the quite different
vegetation of each area.
The last two chapters in this section treat the mapping of two quite different
aspects at of vegetation mapping at local scales. Raimondo presents some of the
first attempts to map plant diversity in Italy, with three examples from areas near
Palermo in Sicily. The mapping work involved photographs, field sampling, and
GIS software to assess biodiversity, degrees of endemism and naturalness. In cul-
tural landscapes, economic and environmental values must be combined with social
and cultural values for effective sustainable management. As a result, people’s
mental representations of remaining patches of more or less natural vegetation may
become important, especially fragments of forest. Roulier et al. consider how to
map people’s cognition of forest fragments in France.

Phytosociology

Traditional phytosociology still has its place, especially in more local and regional
vegetation studies. A good example is the updated phytosociological treatment
of the ultramafic vegetation of California by Sánchez Mata, which proposes some
new contributions to the taxonomic nomenclature for California. In the eastern
Carpathian Mountains of Romania, Oroian and co-authors use traditional phy-
tosociology to identify associations of Petasites and related tall forbs along mon-
tane riverbanks. Ramirez and co-authors use traditional phytosociology to identify
halophytic plant communities, especially those involving Arthrocaulon
(Arthrocnemum) macrostachyum along coastlines and in inland saline areas in the
Iberian Peninsula, Balearic and Canary Islands. In recent years, traditional phy-
tosociology has expanded to treat more integrative, synthetic aspects of vegetation
patterns in the field. This sub-field, known as symphytosociology, studies landscape
mosaics and provides the underpinning for the concepts of sigmeta and geosigmeta
already introduced in the preceding section on mapping. Bioret provides a good
introduction to the concept and uses of geosymphytosociology, as a tool for
Preface vii

bioevaluation. The use of symphytosociology as a tool for landscape monitoring

over time is explained by Theurillat et al., using a case study from the Swiss Alps.
This study, involving data from 1982 to 2014, combined a detailed list of phy-
tosociological vegetation units, aspects of physiognomic vegetation structure,
physical environmental factors, land use, and other human activities.

Ecology

Ecological analysis of vegetation may involove various methodologies, some

applied simultaneously, to understand the effects of climate, substrate, topography,
and other physical factors. Climatic envelope modeling is used by Box to suggest a
climatic basis for the apparent high diversity of plant life forms and vegetation
structures in Mediterranean regions. Also in the Mediterranean region, Acosta
considers habitat types and environmental constraints along gradients in Italian
coastal dune vegetation. In the warm-temperate climatic region of the southeastern
USA, there are only small areas of zonal evergreen broad-leaved forest. Climatic,
edaphic, and other constraints on this forest type, now and under global warming,
are considered by Box and Fujiwara, using relevé as well as physical data. Finally,
the role of riparian buffers in reducing pollution and ameliorating other conse-
quences of intensive agriculture in Romania is treated by Gheorghe and colleagues.
This detailed study considers hydrology, different agricultural crops, and nutrient
loads in different river basins.

Conservation

Conservation is often the ultimate goal of applied vegetation studies. This final
section presents three papers on management strategies and landscape evaluation
for conservation purposes. Especially in Asia but also elsewhere, sacred sites often
preserve natural vegetation. Sacred sites in northern Morocco are described by Deil
and co-authors, who ask critical questions concerning the effectiveness of such sites
for conserving the native flora, vegetation in general, and forest patches in partic-
ular. Cultural preservation can often be a useful strategy for conservation of natural
ecosystems. Fujiwara describes how this is done along the Kaga coast of central
Japan, where preservation of a historically significant coastal dune pinewood
landscape and of habitat for a unique, tree-climbing crab are combined for con-
servation of natural ecosystems. Finally, small islands differ from larger islands in
various characteristics, some affecting conservation strategies and biodiversity.
Different methods for measuring biodiversity, at different fine scales, are considered
by Domina and co-authors on small Mediterranean islands. They conclude that, on
small islands, distribution mapping may be the most effective method for assessing
biodiversity.

Elgene Owen Box

Franco Pedrotti
Contents

Mapping
Basic Semiological Principles to Define Coherent Colour Schemes
for Vegetation Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Gwenhael Perrin and Frédéric Bioret
Plant Species Distributions and Ecological Complexity:
Mapping Sampling-Effort Bias Explicitly . . . . . . . . . . . . . . . . . . . . . . . . 15
Carol X. Garzon-Lopez and Duccio Rocchini
Mapping of Biogeographical Territories: Flora, Vegetation
and Landscape Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Javier Loidi
Dynamic-Catenal Vegetation Mapping as a Tool for Ecological
Restoration and Conservation Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Javier Loidi
Climate Change, Wetland Management and Alpaca Pastoralism
in the Bolivian High Andes Mountains . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Máximo Liberman
Essays on Geobotanic Mapping in the Andes of Bolivia,
with Particular Reference to the Conservation Status
of the Vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Franco Pedrotti
Using Geobotanical Tools to Operationalize Mapping
and Assessment of Ecosystem Services (MAES)
in Southern Portugal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
S. Mesquita, J. Capelo, I. Gama, C. Marta-Pedroso, M. Reis,
and T. Domingos

ix
x Contents

Spatial Representation of Plant Diversity at Geographical Scale:

The Italian Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Francesco Maria Raimondo
Cognitive Mapping of Forest Fragments . . . . . . . . . . . . . . . . . . . . . . . . 177
Frédéric Roulier, Emilie Gallet-Moron, and Guillaume Decocq

Phytosociology
California Ultramafic Vegetation: A Phytosociological Update . . . . . . . . 187
Daniel Sánchez-Mata
Plant Associations of Petasition Officinalis Alliance in the East
Carpathians (Călimani and Gurghiu Mountains Romania) . . . . . . . . . . 205
Silvia Oroian, Mihaela Sămărghiţan, and Erzsébet Domokos
Arthrocnemum Macrostachyum Plant Communities in the Iberian
Peninsula, Balearic and Canary Islands (Spain and Portugal) . . . . . . . . 231
Esteban Ramírez, Lourdes Rufo, Daniel Sánchez-Mata,
Irene Sánchez-Gavilán, and Vicenta de la Fuente
Geosynphytosociological Typology of French Atlantic Coastal
Rocky-Cliff Vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Charlotte Demartini and Frédéric Bioret
Symphytosociology, a Tool for Landscape Monitoring: A Case Study
from the Swiss Alps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Jean-Paul Theurillat, Michele Di Musciano, Olivier Duckert,
and Claude Béguin

Ecology
Why Does Mediterranean Vegetation Seem So Diverse? . . . . . . . . . . . . 287
Elgene O. Box
Coastal Dune Vegetation Zonation in Italy: Squeezed Between
Environmental Drivers and Threats . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Alicia T. R. Acosta
Constraints on Evergreen Broad-Leaved Forests in the Southeastern
United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
Elgene O. Box and Kazue Fujiwara
Role of Riparian Zones in Reducing Pollution of Surface
and Ground Water, Increase Agricultural Production and Nutrient
Acquisition and Storage in River Catchments . . . . . . . . . . . . . . . . . . . . 365
Iuliana Florentina Gheorghe, Daniela Strat, and Simona Mihailescu
Contents xi

Conservation
Vegetation and Flora of Sacred Natural Sites in North-Western
Morocco—Landscape Context and Conservation Value . . . . . . . . . . . . . 391
Ulrich Deil, Birgit Frosch, Holger Jäckle, and Allal Mhamdi
The Kaga Coast (Japan): A Natural Ecosystem and Cultural
Landscape Ensuring Biodiversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
Kazue Fujiwara, Elgene O. Box, Jin Ōsaka, Hideki Sakai,
and Youji Kitaguchi
Measurement and Conservation of Species and Habitat Biodiversity
at Different Scales on Small Mediterranean Islands . . . . . . . . . . . . . . . . 445
Gianniantonio Domina, Anna Maria Mannino, Francesco M. Raimondo,
and Patrizia Campisi
Mapping
Basic Semiological Principles to Define
Coherent Colour Schemes
for Vegetation Mapping

Gwenhael Perrin and Frédéric Bioret

Abstract Colour is the most used visual element in cartography for symbolising
vegetation because, more than many other signs, it translates information in an
implicit and condensed way. While the diversity of vegetation and thematic mapping
preclude the possibility of a universal colorimetric norm, certain pioneer works have
tended to establish a methodological framework allowing the rendering of ecological
and bioclimatic conditions. Thus, the evolution of colorimetric models, linked to the
progression of infographic tools, make it possible to consider novel approaches.
This article proposes an intuitive method for using colorimetric parameters to
represent phytosociological and synphytosociological units. It allows, through
colour only, without any overlay, the representation of three to four ecological
dimensions. Hue and luminosity are used for representing vegetation types or
vegetation series and their physiognomy. Saturation, an overlooked parameter in
mapping, represents the level of anthropization. Used in GIS applications, the
system permits easy establishment of numerically calibrated palettes and con-
tributes to improving map understanding.

Keywords Semiology Cartography
Vegetation series mapping
Colours

Ecological and bioclimatic factors

1 Introduction

Vegetation maps are good tools for scientific knowledge and research but also
represent tools of communication for applications to land management, as they can
be presented to stakeholders and agencies dealing with different types of projects
related to land management or elaboration of nature conservation strategies. Maps
have to be easily understandable if their elaboration integrates basic rules.

G. Perrin (&)
F. Bioret

Univ Brest, EA 7462 Géoarchitecture, 29200 Brest, France
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 3

F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_1
4 G. Perrin and F. Bioret

Fig. 1 Colorimetric space of Gaussen published in Ozenda [9]. The different colours of the series
are located in a 3D volume, with a red-black axis, a yellow-brown axis and a blue vertical axis

A few authors have developed methodological approaches to harmonise and

make more obvious to the reader the symbology of vegetation maps. Gaussen [4, 5]
proposed the use of shades of hues in mapping vegetation series. The CNRS 1:200
000 map of France was used this principle to represent different dynamic stages by
the same hue as the climax community but with colours ranging from light to dark
in order to show the gap between the climax and the actual vegetation. Ozenda [9]
introduced the notion of chromatic synthesis and preliminary colorimetric princi-
ples (Fig. 1), which, however, presents some incoherencies. For example, bistre, i.e.
blackish brown, symbolises vegetation types on organic soils when the balance
between black for shadow and red for light generates a colour close to bistre. The
Ozenda system also presents counter-intuitive choices for the way colours are
organised in more recent systems. More recently, Pedrotti [11] proposes two other
ways to represent vegetation: use of monochrome palettes, from white to black,
with superposition of hatching; use of colours without pre-established rules.
The French national CarHAB program aims to map habitats and vegetation
series at the 1:25,000 scale all over the metropolitan territory up to 2025 [8]. For
this, a preliminary theoretical reflection on semiology and vegetation mapping has
been developed [12] and integrated into a methodological guide.
In his historical reference book, Bertin [1] proposed basic rules for colour use in
representing qualitative or quantitative data, taking our perceptions into account. He
explains that, according to its terminology, the colour—or hue—has a semiological
value only if we can distinguish its colour (in other words, its hue) from its value
(amount of black in the hue). The two graphical variables do not have the same
symbolic properties and cannot be used for the same type of semiological components.
In other words, the hue is not ordered because of its selective force that allows the eye
to distinguish entities as belonging to different categories. It allows representing
quantitative components, as in the case of vegetation types defined according to a
categorisation based on ordinal to quantitative data analysis. Always following Bertin,
an ordinal scale can nevertheless be foreseen within a colour temperature range. In this
Basic Semiological Principles to Define Coherent Colour Schemes … 5

case, the eye is able to establish an order relation, contrarily to schemes based on the
whole chromatic spectrum, from red to purple. On the contrary, the value suits ordinal
components for which it is possible to establish a hierarchical relation between cate-
gories. Numerous semi-quantitative ecological variables expressed along gradients
(trophy, humidity, bioclimatic belts…) can be linked to this component type.
Thus, after having reviewed some aspects of the colour symbolism applied to
vegetation and having presented the colorimetric models commonly used in com-
puter graphics, we will make some proposals useful for the development of
semiological charts.

2 The Colour Metric in Infographics

Among the numerous available colorimetric spaces, three models are widely used
for colour coding by computers. Each model has three parameters: RGB, HSL and
HSV. Different proportions of the three base colours allow reproduction of all the
possible colour impressions.
The RGB system reconstitutes a colour by mixing in additive synthesis the three
basic colours, red (R), green (G), and blue (B). If the RGB model represents a main
advantage in computer coding and storage, it adopts a purely physical conception and
seems less adapted to the human colour vision that we do not perceive as an addition of
primary colours but rather as a complex mixing of sensations associating brightness as
perception of luminous flux intensity and chromaticness as perception of chromacity
factors. Chromaticness results from two visual sensations: the hue that sensitively
renders the dominant wavelength and saturation that renders purity of the colour.
The other two coding systems, HSL and HSV, developed by Smith [14], are
based precisely on this colour perception, as they are said to be perceptual. They
use two common variables, the hue of colour (H), in which all the shades are
included inside a circle, and its saturation (S), varying from 0 to 100%. The result is
defined according to a sensibly different method for each of the two models. The
third parameter differs in the two models: HSV uses the value (V) of the colour, i.e.
the brightness (the distance of one hue from black), while HSL uses luminosity (L),
i.e. the degree of darkness or lightness. The main distinction between the two
models corresponds to the way totally saturated hues are treated. The variables V
and L range between 0 and 1 (Fig. 2).
Fort each hue, the variation of luminosity and purity range can result in four
main colour types: bright (light and saturated), pale (light and unsaturated) deep
(dark and saturated) and dull (dark and unsaturated).
If the HSV and HSL model conceptions are closed and offer the possibility of
varying the saturation and intensity without affecting hue, nevertheless some dif-
ferences make HSL preferable for mapping applications. If pure black (L = 1 and
V = 1) is located at the base of the cylinder for both models, white is located at the
top of the cylinder for HSL (L = 1) and at V = 1 and S = 0 for HSV. The brightest
colours for HSV (V = 1) are equivalent to intermediate luminosity colours in HSL
6 G. Perrin and F. Bioret

Fig. 2 HSL and HSV models (top: radial representation in 3 dimensions; middle: HL and HV
spaces; bottom: colour wheel and SL spaces for two different colours)

(L = 0.5). HSL appears to be a symmetric system offering a wider range along the L
axis, from black to the selected hue to white. It seems to be more intuitive for
varying colour intensity (form dark to light). At constant hue, saturation gradients
are dependent on the luminosity in HSL, which limits intensity variations from one
extreme to the other, reducing the differences of perception.
The most commonly used GIS (ArcGIS, MapInfo, QGIS…) integrate these
models in a more or less complete way, allowing precise numerical parameteriza-
tion of colours. RGB and HSV are usually available, while HSL is less common. Its
use requires a conversion with easily available algorithms. The following equations
do the HSL-to-HSV conversion:

H HSV ¼ H HSL

2L þ SHSL ð1  j2L  1jÞ

V¼
2
2ðV  LÞ
SHSV ¼
V
Basic Semiological Principles to Define Coherent Colour Schemes … 7

3 Colour Symbolism and Its Potential Use in Vegetation

Mapping

It appears to be possible to translate the impression given by an object (natural,

artificial, feeling, state…) by varying the three parameters according to a colori-
metric language. This is the most common confusion in vegetation mapping: lack of
explicitly established rules, at least based on a feeling or a shared symbolism, even
if perceptions can differ inside the occidental world and over historical time [3, 10].
This is why we often recognise invariants in colour choice (red for warm and dry,
blue for cold and wet) but also a broad range of choices from one cartographer to
the other, based on the phenomena to be represented (edaphic gradients vs biocli-
matic gradients) or representation modes. For all these reasons, we propose a
flexible system to harmonise vegetation mapping semiology.
Table 1 proposes a correspondence between colorimetric parameters, symbol-
ism, and vegetation or vegetation series for each opposed value of a colorimetric
parameter (warm hue vs cold, saturated vs unsaturated colours…). The ecological
significance is considered in the light of symbolic interpretation, which is generally
realised according to what the colours seem to indicate. The proposed ecological
and bioclimatic terms can be adapted easily to the cartographer’s set of themes and
to local context.

Table 1 Symbolic and ecological interpretation of the three colorimetric parameters

Variable Value 1 Value 2

Symbolic and ecological / bioclimaƟc Symbolic and ecological / bioclimaƟc
Colorimetric value
equivalence equivalence
Sky, air, atmosphere, cold, sea, water, Fire, heat, drought, oxidaƟon
humidity, reducƟon
Cyan Red
Cryophile, (edapho-) hygrophile, hydrophile, Thermophile, (edapho-) xerophile,
(H=180°) (H=0°)
flooded, pergelid, ultrahyperhumid, boreal drained, infrathermic, ultrahyperarid,
and polar, oceanic tropical and mediterranean, desert
Planet Earth, darkness, night, humidity, Sun, shine, light, clarity, drought,
reducƟon oxidaƟon
Blue Yellow
Hue
Sciaphilous, (edapho-) hygrophile, (H=240°) (H=60°) Heliophilous, (edapho-) xerophile,
hydrophile, flooded, ultrahyperhumid, drained, ultrahyperarid, tropical and
polar, series mediterranean, permaseries
UnidenƟfied natural symbolism VegetaƟon, vitality, ferƟlity, growth
Magenta Green
In opposi on to green: unproduc ve, stress Produc ve, compe ve, eutrophic,
(H=300°) (H=120°)
tolerant, oligotrophic, acidophilic, basiphile, neutrophilic, mull, limnic, zonal,
mor, euhaline, azonal, intrazonal extrazonal
Dark, shady, cloudy, heavy, deep, carbon Bright, sunny, light, surficial
Dark Light
Luminosity
Closed, high, dense, evergreen, sciaphilous, (L=0) (L=1) Open, low, sparse, deciduous,
series, organic heliophilous, permaseries, mineral
Dull, bland, exƟnguished, monotonous, sad, Bright, vibrant, vibrant, varied,
impure, dusty, poor Unsaturated Saturated cheerful, pure, rich
Satura on
Oligotrophic, ruderal, secondary, ar ficial (S=0) (S=1) Eutrophic, undisturbed, primary,
natural
8 G. Perrin and F. Bioret

4 Setting of Colorimetric Variables

4.1 General Principle

Hue appears to be the main way to represent particular ecological factors.

A colorimetric gradient can be produced by the opposition of divergent hues for
elaborating a colorimetric spectrum (e.g. gradient from red to cyan). The temper-
ature contrast between warm and cold colours is mainly used [6]. The polar
chromatic of the range should be adapted for mapping themes and the cartogra-
pher’s objectives (Table 1).
According to needs, two other colorimetric variables, L and S can be added as
secondary ecological dimensions in order to represent, respectively, physiognomy
—in the way Itten considers the progressive light-dark contrast—and the level of
vegetation or vegetation-series degradation, by means of quality contrast (see
horizontal and vertical axes of Fig. 3). This method permits one, in a certain way, to
reconstitute the state of a dynamic system. Other sets of themes can also be
developed, always by following the elements listed in Table 1.

Stronger (S ↗)

Darker (L ↘) Lighter (L ↗)

Weaker (S ↘)

Fig. 3 The four divergent colours of a same hue

Basic Semiological Principles to Define Coherent Colour Schemes … 9

5 Symbolisation of First-Order Ecological Dimensions

The construction of a hue range can follow some simple rules, making it possible to
use the entire available spectrum and to obtain the best balance possible.
In the case of a single ecological variable, the cartographer will define two hues
located at the extremes of the spectrum in order to define the palette.
In the case of two ecological variables and in order to exploit one half of the
chromatic circle, a nominal variable is defined first, in order to subdivide the circle
into two parts (e.g. acidophilous in magenta vs neutrophilous in green, oligotrophic
vs eutrophic, azonal vs zonal…) (Fig. 4).
A second ecological axis can be defined between two other chromatic poles, in
order to symbolise an ordinal variable (e.g. from xerophilous in red to hydrophilous
in cyan). Each vegetation type or vegetation series can then be located on the circle
according to its values for the two variables.
In certain cases it is necessary to divide the circle in other ways. One triad could
be used to symbolise zonality, according to a humidity gradient and the following
scheme: from yellow to cyan for zonal vegetation or vegetation series (around
green), from cyan to magenta for azonal tending to hygrophilous vegetation (around
blue), and from magenta to yellow for azonal tending to xerophilous vegetation
(around red).

Hyperxerophile Hyperxerophile Azonal edaphoxerophile Thermomediterranean

(H=350°) (H=10°) (H=300°) (H=0°)
Xerophile Xerophile N
A (H=340°) (H=20°) Azonal edaphohygrophile Thermotemperate
e
c (H=240°) (H=22.5°)
u
i
t
d Mesophile Mesophile
r
i (H=310°) (H=80°) Cryorotemperate Mesomediterrnean
o
p (H=180°) (H=45°)
p
h
h
i Hygrocline Hygrocline i
l (H=290°) (H=100°) l Orotemperate Mesotemperate
o
o (H=157.5°) (H=67.5°)
u
u
s Hygrophile Hygrophile
s
(H=210°) (H=150°) Oromediterranean Supramediterranean
AquaƟc AquaƟc (H=135°) Supratemperate (H=90°)
(H=190°) (H=170°) (H=112.5°)

Fig. 4 Examples of chromatic symbolizations (diagram on the left: symbology adapted to

vegetations according to a humidity gradient: left hemisphere coding for acidiphilic vegetations,
right for neutrophilic ones; right diagram: complex symbology adapted to zonal series following a
thermal gradient and to azonal series following a moisture series gradient: green-tending colours
code for climatophilic series, magenta-tending colours correspond to azonal series)
10 G. Perrin and F. Bioret

6 Symbolisation of Physiognomic, Dynamic

and Anthropogenic Dimensions

Once the hues are defined, luminosity and saturation levels are fixed. These two
parameters are usually applied to ordinal type variables, and ecological factors to be
described are categorised above all according to a quantitative scale. Thus, colours
range from pale for more open vegetation, to intermediate luminosity for herba-
ceous and shrubby formations, to dark for closed forests (Table 2). It is similar for
the three main vegetation series types of Rivas-Martínez (Table 3).
By the colour language described above, primary vegetation types and low
levels of disturbance are represented by saturated colours, and secondary vegetation
and high levels of disturbance by unsaturated colours; achromatic colours (i.e. grey)
correspond to artificial environments (Table 4). The openness or density of artificial
environments is represented in the same way, by a colour gradient from light to
dark (Table 5). Addition of the saturation variable eventually provides a great
variety of palettes, as illustrated by Table 6 and Table 7.

Table 2 Colorimetry of formations types

S=1 H
0 30 60 90 120 150 180 210 240 270 300 330 360
Sparse annual herbaceous 0.8
Open perennial herbaceous 0.7
Closed perennial herbaceous 0.6
Dwarf shrubby 0.5 L
Shrubby 0.4
Open or pioneer woody 0.3
Closed woody 0.2

Table 3 Colorimetry of series types

S=1
H

0 30 60 90 120 150 180 210 240 270 300 330 360

Permaseries 0.75
Minoriseries 0.50 L
Series 0.25
Basic Semiological Principles to Define Coherent Colour Schemes … 11

Table 4 Colorimetry of the degree of artificialization of a vegetation type

L=0.5 H
0 30 60 90 120 150 180 210 240 270 300 330 360
Primary 1.00
Replacement secondary 0.75
Regressive secondary 0.50 S
Anthropogenic 0.25
Ar ficial 0

Table 5 Colorimetry for artificial units

S=0
Annual crops 0.8
Perennial herbaceous crops, temporary meadows 0.6
L
Shrub planta ons 0.4
Tree planta ons 0.2

Table 6 Colorimetry of the formation type according to the degree of artificialization

H1=0° H2=180°
S
1.00 0.75 0.50 0.25 0.00 0.25 0.50 0.75 1.00
Sparse annual herbaceous 0.8
Open perennial herbaceous 0.7
Closed perennial herbaceous 0.6
Dwarf shrubby 0.5 L
Shrubby 0.4
Open or pioneer woody 0.3
Closed woody 0.2
12 G. Perrin and F. Bioret

Table 7 Colorimetry of the degree of artificialization according to the formation type

H1=0° H2=180°
L
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Primary 1.00
Replacement secondary 0.75
Regressive secondary 0.50 S
Anthropogenic 0.25
Ar ficial 0.00

7 Additional Symbologies: Outlines, Overlays

and Transparencies

Maps are not only constituted of solid colours. They may also integrate other
symbolic elements to provide complementary thematic information that is useful to
the cartographer’s discourse, or more simply to the spatial localisation that can
guide the reader’s eye. Use of these other symbolic elements requires some rules.
An area outline represents a limit or a boundary between two polygons and is
shown by a line. The nature of this limit could potentially be represented by one or
several graphic variables specific to linear forms (thickness, form, grain…).
Nevertheless, since polygons are generally contiguous to other polygons, setting up
such a methodology seems complicated. In order not to multiply superfluous signs,
it appears preferable to suppress less visible outlines when surfaces become smaller
or when the scale increases. Their use could be reserved for particular cases such as
insufficient contrast between adjacent colours or for thematic information useful for
locations, such as administrative entities.
The use of fill symbols overlying flat tints allows restitution of additional
information. These are composed of various visual variables: form (dots, lines,
symbols), orientation of the form, grain, or colorimetric parameters [1]. In order to
make the message less confused, overlays should be used with moderation, for two
reasons. The first risk is to overcharge the senses by multiplying information from
different themes. The second risk is to create an overcharge of signs. In general,
overlays should be complementary and not redundant with the main colour raster.
In principle, overlays should not be used to describe vegetation types or vegetation
series, but rather should be used for additional information useful for interpreting
the map. Different topics may be involved:
• for disturbance: grazing in dots, mowing in hatching, fire…; grain can express
the intensity of the phenomenon;
• dynamic facies: colonisation by shrubs or trees can be represented by dots or
symbols of different grains according to the level of colonisation;
Basic Semiological Principles to Define Coherent Colour Schemes … 13

• pedologic or geologic information: turf, sand, lithology, etc. can be represented

by filling polygons with dots or other more suitable symbols.
It also seems important to ensure that the map integrates enough landmarks to
offer the best reading conditions for the reader. Geographic maps tend from the
outset to erase remarkable landscape elements, by generalising information. This
aspect is often neglected, with visual elements poorly designed (e.g. topographic
raster maps overprinted on thematic map). Weighted addition of vector data can
represent a solution in this case. Another approach consists in using transparencies
by merging thematic vegetation layers with a raster background (e.g. shaded relief
generated from DEM, aerial photographs), thanks to layer-rendering tools of the
mapping software, which improves locations [2, 13].

8 Conclusion

Despite the lack of bibliographic references and of any uniformly recognised

semiological chart for vegetation mapping, it seems interesting to propose some
principles for increasing the readability of maps, especially for integrated mapping
of vegetation series and dynamic maps. In any case, the colour choice linked to an
unequivocal typology appearing in the legend is a preliminary principle in vege-
tation mapping [7].
Also, the main purely semiological question is: can palettes be harmonious? If a
landscape is heterogeneous, due to its ecological heterogeneity and often to its high
level of human impact, is it preferable a priori that mapping appears disharmonious.
Potentials of the colorimetric spaces of Munsell or the Natural Colour System
elaborated for colour harmony more adapted to the human eye should be explored in
order to check whether their adaptation to represent vegetation is possible.

References

1. Bertin J (2005) Sémiologie graphique. Les diagrammes, les réseaux, les cartes, 4th edn.
Editions EHESS, Paris
2. Bioret F, Caillon N, Glemarec E (2014) Cap d’Erquy et Cap Fréhel. Doc Phytosociologiques,
s.3, 1:25–65
3. Caivano JL (1998) Color and semiotics: a two-way street. Color Res Appl 23:390–401.
https://doi.org/10.1002/(SICI)1520-6378(199812)23:6%3c390::AID-COL7%3e3.0.CO;2-%
23
4. Gaussen H (1936) Le choix des couleurs dans les cartes botaniques. Bull De La Société
Botanique De France 83:474–480. https://doi.org/10.1080/00378941.1936.10837343
5. Gaussen H (1947) L’emploi des couleurs en cartographie. Comptes Rendus Hebdomadaires
des Séances de l’Académie des Sciences 224:450–452
6. Itten J (1961) Kunst der Farbe. Otto Maier, Ravensburg
7. Küchler AW (1967) Vegetation Mapping. Ronald Press Company, New York
14 G. Perrin and F. Bioret

8. Lalanne A (2016) Le programme de cartographie nationale des végétations naturelles et

semi-naturelles de France: CarHAB. Doc Phytosociologiques 6:236–244
9. Ozenda P (1986) La cartographie écologique et ses applications, p 159
10. Pastoureau M, Simonnet D (2005) Le petit livre des couleurs. Panama, Paris
11. Pedrotti F (2013) Plant and Vegetation Mapping. Springer, Heidelberg
12. Perera S (2016) Compte-rendu du groupe de travail sur la sémiologie CarHab.
Représentations graphique et cartographique de la végétation
13. Perrin G (2014) Etang du Moulin Neuf. Doc phytosociologiques, s.3, 1:123–154
14. Smith AR (1978) Color gamut transform pairs. ACM SIGGRAPH Comput Graph 12:12–19.
https://doi.org/10.1145/965139.807361
Plant Species Distributions
and Ecological Complexity: Mapping
Sampling-Effort Bias Explicitly

Carol X. Garzon-Lopez and Duccio Rocchini

Abstract In geobotany, mapping plant species distributions properly is crucial to

guarantee a proper estimate of their dispersal variability in space and time, also
considering habitat suitability. In most cases, uncertainty in the modelling proce-
dures has been disregarded. However, hidden uncertainty or bias may hamper
robust estimates of the distribution of plant species or species assemblages. In this
paper, we propose an approach to mapping uncertainty properly, mainly deriving
from sampling effort bias, when mapping plant species distributions.

Keywords Biodiversity
GBIF
Sampling schemes
Uncertainty

1 Introduction

In ecology, a number of studies have dealt with the prediction of species distri-
bution and diversity over space and its changes over time, based on a set of
environmental predictors related to environmental variability, productivity, spatial
constraints and climate drivers [10].

C. X. Garzon-Lopez
Ecology and Vegetation Physiology Group (EcoFiv), Universidad de los Andes,
Cr. 1E No 18A, Bogotá, Colombia
D. Rocchini (&)
Center Agriculture Food Environment, University of Trento, Via E. Mach 1, 38010 S.
Michele All’Adige, TN, Italy
D. Rocchini
Centre for Integrative Biology, University of Trento, Via Sommarive, 14, 38123 Povo, TN,
Italy
D. Rocchini
Research and Innovation Centre, Department of Biodiversity and Molecular Ecology,
Fondazione Edmund Mach, Via E. Mach 1, 38010 S. Michele All’Adige, TN, Italy

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 15

F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_2
16 C. X. Garzon-Lopez and D. Rocchini

Species distribution models have been acknowledged as the most powerful

methods to map the spread of plant (and animal) species. The basic approach used
to create maps based on predictors is to rely on linear models to create gridded
landscapes of potential distribution of species based on point or local polygon
data [5].
In most cases, the output is a density function in two dimensions representing the
distribution Sx of species x. In general, boundaries are sharply defined based on
thresholds of predictors/factors (e.g. when mainly based on land cover) or con-
tinuous, if based on the continuous variability of predictors (e.g. the continuous
variation of temperature).
Uncertainty in such models derives mainly from pseudo-absences in input data
[8] as well as from model bias, i.e. the error deriving from the model being chosen
(GAM, GLM, Maximum entropy models, etc.,). Hence, the representation of
uncertainty in two dimensions is strongly suggested but is disregarded in most
cases. However, its importance is apparent. In fact, areas with a high or low
probability of species occurrence might also be related to a high or low error rate.
As an example, in case of invasive species, it might be crucial to represent
uncertainty spatially, in order to allow better decision making.
As an example, a high (or low) invasion rate might be related to high or low
errors in the output model being used by decision makers. The most dangerous
situation occurs when a low predicted invasion rate is related to a high error in the
modelling procedure. In this case decision makers might underestimate the effort
needed against invasion, suspected to be low from the species distribution map.
In this case, misconceived use of a species distribution map might be dangerous,
for example in case of a low dispersion probability of an invasive species but with a
high error in the model. The prediction of the distribution of an invasive species
might be low but with a high error; hence its spread could be underestimated in
some parts of an area.
The aim of this study is to provide a straightforward and robust mapping pro-
cedure to show explicitly the spatial uncertainty related to sampling problems like
sampling effort (relying on commonly used datasets) when dealing with species
distribution modelling.

2 Uncertainty Related to Sampling Effort: Representing

Sampling Bias by a Diffusion-Density Algorithm

For bias related to sampling effort, we will rely on one of the most used datasets in
biodiversity study at large spatial extents, namely the Global Biodiversity
Information Facility (GBIF) dataset (https://www.gbif.org/). GBIF data comprise a
huge range of species occurrence observations collected with a wide variety of
sampling methods, ranging from well-established plot censuses to direct observa-
tions collected during field trips. Consequently, some of the data points are at the
Plant Species Distributions and Ecological Complexity … 17

center of censused grids (each point comprises the species located at a specific-size
quadrant) or correspond to single observations of one (or more) individuals of the
same species. These differences also depend on the methodologies used to observe/
record occurrences per taxon. Plots, and plots within transects, are common designs
in vegetation censuses, while transects, point counts and live traps are preferred in
the case of animals.
The variation in factors, such as per country biodiversity monitoring schemes,
funding schemes, focal ecosystems, accessibility to remote areas, and others, adds
another source of variation, especially at multinational scales [1].
Undoubtedly, all those sources of variation result in non-homogeneous sampling
that has important consequences not only for the development of accurate species
distribution maps but, more importantly, for conservation and management deci-
sions [14]. The aim of this study was to show explicitly the spatial uncertainty in the
sampling effort of the GBIF data.
In this study we aim to quantify and map the uncertainty derived from the
variation in observations due to differences in sampling efforts. In particular, car-
tograms are used, in which the shape of objects (countries and grid cells) is directly
related to a certain property, in our case to uncertainty. Cartograms build on the
standard treatment of diffusion, in which the current density is given by:

J ¼ vðr; tÞpðr; tÞ ð1Þ

where v(r, t) and p(r, t) are the velocity and density at position r and time t. See
Gastner and Newman [9] for additional information.
Cartograms facilitate the visualization of spatial uncertainty in the results by
changing the size of the polygons based on the density of information contained
(number of observations, variation, etc.). In this paper, the cartograms were
developed using the free, open-source software ScapeToad (https://scapetoad.
choros.ch/).
After having superimposed a grid on the GBIF set over Europe, we modified the
shape of different grid cells based on their relative oversampling, quantified as the
number of dates in which a cell was sampled. In this, i) the colour of the map cells
represents the occurrence possibility of a certain species; while ii) cell size is related
to the sampling effort. The greater the distortion, the higher will be the number of
visits in a certain area, hence the sampling effort bias.
By this strategy, the maps generated show the differences in species observations
per area in Europe for three different plant species: Quercus coccifera, Fagus
sylvatica and Betula pendula. Their differences in relative sampling bias are shown
in Fig. 1.
18 C. X. Garzon-Lopez and D. Rocchini

Fig. 1 Examples of
cartograms for different plant
species, from those sampled
with a lower sampling bias to
those based on cells sampled
several times, hence with high
sampling bias on occurrences
of the species
Plant Species Distributions and Ecological Complexity … 19

3 Discussion and Open Challenges

Mapping sampling bias properly is a crucial step when aiming to develop robust
species distribution models or to represent species variability over space (and time).
Various methods have been proposed to account for uncertainty in such models.
One of the most important is certainly related to Bayesian statistics, in which the
model can be’informed’ with previous knowledge related to the distribution of a
certain species. Uncertainty, in our case based on sampling effort bias, might then
be directly part of the species distribution model, as additional information about
spatial bias. This could be particularly important when forecasting species distri-
bution in time, directly taking into account potential oversampling and avoiding
hidden uncertainty [13].
On the other hand, tests of oversampling could be performed by relying on
virtual species in order to estimate synthetically the amount of effort due to properly
sampling a certain species and to avoid oversampling, and hence a higher possi-
bility of occurrence. This is true considering both single species distributions and
species assemblages, in which complexity is expected to increase exponentially. In
such cases, correlative distribution models could benefit for complete control during
the experimental phase of input and response (e.g. species or assemblage distri-
bution) variables [7].
Mapping of different objects in 2-D space has always been a peculiar problem in
geobotany and in cartography more generally [2]. This is particularly true when
dealing with the distribution over space of plant species and community assem-
blages [6]. We think that the cartogram-based approach proposed in this paper has
an explicitly high robustness and an implicitly high reproducibility, since it is based
only on free, open-source algorithms. We refer to Keil et al. [12] for a dissertation
about the development of free, open-source concepts in recent years.
Besides free, open-source methods, free data are crucial for further development
of proper species distribution models, including both free environmental data and
joint field data archiving procedures [11]. Future developments in SDMs should
include the visualization of species distribution model predictions combined with
the map of uncertainty presented here. This could lead to a proper spatial
cross-correlation investigation to check for potential autocorrelation in the sampling
process [3]. This is true when considering mapping of plant species, and more so
for vegetation properties at global scales (e.g. [4]).

References

1. Barbosa AM, O’Hara RB (2015) fuzzySim: applying fuzzy logic to binary similarity indices
in ecology. Methods Ecol Evol 6:853–858
2. Baselga S (2018) Fibonacci lattices for the evaluation and optimization of map projections.
Comput Geosci 117:1–8
20 C. X. Garzon-Lopez and D. Rocchini

3. Chen Y (2015) A new methodology of spatial cross-correlation analysis. PLoS ONE 10:
e0126158
4. Crowther TW, Glick HB, Covey KR, Bettigole C, Maynard DS, Thomas SM, Smith JR,
Hintler G, Duguid MC, Amatulli G, Tuanmu M-N, Jetz W, Salas C, Stam C, Piotto D,
Tavani R, Green S, Bruce G, Williams SJ, Wiser SK, Huber MO, Hengeveld GM, Nabuurs
G-J, Tikhonova E, Borchardt P, Li C-F, Powrie LW, Fischer M, Hemp A, Homeier J, Cho P,
Vibrans AC, Umunay PM, Piao SL, Rowe CW, Ashton MS, Crane PR, Bradford MA (2015)
Nature 525:201
5. Elith J, Graham CH (2009) Do they? How do they? WHY do they differ? On finding reasons
for differing performances of species distribution models. Ecography 32:66–77
6. Feilhauer H, Schmidtlein S (2011) On variable relations between vegetation patterns and
canopy reflectance. Ecol Inform 6:83–92
7. Fernandes RF, Scherrer D, Guisan A (2018) How much should one sample to accurately
predict the distribution of species assemblages? A virtual community approach. Ecol Inform
48:125–134
8. Foody GM (2011) Impacts of imperfect reference data on the apparent accuracy of species
presence-absence models and their predictions. Glob Ecol Biogeogr 20:498–508
9. Gastner MT, Newman MEJ (2004) Diffusion-based method for producing density-equalizing
maps. PNAS 101:7499–7504
10. Guisan A, Thuiller W (2005) Predicting species distribution: offering more than simple habitat
models. Ecol Lett 8:993–1009
11. Huang X, Qiao G (2011) Biodiversity databases should gain support from journals. Trends
Ecol Evol 26:377
12. Keil P, MacDonald AA, Ramirez KS, Bennett JM, Garcia-Pena GE, Yguel B, Bourgeois B,
Meyer C (2018) Macroecological and macroevolutionary patterns emerge in the universe of
GNU/Linux operating systems. Ecography 41:1–13
13. Ramin M, Arhonditsis GB (2013) Bayesian calibration of mathematical models: Optimization
of model structure and examination of the role of process error covariance. Ecol Inform
18:107–116
14. Rocchini D, Hortal J, Lengyel S, Lobo JM, Jiménez-Valverde A, Ricotta C, Bacaro G,
Chiarucci A (2011) Accounting for uncertainty when mapping species distributions: The need
for maps of ignorance. Prog Phys Geogr 35:211–226
Mapping of Biogeographical Territories:
Flora, Vegetation and Landscape
Criteria

Javier Loidi

Abstract One of the tasks of biogeography has traditionally been the definition of
territorial units by means of their biotic content. Five main ranks have been rec-
ognized, in a downward scale: kingdom, region, province, sector and district. It is
assumed that for this purpose biogeography can be used as a synonym of phyto-
geography, as plants are particularly indicative of the substrate (soil, bedrock,
geomorphology) and atmospheric (climatic) conditions. In this chapter, the way in
which traditional plant taxonomical information is enriched with information about
vegetation (plant communities) and a certain type of landscape analysis are
explained, as far as they reflect the conditions prevailing in each territory. Floristic
specificity reaches the level of family at kingdom level, while it decreases as we
move towards the lower ranks, being null in the districts. Plant communities have
high specificity at the level of formation at kingdom level with all the classes being
endemic at this level; at the destrict level of district on the other hand, the specificity
is null. Concerning landscape, the analysis of the repeated ordered arrangement of
vegetation units across a gradient (vegetation zonation) in each territory is con-
sidered. The normal topographic zonation is the crest-slope-valley bottom model
(geosigmetum). The rest of the eco-topographic zonations, i.e. those considered
exceptional (permageosigmetum), are also considered. They differ minimally
between the lower rank units, the districts, i.e. they differ in at least one geoper-
masigmetum. Combining information from three sources: taxonomy, vegetation
and landscape analysis, it is possible to establish efficiently accurate and precise
limits between territorial units.

Keywords Biogeography Provincialism Floristic element Phytogeographical




element Geoflora Geosigmetum Biogeographic kingdom Biogeographic





region Biogeographic province Biogeographic sector Biogeographic district

J. Loidi (&)
Department of Plant Biology and Ecology, University of the Basque Country (UPV/EHU),
P.O. Box 644, 48080 Bilbao, Spain
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 21

F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_3
22 J. Loidi

1 Introduction. Concept of Biogeography

The term Biogeography is understood by the scientific world to mean a discipline

that deals with the distribution of living beings in a geographical scale; in words of
Brown and Lomolino [10], it attempts to document and understand spatial patterns
of biodiversity. It could be defined as the study of the geographical distribution of
organisms, but this simplicity hides the great complexity of the subject [31] as it
attempts to find patterns of distribution in both the present and in the past [23]. Such
patterns of species distribution are usually explained by means of a set of historical
factors such as speciation, extinction, continental drift and glaciation, in combi-
nation with current environmental conditions, such as climate, substrate, geography
and orography, thus, the current distribution of species over the earth is the result of
the current and the past physical and chemical conditions of the substrate and
atmosphere. The inclusion of comprehensive timescale has been part of the defi-
nition since the early geobotanists, such as A. Grisebach [24] and Ch. Flauhault
[20]. Therefore, Biogeography is devoted to the study of the distribution of species
and ecosystems across geographical space and through (geological) time.
If we focus on the current distribution of a taxon, we see that it is the result of: a)
historical processes, such as the origin and location of the lineage and its evolution,
in the frame of climatical and geographical changes; b) the dispersal possibilities,
determined by the plant´s abilities (anemochory, endozoochory, barochory, etc.)
and the geographical barriers and migratory pathways; and c) the physiological and
current ecological conditions enabling its survival [35]. The former two can be
considered the core of the genuine “biogeographical” component of the concept
while the ecological part deals with the current environmental opportunities and
constraints driving species distribution. Both aspects, equally important, need to be
clearly separated. Certainly, the distribution of living beings is driven mainly by
ecological factors, but the other factors, i.e. the dispersal capacity of the organisms
and their evolutionary history in the geologic time scale, are also necessary to
explain their distribution at a geographical scale. Not considering the genuinely
biogeographical factors has led to a widespread confusion between biogeography
and ecology, to the extent that many texts hardly distinguish between them [46],
and the boundaries between both disciplines have become blurred [30].
Biogeography, sensu stricto, takes into account phenomena such as migration,
geographical distribution, speciation, extinction, etc., which are not uniquely
explained by current environmental conditions but by ecological and geographical
conditions that existed in the past and evolved with time. Very often we have
difficulties to explain the reason of the geographical limit of a species range if a
clear ecological or geographical barrier is lacking. In that case, we can guess that
such limit has to do with competitiveness with other species or simply because it
needs more time to expand beyond its current border until it reaches its natural
ecological limit.
In any case, it is clear that biogeography is a synthesis discipline which
encompasses a set of phenomena governing the distribution patterns of organisms,
Mapping of Biogeographical Territories ... 23

from their current or past dispersal abilities to their migrations and ecological
adaptations, in the frame of a changing world in a historical perspective: climatic
changes, geographical changes (continental drift, orogenic processes, island
upliftments and sinkages and many others). All of them have their influence in the
present day distribution of living beings. Thus, biogeographical research necessarily
combines information and concepts from many fields: the physiological and eco-
logical opportunities and constraints influencing the dispersal of organisms, the
dispersal strategies of the different species, the geological, geographical and cli-
matic changes operating at global spatial scales combined with the evolutionary
phenomena which take place in the geological-scale time frame.

2 Definition of Biogeographical Territories

One of the products of traditional biogographical surveys has been the establish-
ment of territories on earth or seas characterized by their biodiversity content. Thus,
the definition of such territories is an activity which has been called provincialism
[10]. It has produced a hierarchical system of units ranked, in a descending order, as
realm (kingdom), region, province, sector and district. Botanists and zoologists
have contributed to this field since the early nineteenth century [32]. We can
mention A.P. De Candolle [13, 14] and also Schow [42], who produced the first
map of biogeographical regions covering the entire world. Since then several
important syntheses covering the globe or large areas of it have been produced,
including those by Drude [15], Meusel et al. [29], Good [23], Takhtajan [45] and
Schroeder [43].
This definition of territories provides a description of the different areas by
means of their current biological content, which is related to their current climatic,
geological and geographical conditions, and also to the evolutionary history of their
biota, forged by the climatic and geological history. From that point of view,
provincialism can be considered a relevant task in biogeography.
Biogeography “sensu stricto” considers organisms of all groups (plants, animals,
fungi, algae, bacteria, etc.), however, there is a tendency of describing biogeo-
graphical territories by using preferably plants, as they are particularly indicative of
the substrate (soil, bedrock, geomorphology) and atmospheric (climatic) conditions.
Thus, the term biogeography is often used as a synonym of phytogeography. In the
following paragraphs I shall focus on the vascular plants and the vegetation.

3 Floristic Approach

Initially, all the surveys and maps of the biogeographical units were done by using
taxa, either high rank such as families or genera or low rank such species and
subspecies. In this approach the floristic content of the different territories is
24 J. Loidi

considered. On the other hand, the distribution area of the taxa follows certain
regular patterns. As Myers and Giller put it “The existence of non-random patterns
in species distribution implies the operation of some general causal processes and
from an understanding of them one can work towards a reconstruction of the
history of life that can explain the present day distribution of species” [31]. In that
sense, concepts such as floristic element were coined [22]: this can refer to a group
of plants having similar distribution areas (geographical element), or having orig-
inated in the same region (genetic element), or that immigrated at the same time
(historical element) or following the same route (migratory element). Initially this
concept was introduced by Christ [12] in a geographical sense, but later
Braun-Blanquet (1923) used the term phytogeographical element with the aim to
include the flora and plant communities characteristic of a territory, i.e. “the floristic
and phytosociologic expression of a vast and defined territory”, so that we could
speak about the Mediterranean element or the Boreal element as the group of plants
and communities living exclusively or preferentially in the mentioned areas
(Mediterranean or Boreal) and constituting their living features. The different
floristic elements which participate in the flora of a territory are the footprints of the
successive migrations arrived in the past, and the endemics are the result of the local
biological evolution and witnesses of geographical isolation [41].
Another concept used in biogeography is that of geoflora, particularly applied to
the Arcto-Tertiary Geoflora and the Paleotropical Geoflora [28, 47]. Those terms
were established by Engler [17] and have been used mostly in paleobotany or
paleobiogeography [4, 28, 47] referring to the flora constituent of large vegetation
zones which persisted unchanged over long periods of time. Those species have thus
a common evolutionary history along a relevant period of their existence on earth
and reflect certain environmental conditions that existed in the past or at present.
Traditionally, distribution areas of taxa have been studied (chorology), and their
distribution patterns have been a fundamental source for the construction of the
biogeographical theory and spatial units. Some outstanding examples of this
activity are shown in the “Atlas of Flora Europaea, vols. 1–16” [26], the “Atlas of
British Flora” [34], the “Atlas partiell de la flore de France” [16] or the “Atlas der
Farn- und Blütenpflanzen der BRD” [25].
Such information is used to apply the knowledge of the species pool or floristic
content of the different territories, which is considered the main criterion for their
biogeographical definition. The species pool concept has already been used in
relation to diversification and migratory phenomena in biogeographical analysis
[48]. As a result, a number of biogeographical territorial maps (also called chori-
onomical synthesis) have been drawn, with a considerable degree of accuracy,
especially on broad geographical scales. All of them exclusively use the distribution
of taxa to define the biogeographical territories [23, 45].
The use of plant taxa is profoundly indicative of both the plant diversity of the
concerned territory and also of relevant evolutionary and paleoecological issues
affecting it. However, in the definition of lower rank biogeographical categories,
and particularly for their geographical delimitation, the flora or the floristic concepts
have a poor diagnostic power [41].
Mapping of Biogeographical Territories ... 25

4 Vegetational Approach

Since the spatial distribution pattern of plants (or other organisms) is conditioned by
the multiple environmental factors converging in each territory of the world, it is
possible to figure out that the environmental scenarios have a geographical vari-
ability, consequently responsible for the variability, not only of the territorial floras
or faunas, but also of the ecological frameworks which determine and organize the
existing communities, and that question is central in biogeography.
With the floristic approach we can at most know the plant species (or other
higher rank taxa) list of any territory, but no information is provided about the
relative abundance, distribution pattern and the assemblage patterns into commu-
nities of each of these taxa. Thus, in a step forward, it could be relevant to know
how this species pool assemble into communities, i.e. how plant species combine to
form associations [19]. Such phenomena depend on local climatic, lithological and
geomorphological conditions, as well as on the human land use system, the bio-
geographical territories are the specific scenario in which those phenomena occur
(Fig. 1).

Fig. 1 Schematic representation of the biogeographical synthesis in a given area

26 J. Loidi

Fig. 2 Floristic and vegetational content of the biogeographical categories

The abundance and distribution pattern of plants is of basic importance in

biogeography. They are directly determined by the geology, climate and history of
each territory. Consider, for instance, two common plants in Europe: Urtica dioca
and U. urens, both linked to strongly nitrified habitats. The first is particularly
common in the temperate climatic regions of Europe, while the second is prevalent
in the Mediterranean part. Nevertheless, they are both profusely represented in both
regions: U. dioica can also be found in the hydromorphic soils in Mediterranean
Europe, while U. urens is frequently found in the relatively dry and continental
areas of the temperate part of the continent. Dot maps do not make this situation
altogether clear, as both species are represented by dense clouds of dots in both
parts of Europe, masking their real distribution pattern and relative abundance. At a
more detailed scale we can see that a rocky country will have abundant populations
and communities of chasmophytic plants, while a neighbouring one with less rock
outcrops will have lower abundance of those particular taxa and vegetation. Such
differences will be poorly expressed by comparing floristic catalogs or by observing
dot maps. Indeed, the latter can even give a distorted idea of the true abundance and
distribution of some plants; even some localities are famous for having rare plants
that are often located on the edge of their area and which are over-represented in
herbaria and literature. Quantifying the “role” of every plant in a particular territory
and considering the assemblage rules among species is a task for a more accurate
biogeography, as it reflects geological, climatic and historical phenomena. Taking
vegetation into account is one way of advancing in this direction.
Mapping of Biogeographical Territories ... 27

The use of vegetation units in the form of phytosociological plant communities

in biogeography started early in the twentieth century with Flauhault [20, 21],
Braun and Furrer [7] and Braun-Blanquet [8, 9] and has been widely accepted later
[39, 41]. They established the first rules for applying knowledge of syntaxa and
their distribution to the definition of territorial units. Figure 2 summarizes the
taxonomic and vegetational content in the different biogeographical categories.

5 Landscape Approach

For defining higher rank biogeographical categories, kingdom or region, the array
of taxonomic criteria are sufficient, as they are large areas bearing profound dif-
ferences in their flora resulting from very different evolutionary histories; we can
easily understand that floristic differences are enough to sustain the independence of
Holarctic and Paleotropic kingdoms. When descending in rank, i.e. province or
sector, the vegetation analysis in terms of plant communities (syntaxa) becomes
important for documenting the units and especially for a definition and a precise
drawing of the limits between the categories in a map. Floras can be too similar
between two neighbouring provinces to provide criteria for such precise delimita-
tion. However, in the low ranks, limits cannot be precisely defined by using either
the flora or the plant communities, we need more precise criteria which are provided
by a specific form of landscape analysis called Integrated phytosociology [1] or
dynamic-catenal phytosociology [37].
The incorporation of this landscape analysis in the provincialism has been car-
ried out in some territories, mostly in southern Europe, in the last decades of the
twentieth century [3, 5, 6, 33, 36, 44]. On a detailed scale, concerning mostly the
low-rank units, Alcaraz et al. [2] and Loidi and Fernández Prieto [27] made sub-
stantial contributions.
In order to briefly explain the application of a certain type of landscape analysis
to the biogeographical definition of territories, Alcaraz [1] attempts to formalize the
criteria (under the name of integrated phytosociology). He tries to define the basic
unit, i.e. the district, as the absolute homogeneous territory from the biogeo-
graphical point of view, resulting from its climatic and geomorphological homo-
geneity. As indicated above, differentiating closely-related neighbouring districts
using only floristic criteria or a bare list of plant communities is almost impossible.
Such lists of species and of plant communities share almost all of the elements and
would not show any substantial differences between the districts. However, in
applying landscape analysis, the rules of distribution of communities into the
landscape can change minimally, so that it can result in certain and clearly different
distribution patterns across the territory. To detect this, we can focus on the ordered
arrangement of vegetation units across a gradient, i.e. the vegetation zonation,
which, when repeated in a certain territory, can be typified and named with the
prefix geo-. For that, the name [50] geosigmetum was established by
Rivas-Martínez et al. [37, 38]. These zonations can follow a universal topographic
28 J. Loidi

Table 1 Summary of the defining criteria for the biogeographical categories

Flora Plant communities Integrated phytosociology
Kingdom Endemic families, Practically all the classes
(Realm) subfamilies and tribes; very are endemic, only
high generic and specific synatropic or aquatic ones
endemism. Paleogenic are cosmopolitan or
endemism of high rank subcosmopolitan
Region High generic and specific Several exclusive classes
endemism. Few families and orders. Classes with
endemic or with their optimum within the
predominant position in the region involved
region. Paleogenic
endemism of intermediate
rank (genus)
Province Some endemic genera Some alliances or Some exclusive
(usually monotypic or suballiances endemic. climatophilous sigmeta.
oligotypic). Still relevant Relevant association Particular altitudinal
specific endemism; endemism (mountain) vegetation
paleogenic endemism of zonation (cliserial
low rank (species). geosigmetum)
Statistically distinct
assemblages of species
Sector Low or absence of species Some associations endemic Several exclusive
endemism; long isolated climatophilous geosigmeta
islands increase endemic
element
District Usually no endemism in No syntaxonomical Geomorphologically
the flora. Slight floristical endemism; in some cases homogeneous territory. At
differences with neighbour lower rank syntaxa like least one climatophilous
districts are due to subassociation, variant and geosigmetum and some
geomorphology: relief or facies are exclusive. Some geopermasigmeta exclusive
substrata, or to slight associations are lacking in to the territory
differences in climate neighbouring districts

pattern crest-slope-valley bottom model. This gradient model is universal and can
be studied in any territory of the world, being called the climatophilous geosig-
metum. In a broader scale, the zonation of vegetation belts along the altitudinal
gradient of a mountain has been called cliserial geosigmetum. Cliserial geosigmeta
are strongly linked to regional environmental conditions (regional climate and
substrate) and thus have a high biogeographical specificity (Fig. 6). Other cases are
those gradients occurring in exceptional situations, as in highly stressed areas under
extremely limiting conditions such as coastal dunes, salt marshes, cliffs, screes,
ponds, mires, fens, etc. In those cases the geosigmetum becomes formed by per-
manent vegetation types and the term geopermasigmetum is used [38]. It has been
established that the minimum difference needed to separate two districts is that they
differ in at least one permageosigmetum present in one of them and absent in the
other,the rest, i.e. all climatophilous geosigmeta, as well as the rest of the perma-
geosigmeta, may be shared. Logically, the greater the differences in geosigmeta
Mapping of Biogeographical Territories ... 29

Fig. 3. Floristic kingdoms and subkingdoms of the world, after Takhtajan [45]. A–C Holarctic
kingdom (A Boreal subkingdom, B Tethyan subkingdom, C Madrean subkingdom); D–H
Paleotropical kingdom (D African subkingdom; E Madagascan subkingdom; F Indo-Malaysian
subkingdom; G Polynesian subkingdom; H Neocaledonian subkingdom), J Neotropical kingdom;
K Cape kingdom; L Australian kingdom; M Anctarctic kingdom

(climatophilous or permageosigmeta), the greater the biological distance between

the territories analyzed. This system helps to provide a very accurate definition of
the units and to draw the frontiers between those districts; each piece of land can be
submitted to a rigorous and quick analysis leading to its inclusion in one district or
the other. In practical work, such frontiers have been drawn at a relatively detailed
scale, sometimes at 1:50.000. This is the starting point for efficient work in bio-
geographical provincialism at the elementary level.
Table 1 shows the floristic, pure phytosociological and integrated phytosocio-
logical criteria that can be used to characterise each biogeographical rank. They are
taken from Braun-Blanquet [9], Takhtajan [45], Rivas-Martínez [36–38] and
Alcaraz [1]. In it, the decreasing importance of floristic criteria when descending in
rank parallel to the increasing importance of landscape criteria becomes evident.
Vegetation units play an intermediate role.

6 Definition of the Biogeographical Territorial Ranks

6.1 Kingdom (Realm)

It is the highest rank unit. It is usually a territory of great extent, encompassing one
or several continents and concealing entirely several high rank taxonomic entities,
such as families, subfamilies or tribes, as well as most of the genera and practically
all the species occurring in it. Often one particular vegetation formation is exclusive
30 J. Loidi

of one kingdom. In the world there are six kingdoms traditionally recognized:
Holarctic, Paleotropical, Neotropical, Capensis, Australian and Antarctic (Fig. 3).
Some authors include the Capensis in the Paleotropical [35] and even split the
Antarctic and allocate its fragments to the neighbouring kingdoms [11].

6.2 Region

The kingdoms are divided into regions, which are vast territories characterized by a
high specificity in their flora, having often paleogenic endemics such as a small
family or some families which dispersal centre lies in the region, so that although
not strictly endemic, they can be considered having their optimum in the concerned
region (e.g. Cistaceae for the Mediteranean region). The level of endemicity
increases with descending taxonomic rank, so that several genera and most of the
species are endemic to the region. The regions can be characterized by some
vegetation classes exclusive to them and by some others for which the optimum is
in the concerned region, eg. Quercetea ilicis has its optimum in the Mediterranean
region while Molinio-Arrhenatheretea develops mostly in the Eurosiberian region.
At lower levels, such as order and alliance, endemicity increases. There are 35
regions around the world [45] (Fig. 4).

Fig. 4 Biogeographical regions of the world [23]

Mapping of Biogeographical Territories ... 31

Fig. 5 Biogeographical map of Europe with the regions and provinces [39]

6.3 Province

It is a territory of regular extent, highly variable depending on spatial niche diversi-

fication. Endemicity hardly reaches the genus level with some mono- or oligospecific
genera. The number of endemic species is high. Syntaxonomically, there are endemic
alliances but at the association level there are more, even in vegetation types repre-
senting mature stages of succession. Vegetation zonation patterns (geosigmeta) are
entirely particular to the province (Fig. 5). Vegetation belts zonation in mountains
(cliserial geosigmeta) are characteristic of the province (Fig. 6).

6.4 Sector

It is a territory of modest extent (Fig. 7) where endemicity concerns mostly sub-

species, reaching to the species level in few cases (this is not applicable to mountain
areas and to true islands, where endemicity increases). The floristic differences refer
to the species pool. It can have some endemic associations but not of mature stages
of succession. One important feature to distinguish sectors belonging to the same
province is their specific zonation pattern (geosigmetum) in the crest-slope-valley
bottom model. Slight variations in cliserial geosigmeta can also be used to separate
closely related sectors [2].
32 J. Loidi

Fig. 6 The differences in the altitudinal zonation pattern in different mountains are the result of
profound differences in climate, substrate and/or vegetation history and can thus be used as a
criterion for differentiating biogeographics units, such as provinces or even sectors, depending the
importance of such differences. Between A and D there is a difference in substrate while between
A and B and C there are differences in climate. In both cases, the differences entail a change in the
zonation pattern (cliserial geosigmetum) which, if accompanied by differences in historic elements,
can be used to substantiate with more or less arguments, the independence between biogeographic
territories

6.5 District

It is the lowest unit in the classification system, and also the most homogeneous in
terms of mesoclimate, relief, lithology and history of the flora living within it. It is a
territory of small extent (Fig. 8) without endemicity at any taxonomic level (per-
haps some accidentally). As regards vegetation, there is no endemicity even in the
association level. For Braun-Blanquet [9] a district is an area defined only by
“particular communities of lower rank (variants, facies), lacking endemic species”.
The differences with a contiguous district in terms of flora and vegetation are due to
slight changes in mesologic conditions, not due to facts related to a “phylogenetic
Mapping of Biogeographical Territories ... 33

Fig. 7 Biogeography of the Iberian Peninsula at the level of sector (Rivas-Martínez et al. [40])

phytogeography” in the sense of Engler [18]. Regarding landscape, the pattern of

distribution and sorting of the plant communities concerning their zonal and suc-
cessional relationships remains within its limits, i.e. the formal zonation pattern, or
geosigmetum, remains unaltered within the district [1]. Any change in these pat-
terns (affecting at least one geopermasigmetum) due to a slight change in climate,
lithology ot relief, means a change to another district within the same sector.
34 J. Loidi

Fig. 8 Biogeography of the North-west of the Iberian Peninsula at the level of district
(Rivas-Martínez et al. [40])

References

1. Alcaraz F (1996) Fitosociología integrada, paisaje y biogeografía. In: Loidi J (ed) Advances in
phytodociology, pp 59–94. Ed. UPV/EHU
2. Alcaraz F, Sánchez-Gómez P, De La Torre A (1991) Biogeografía de la provincia
Murciano-Almeriense hasta el nivel de sub sector. Rivasgodaya 6:77–100
Mapping of Biogeographical Territories ... 35

3. Asensi A 1996 Fitosociología y paisaje (Una aproximación histórica). In: Loidi, J. (ed.)
Avances en Fitosociología, pp 43–58. Ed. UPV/EHU
4. Axelrod DI (1966) Origin of deciduous and evergreen habits in temperate forests. Evolution
20(1):1–15
5. Biondi E (1996) L’analisi fitosociologica nello studio integrato del paesaggio. In: Loidi J
(ed) Avances en fitosociología, pp 13–22. Ed. UPV/EHU
6. Blasi C (ed) (2010) La Vegetazione d’Italia con Carta delle Serie di Vegetazione in scala
1:500000. Palombi Editori, Roma
7. Braun J, Furrer E (1913) Remarques sur l’étude des groupements de plantes. Soc Languedoc
Geogr Bull 36:20–41
8. Braun-Blanquet J (1919) Essai sur la notion d’element et de territoire phytogeographyques.
Arch Sci Phys Nat Geneva Ser 5, 1:497–512
9. Braun-Blanquet J (1951) Pflanzensoziologie. Grundzüge der Vegetationskunde. Springer,
Wien
10. Brown JH, Lomolino MV (1998) Biogeography, 2nd edn. Sinauer, Sunderland
11. Cox B (2001) The biogeographic regions reconsidered. J Biogeogr 28:511–523
12. Christ H (1867) Ueber die Verbreitung der Pflanzen der alpinen Region der europäischen
Alpenkette, Basel
13. De Candolle AP (1820) Essai élémentaire de Géographie botanique, vol 18. du Dictionnaire
des Sciences Naturelles. F.G. Levrault, Strasbourg
14. De Candolle AP (1855) Géographie Botanique Raisonné, vol 2. V. Masson, Paris
15. Drude O (1884) Die Florenreiche der Erde (The floristic realms of the Earth). Petermanns
Geogr. Mitt., Erg änzungsheft, p 74
16. Dupont P (1990) Atlas partiel de la flore de France. Secrét Faune et Flore Mus Nat D’Histoire
Naturelle, Paris
17. Engler A (1879–1882) Versuch einer Entwicklungsgeschichte der Pflanzenwelt seit der
Tertiärperiode. Leipzig Engelmann, pp 1–2
18. Engler A (1899) Entwickelung der Pflanzengeographie in den letzten hundert Jahren und
weitere Aufgaben derselben. Wissenschaftliche Beiträge zum Gedächtnis der hundertjährigen
Wiederkehr des Antritts von Alexander von Humboldt’s Reise nach Amerika. Gesellschaft für
Erdkunde zu Berlin, Berlin
19. Eriksson O (1993) The species pool hypothesis and the plant community diversity. Oikos 68
(2):371–374
20. Flauhault C (1901a) La flore et la Végétation de la France. In: Coste, H. Flore descriptive et
illustrée de la France, de la Corse et des contées limitrophes, vol 1. Libr. Sci. Et Techn. Albert
Blanchard, Paris, pp 1–52
21. Flauhault C (1901) Premier essai de nomenclature phytogeographique. Bull Spc Langedoc
Géogr 24:157–192
22 Font Quer P (1953) Diccionario de Botánica Ed. Labor, Barcelona
23 Good R (1974) The geography of the flowering plants, 4th edn. Longman Group Ltd.,
London, p 557
24 Grisebach A (1884) Vegetation der Erde, vol 1. Verlag von W. Engelmann, Leipzig, p 230
25 Haeupler H, Schönfelder P (1989) Atlas der Farn- und Blütenpflanzen der Bundesrepublik
Deutschland Ed. Eugen Ulmer, Stuttgart
26 Jalas J, Suominen J (1972–2013) Atlas Florae Europaea, vol 1–16. Committee for mapping
the Flora of Europe, Helsinki
27 Loidi J, Fernández Prieto JA (1986) Datos sobre la biogeografía y la vegetación del sector
castellano-cantábrico (España). Docum. Phytosociologiques N.S. 10:323–362
28 Mai DH (1991) Palaeofloristic changes in Europe and the confirmation of the
Arctotertiary-Palaeotropical geofloral concept. Rev Palaeobot Palynol 68:29–36
29 Meusel H, Jäger E, Rauschert S, Weinert E (1965–1992) Vergleichende Chorologie der
zentraleuropäischen Flora. Ed Gustav Fischer, Jena
30 Müller P (1986) Biogeography. Harper & Pow, New York, p 377
31 Myers AA, Giller PS (eds) (1988) Analytical biogeography. Chapman & Hall, London, p 578
36 J. Loidi

32. Nelson G (1978) From Candolle to Croizat: comments on the history of biogeography. J Hist
Biol 11:269–305
33 Pedrotti F, Venanzoni R, Moretti G, Cianficconi F (1987) Biogeografia e sue prospettive
attuali. Aspetti e problem della geografia, pp 557–588. Ed. Marzorati, Milano
34 Perring FH, Walters SM (1962) Atlas of British Flora. Ed. Nelson, London & Edinburgh
35 Pfadenhauer JS, Klötzli FA (2014) Vegetation der Erde. Springer, p 643
36 Rivas-Martínez S (1987) Memoria del mapa de series de vegetación de España, 1:400.000.
ICONA, Madrid
37. Rivas-Martínez S, et al (2007) Mapa de series, geoseries y geopermaseries de vegetación de
España Memoria del mapa de vegetación potencial de España] Parte I. Itinera Geobotanica
17:5–436
38 Rivas-Martínez S, et al (2011) Mapa de series, geoseries y geopermaseries de vegetación de
España [Memoria del mapa de vegetación potencial de España]. Parte II. Itinera Geobotanica
18 (1):5–424
39 Rivas-Martínez S, Penas A, Díaz González TE (2001) Biogeographic map of Europe, León
40 Rivas-Martínez S, Penas A, Díaz González TE, Cantó P, Del Río S, Costa JC, Herrero L,
Molero J (2017) Biogeographic units of the Iberian Peninsula and Baelaric Islands to district
level. A concise sinopsis. In: Loidi J The Vegetation of the Iberian Peninsula
41 Roisin P (1969) Le domaine phytogéographique atlantique d’Europe. Les presses
agronomiques de Gembloux A. S. B. L. Ed. Duculot. Gembloux, p 262
42 Schow JF (1823) Grundzüge einer allgemeinen Pflanzengeographie, Berlin
43 Schroeder F-G (1998) Lehrbuch de Pflanzengeographie. Verlag Quelle & Meyer, Wiesbaden,
p 457
44 Schwabe A (1997) Sigmachorology as a subject of phytosociological research.
Phytocoenologia 27(4):463–507
45 Takhtajan A (1986) Floristic regions of the world. University of California Press, Berkeley
46 Vincent P (1990) The biogeography of the British Isles. Routledge, New York, p 315
47 Wolfe JA (1981) Paleoclimatic significance of the Oligocene and Neogene floras of Eastern
Asia and relation to forests of other regions of the Northern Hemisphere and Australasia.
United States Geol Surv Prof Pap 1106:1–37
48 Zobel M, Otto R, Laanisto L, Naranjo-Cigala A, Pärtel M, Fernández-Palacios JM (2011) The
formation of species pools:historical habitat abundance affects current local diversity. Glob
Ecol Biogeogr 20:251–259
49. Braun-Blanquet J (1923) L'origine et le développement des flores dans le massif central de
France avec aperçu sur les migrations des Flores dans l'Europe sud-occidentale. Annales de la
Société linnéenne de Lyon et des Société botanique de Lyon, Société d'anthropologie et de
biologie de Lyon réunies 69(1):1–16
50. Loidi J (2017) Dynamism in Vegetation. Vegetation changes in a short time scale. In: Loidi J
(Ed.) The vegetation of the Iberian Peninsula vol. 1:81–99.
Dynamic-Catenal Vegetation Mapping
as a Tool for Ecological Restoration
and Conservation Policy

Javier Loidi

Abstract Traditional or standard vegetation mapping represents a static view of

the ecosystems living on earth. The units represented in the map are vegetation
types that express one or several features of the existing reality, i.e. species com-
position, size, structure, etc.; they express little, however, about the dynamic
context in which they develop and even less about the spatial relationships network
of which they form part. This chapter presents an approach that develops a system
of complex geobotanical entities expressing the temporal and spatial relationships
between simple vegetation units found in the field. The basic relationships are the
successional and ecological gradients. Consequently two basic concepts are
established: the sigmetum (or vegetation series), in which the vegetation units are
related by successional series; and the geosigmetum, which encompasses the
zonation associated with an ecological gradient. Within these categories variants
such as minorisigmeta and geominorisigmeta as well as permasigmeta and geop-
ermasigmeta are considered. Maps drawn using these units provide valuable syn-
thetic information for ecological restoration and land management.

Keywords Geosigmetum Potential natural vegetation
Sigmetum
Tessela



Vegetation catena Vegetation series

1 Introduction

Vegetation Science—or Geobotany—has developed working on the initial intuitive

idea that there are patterns or laws which determine the existence and distribution of
the different plant communities (vegetation units) observed in the landscape [36].
After two centuries, we have constructed different explanatory scientific models that
formalize such organization. A milestone in this history was the formalization and

J. Loidi (&)
Department of Plant Biology and Ecology, University of the Basque Country (UPV/EHU),
P.O. Box 644, 48080 Bilbao, Spain
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 37

F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_4
38 J. Loidi

systematic description of plant communities after their floristic composition (phy-

todiversity content), led by the Zürich-Montpellier school, developed mostly in
continental Europe and Japan. This approach has been extensively called phy-
tosociology [13]. A further level of integration in this survey for patterns in veg-
etation distribution was to address the possibility of recognizing complexes or
assemblages of plant communities which can reflect some existing patterns in the
distribution, not only of plants, which produce a regularly based plant community
typology, but also in the spatial and temporal distribution of the plant communi-
ties, this would enable them to be grouped into categories of higher level. This idea
of a higher level of integration of vegetation units has been underpinning the
thinking of vegetation scientists more or less intuitively since early times of phy-
tosociology and geography [12, 67]. It was fueled by the assumption of the suc-
cessionist postulates of Clements [17, 18] in the early 20th Century, which
permitted the universal acceptance of the dynamic nature of vegetation. However,
until the seventies, no serious effort was made to produce a scientific construction
that would order the previously observed patterns of relations between different
plant communities. Nonetheless, in the long period between 1910 and 1973 there
were some fundamental theoretical contributions that made further developments
possible. The first was by Braun-Blanquet, who accepted the Clementsian suc-
cession theory, in his original “Vocabulary of Phytosociology” [15] there are
several concepts which group a number of plant associations in time (série) or space
(climax-complexe, complexe des groupements). Tüxen [71] made a crucial con-
tribution by establishing the concept of Potential Natural Vegetation (PNV), largely
used by European and Japanese phytosociologists [35]. Later, Bolòs [10, 11]
defined some important concepts, such as tessella, and discussed the catenal
arrangement of plant communities, accepting the ideas of vegetation series, PNV,
primitive vegetation and climax. But the first attempt to construct a formal theo-
retical model begins with Tüxen [72], and has continued with many later contri-
butions: Géhu [25–31], Beguin et al. [3], Béguin and Hegg [4], Tüxen [73, 75] and
Rivas-Martínez [54]. The work of Géhu and Rivas-Martínez [33] on basic con-
ceptions in phytosociology brought this phase to an end. This period was followed
by another in which the above-mentioned concepts were developed and tested in
the field, often accompanied by different kinds of maps [43, 58]. Some compre-
hensive and conceptual reviewing papers about this issue of vegetation complexes
are worth being mentioned, such as those by Matuskiewicz [46], Theurillat [69],
Rivas-Martínez [60], Gillet and Gallandat [34], Schwabe [68], Vigo [76] and
Biondi [6].

2 What is Dynamic Catenal Phytosociology?

The term “integrated phytosociology” [32], initially established and later replaced
by dynamic-catenal phytosociology [64], was coined by Rivas-Martínez [58] in an
attempt to integrate the fields of phytosociology, synphytosociology and
Dynamic-Catenal Vegetation Mapping ... 39

geosynphytosociology. Integrated phytosociology could be seen as a sort of phy-

tosociology squared or even cubed [33] or as a kind of high phytosociology (hohere
Pflanzensoziologie), as Tüxen put it [74]. Recently, the terms integrated phytoso-
ciology and landscape phytosociology tend to be replaced by the more explicit
dynamic-catenal phytosociology [65]. This discipline studies the associations of
associations, or association complexes, distinguishable in the landscape, their
organization patterns and the factors determining such organization.
The main issue to be settled when starting a scientific grouping of plant com-
munities is the selection of the criteria to be applied, i.e. to determine the assembly
rules governing such groupings. The definition of such criteria and their relative
acceptance took decades, from the beginning of phytosociology to the early sev-
enties. When phytosociology was still in its infancy, only a handful of authors
actually wrote down ideas on the question of how communities could be combined
within the landscape, and occasionally left the limits of the proposed combinations
unclear. A good example is the basic work of Braun-Blanquet and Pavillard [15], in
which they defined their “complex of communities” (Gesellschaftkomplex) as a
combination of plant communities presenting a repetitive pattern in the landscape
due to ecological and geomorphological coincidences. This is still a rather loose
concept with no clear-cut limits. Later contributions helped to make things some-
what clearer, particularly Tüxen’s contribution [72], who decided that the sample,
i.e. the sigma-relevé or relevé of the plant communities of a particular area, should
be made within the limits of the area of a single association of Potential Natural
Vegetation.
In a synthetic way we can establish that there are only two different criteria to be
used to distinguish groups of vascular plant communities on the territory and make
sensitive groupings out of them: the successional and the zonal. This is because
succession and zonation are the two main phenomena involving changes in plant
communities in time or space respectively and thus encompassing all the variability
in plant communities in a given area; succession is related to disturbance, which
includes human influence. In other words, any plant community will have a
determined position in succession and zonation, we only need to know how to place
it in the correct position (Fig. 1). Thus, if we are able to use both criteria for
organizing the existing plant communities in any area, we will be able to somehow
express such phenomena in the frame of landscape ecology involving the existing
phytodiversity. Vigo [76] addressed this question using the term serie for the
temporal dimension and arrangement for the spatial one. Thus the task of
dynamic-catenal phytosociology is to apply both criteria defining two integration
levels, one for each criterion.
40 J. Loidi

Fig. 1 Plant community diversity in any area is a result of two drivers: 1. Mesological gradient
and 2. Successional position, being the latter narrowly related to disturbance regime. Plant
communities A and B are determined by their particular position both in the gradient and in the
succession.

3 Main Levels of Integration

The successionist criterion groups communities that succeed each other in the same
area (serie); this type of process would end in a stable maximum in equilibrium with
the current environmental conditions. This implies, of course, the acceptance of the
concept of Potential Natural Vegetation (PNV) as formulated by Tüxen [71], as
well as the idea of vegetation series [12] and tessella [10, 11].
The other criterion is based on the zonation of plant communities caused by one
or several mesologic gradients: The zonal distribution of different vegetation types
is called a catena (= chain) and used to be represented by a schematic section of the
terrain. The scale of the zonation varies according to the gradient represented;
changes in vegetation due to climatic gradient use very reduced scales involving for
example a whole mountain or an entire region, while zonation due to soil moisture
gradient in a pond will use a much more detailed one. So it is clear that a catena is a
different type of representation to a successionist model. In the latter, the plant
communities represented share the same ecological space, while in the former
ecological parameters change causing change in the vegetation types. The term
catena was initially used in soil science [49, 66] to express the sequences of soil
types in a valley or small region [48], and later was formalized by Bolòs [10] for
vegetation science.
Dynamic-Catenal Vegetation Mapping ... 41

Table 1 Integration levels in phytosociology.

Level Studied units Driving factor Scientific discipline
1 plant communities Floristic composition classic phytosociology
2 sigmeta succession synphytosociology
(or vegetation series)
3 geosigmeta catena geosynphytosociology
(or geoseries)

Thus, integrated or dynamic-catenal phytosociology as landscape science [2, 5]

is involved in assemblages of communities linked by dynamic relationships on the
one hand and by zonal ones on the other. In the words of Rivas-Martínez [59], it
develops on three levels: associations (Braun-Blanquet or classic phytosociology),
vegetation series, also called sigmeta (synphytosociology), and geoseries, also
called geosigmeta (geosynphytosociology). The first refers to assemblages of
plants, the second to succession and the third to zonation (Table 1).

4 Main Concepts
Potential Natural Vegetation (PNV). This idea has been widely debated and is
one of the key concepts of dynamic-catenal phytosociology. The old idea of
Clements [18] that there was an optimal phase in vegetation in balance with
regional climate beyond which succession could never achieve a higher degree of
complexity and size, called the climax, is a concept that resonates with the logic of
ecology. Any system has its carrying capacity, which cannot be exceeded, and the
vegetation also reaches that carrying capacity or limit in the possibilities of
development according the climate, the available resources and the available flora
of the region (plants having the possibility to reach a site by their own dispersal
means). This initial idea was strongly connected with the regional climate and
supposed that different substrata would converge towards the same climax if there
was a similar climate. This accepted the presumption that such a climax would be
reached after a sufficient length of time, after which no disturbance would happen.
The idea was modified by Tüxen [72], who formulated the concept of Potential
Natural Vegetation (Potentielle Natürliche Vegetation). He wrote. “… is the veg-
etation that would finally develop if all human influences on the site and its
immediate surroundings would stop at once, and if the terminal stage would be
reached at once”. This text apparently suggests that we should have a lot of time at
our disposal to test the validity of each type of PNV we hypothesize. But over such
a long time, environmental conditions could change, such as the climate for
instance, deviating the succession tendency in its course and reaching to an
unexpected end. However, it is not the case for the concept, since Tüxen’s paper
clearly explains that the PNV assumption is according to the current (heutige)
42 J. Loidi

Fig. 2 Hypothetical character in PNV construction [77]. The lower the position of the community
in the successional series is, i.e. higher hemerobie [9], the greater the hypothetical character in the
PNV is in that site.

conditions. The PNV has to be imagined as if succession would happen instanta-

neously (schlagartig). Then it is not a set of hypothesis for succession processes, it
is a set of assumptions of ecological potentialities in a given territory. However, in a
landscape covered by a variety of plant communities forming a mosaic, the
assumption of its PNV has a variable degree of hypothetical character [77],
depending on the distance a concrete plant community found in a site is from the
PNV, the more hemerobic a community is [9] the longer the distance to the PNV is
(Fig. 2).
Intense discussion of the PVN concept has continued until recently [16, 42,
45, 50] and, although some authors consider it unnecessary [19], a large number of
vegetation scientists use it as a descriptor of territory in land planning for conser-
vation and as a target guide in restoration projects.
Tessella. This term is taken from the Latin tesella (paving stone)—Fliese in
German [80], piastrella in Italian [53]—and has been used profusely with different
meanings. In general, it has been used to refer to each of the surface elements that
compose a landscape, like each of the pieces of a mosaic. Each tessella coincides
with each of the visible plant communities; their geometrical shape produces a
perfect assemblage, like the pieces of a puzzle. In this sense it is purely descriptive
and does not define any ecological quality of the element considered.
Dynamic-Catenal Vegetation Mapping ... 43

Fig. 3 Scheme of plant communities in a Sigmetum (Querco rotundifolae-Sigmetum) lying on its

tessella. The forest (PNV) and its substitution communities: mantle, scrub and other herb
communities whose appearance is dependent on human influence (from [43]). 1- Quercetum
rotundifoliae (PNV); 2- Rhamno-Quercetum cocciferae (mantle); 3- Salvio-Ononidetum fruticosae
(scrub); 4- Ruto-Brachypodietum retusi (grassland); 5- Trachynion (annual grassland); 6-
Lygeo-Stipetum lagascae (steppic grassland); 7- Pinus halepensis plantations; 8- Olive tree
plantations; 9- Vineyards; 10- Cereal crop fields.

Another meaning of tessella is the sense in which it was used by Bolòs [10, 11],
who gave an inspired ecological and successionist definition of the term: “any
ecologically homogeneous portion of land and thus able to sustain a determined
stable community as well as its substitution associations … The tessella is the
elementary component of the mosaic of pieces of ground of different quality which
constitute the territorial basis of the landscape. Homogeneous in its inherent
properties, each tessella used to be occupied, in a primeval landscape, by a par-
ticular community. However, in human-influenced landscapes, the homogeneity of
the tessellar areas is often masked because of the different treatment applied to its
different parts.” The tessella is thus a piece of land, a physical territory homoge-
neous in its intrinsic mesologic conditions: climate, soil, topography, hydric and
chemical conditions of substratum, etc. Such homogeneity results in an ecological
equipotentiality which supports a unique association of PNV: this occurs only if
enough disturbance-free time passes. PNV is usually scarce in the real landscape of
most areas, strongly influenced by human activity; most of the surface is covered by
44 J. Loidi

seral communities (Ersatzgesellschaften) of different kinds, depending on the

various treatments given by man. From this circumstance, it is possible to relate all
these communities and group them together to build a higher entity (Fig. 3).
This interpretation of tessella, largely accepted by Rivas-Martínez and other
authors, implies the acceptance of two main ideas:
1. The assumption of the concept of PNV, which may coincide -or not- with
primitive vegetation.
2. The almost reciprocal one-to-one relationship between tessella and PNV asso-
ciation. Each tessella (or approximate pluri-tessellar complex) sustains a unique
association of PNV due to the equipotentiality of the entire tessellar area.
Inversely, each PNV association occupies its own tessella, each PNV associa-
tion has to have its defined place.
Sigmetum (also called Vegetation Series). Sigmetum is the oldest concept of
integrated phytosociology, and was already used, under the name of series, by
Braun and Furrer [12] and Braun-Blanquet and Pavillard [13]. They distinguished
the initial and transformation stages which terminate in a final stage to bring the
dynamic process called succession to a close [18]. However, the idea of series was
only extensively used much later, especially in France [24, 51, 52], Spain [10, 41,
43, 44, 55–58] and Italy [6–8]. Sigmetum is a geobotanical concept which includes
one PNV association as well as all the communities which substitute it as a result of
regressive succession (Fig. 3). In Bolòs’ words sigmetum “is the ordinated group
of communities which, in the same place, substitute each other in time” (Fig. 4).
Hence, the sigmetum expresses all the vegetal content organized in communities,
linked by dynamic relationships, which are living on and being supported by a
tessella (Fig. 5). Certainly sigmetum and tessella are as closely related concepts, as
PNV and tessella, and are often used as synonyms. However, since sigmetum
comprises all the communities living on the tessella, its use is more appropriate
when speaking about the tessella’s vegetation. Other terms used in the literature,
such as synassociation [54], sigmassociation [21, 72, 73–75] or, more repeatedly,
vegetation series, express the same concept [55–58, 60, 62, 64, 65].
A sigmetum in a particular area is usually built following chrono-sequences, i.e.
space-for-time substitutions. Each sigmetum is built up by detecting the most
mature vegetation type within each tesella, attributing to it the PNV role and adding
the rest of the plant communities (seral stages) observed in the vicinity within the
tessella. The seral stages are usually numerous and diverse, being necessary to take
into account the hemeroby [9] degree associated with each of them (Fig. 6). The
whole set constitutes the sigmetum.
Dynamic-Catenal Vegetation Mapping ... 45

Fig. 4 Schematic representation of the concept of sigmetum and its dynamic significance. In a
particular place, a plant community (association) is conditioned by a disturbance regime (1), when
disturbance ceases, secondary succession drives the vegetation towards PNV along several stages
represented by seral communities 1 to 3. If the disturbance regime is a different one (2), the
resulting association is different (transformed association) and if that second disturbance regime
ceases, the secondary succession will drive the vegetation to the same endpoint but along a
different set of stages, seral communities 1′ to 3′. All the communities of the seral stages in
addition to the PNV constitute the Sigmetum or Vegetation Series.

It is important to formalize and systematize sampling procedures, establishing a

sampling methodology that renders all collected samples comparable to each other
as a condition for building a typology on which new models can be based. A first
step in this direction was made by Bolòs [10], who proposed the following scale to
assess the abundance of the associations existing within a sigmetum when making a
sigma-relevé:
1. Dominant community (dom): covering over 50% of the area
2. Extensive community (ext): >10% <50%
3. Localized community (loc): covers small areas but occurs regularly in favorable
sites
4. Sporadic community (spor): only exceptionally observed
Béguin et al. [3] and later Géhu and Rivas-Martínez [33] proposed, with much
more success, the use of the Braun-Blanquet scale to assess the abundance of each
46 J. Loidi

Fig. 5 Schematic representation of possible plant communities existing in a sigmetum, their

successional relationships and associated disturbance regimes.

community of the sigma-relevé. Certain symbols can be also added to indicate the
type of distribution of the community on the land:
O: Spatial distribution (e.g. Cynosurion meadows in a bocage landscape)
/: Linear distribution (e.g. Prunetalia hedges in the same landscape)
0: Spatial-linear distribution, as in a broad fringe
•: Point distribution.
This system combines an assessment of the abundance of each association
occurring in one tessella with an estimation of their quality or nature expressed
through their distribution type on the land. It is possible, then, to collect samples
(sigma-relevés) which can be tabulated and treated numerically by using multi-
variate methods or other statistical analysis.

• Ecologic Variability. Despite the homogeneity of the mesological parameters

over their area (tessella) as a whole, ecological and geographical factors usually
give the sigmetum a certain internal variability [1, 41]. Ecological variations are
generally related to minor lithological and trophic changes incapable of pro-
moting substantial changes in the PNV. Such variations may influence changes
in some of the substitution communities such as scrub or grasslands. A good
example of this is provided by gypsum soils, frequent in vast areas in central and
Dynamic-Catenal Vegetation Mapping ... 47

Fig. 6 Landscape representing a fragment of a successional series, with the woodland, the mantle
and the grassland, living on a mesologic homogeneous area (tessella). Each of the three plant
communities is determined by its position in the successional series and the disturbance regime
exerted upon it: the woodland represents a terminal position in the succession, probably the PNV
or a near stage; the mantle is a previous stage in the succession and is linked to episodic
disturbance due to fire or cutting; the grassland is a more initial stage in the successional trend and
is linked to a heavy grazing regime by sheep or cattle.

southern Spain, which determine the existence of specific scrub communities

but have no influence on the PNV of those areas. Another example is the
neighbouring effect observed in the transitional zones between two vegetation
belts in a mountain. In this case, the variability of a sigmetum is due to the
border effect and depends on the number of sigmeta in its vicinity which comes
into contact with.
• Variations due to biogeography. Geographical (or, more accurately, biogeo-
graphical) factors give rise to another type of variability within a vegetation
series or sigmetum. If the area of a sigmetum (tessella) is large, floristic dif-
ferences inevitably appear in its different “regions”. They often support
biogeographically-related phytosociological diversity, usually involving sub-
stitution stages of the PNV. Loidi [41] shows how the central Spanish associ-
ation Quercetum rotundifoliae (incl. Asparago-Quercetum rotundifoliae), which
occupies a vast area, can be replaced by no less than 12 geographical associa-
tions of scrub.
Both types of variability, ecological and biogeographical, are essentially different
in kind and should be expressed at different levels. The former is due mainly to
lithological changes or to the border effect, while the latter is the result of dif-
ferences in climate (past or current) and often of differences in vegetation history.
This suggests that biogeographical variability is more significant than ecological
variability and should therefore prevail when constructing a typology.
48 J. Loidi

• Variability due to alterations in the successional trend. Often, environmental

conditions are extreme in the sense that they impede the completion of a suc-
cessional trend. This happens in places where exceptional, often stressing,
conditions occur, leading to mosaics of microhabitats occupied by plant com-
munities more dependent on such conditions than on the general climate. These
plant communities, apparently arrested in succession, have traditionally been
called permanent communities. This term indicates that the stressing conditions
are more relevant than the general climate, which plays a secondary role in
conditioning the life of the community. Under such conditions, the succession is
slowed down or even canceled, remaining the community stable throughout
time (permanent). Each of these permanent communities usually develops on a
small-sized tessella which supports a small-sized sigmetum called a per-
masigmetum (perma-series). In these situations, the permasigmetum often has
a unique plant community; seral stages are lacking or occasionally restricted to
annual communities. In this category we can also include the extremely cold
climates, i.e. polar or alpine, as far as they impose severe conditions on plant
life. The permasigmeta often appear imbricated forming a mosaic or arranged
along a gradient [61, 64].
A variant in this is the minorisigmetum (minori-series), a concept initially
launched by Lazare and Lanniel [40] under the name of mesosérie and later
renamed as curtosérie or série tronquée (truncated) [39]. Finally the last term for
this concept was proposed by Rivas-Martínez et al. [65] as minorisigmetum.
Similarly to the permasigmetum, the minorisigmetum occurs in situations in which
the succession is restricted or blocked by a stressing condition which does not allow
it to reach the potential vegetation corresponding with the climate and substrate.
The difference with the permasigmetum is that the seral stages can be various and
have perennial non-nitrophilous stages, such as grasslands or scrub, completely
lacking in the permasigmetum. It is as if the higher potential vegetation stage is
prevented by the stressing conditions but not the seral ones. They can be found in
coastal areas (dune, cliffs and marshes), river banks, peat bogs, wind exposed crests,
etc., usually adjacent to the perma-series subjected to more stressing conditions.
The minorisigmetum can be regarded as an intermediate category between the
permasigmetum and the standard sigmetum.
Geosigmetum. This concept deals with the phenomenon of zonation. It can also
be called Geosigmassociation and Geoseries and is the basic unit of what is known
as geosynphytosociology. It is formed by a group of sigmeta arranged along a
gradient: each sigmetum, occupying its own tessella, will span a certain stretch of
the gradient. Thus, the geosigmetum has a spatial expression and represents a
catena of sigmeta (Fig. 7). Geosigmetum is the most synthetic concept in phy-
tosociology and, owing to the extreme variability and complexity of the situations
occurring in nature, some approaches must be made in order to formalize obser-
vations and sampling methodology and to develop a system able to organize the
resulting knowledge.
Dynamic-Catenal Vegetation Mapping ... 49

Fig. 7 Schematic
representation of a
Geosigmetum. The sigmeta
are arranged along a slope
gradient, a complex gradient
which encompasses soil
depth, nutrient richness and
soil moisture.

Fig. 8 Schematic representation of the topographic geosigmetum. The crest-slope-piedmont-

valley bottom model is of universal application. A: edphoxerophile sigmetum (the most
xerophytic), B: climatophile sigmeta (B1: submesophytic, B2: mesophytic, B3: temporihygropile).
C: edaphohygrophile (the most humid). They are arranged along a soil-moisture and soil-depth
gradient, favored by the gravity-driven oblique runoff of water and fine particles in the soil.
Erosion is prevalent on the crest and sedimentation on the valley bottom, where fine particles,
nutrients and water accumulate.
50 J. Loidi

Types of Geosigmeta
Several attempts have been made to study the variability of geosigmeta after
observation in the field. Theurillat [70] was the first, resulting from extensive field
work in the Swiss Alps. He established four major groups of geosigmeta depending
on their extent. Another classification of geo-permasigmeta was done by Alcaraz [1],
and it partially inspires what we propose here. We can divide the geosigmeta into
three main types depending on scale: topographic, altitudinal, and geographic [64].
GEOSIGMETA REFLECTING GENERAL CONDITIONS. The three categories expressed
below refer to zonation models at three spatial scales, from that representing the
general conditions of local climate (mesoclimate), to the climatic variation patterns
across vast areas, in the frame of the lithology and geomorphology existing in the
different territories of the world.
A). Topographic geosigmetum. Its expression can be outlined in the
crest-slope-piedmont-valley model (Fig. 8), and it bears the representation of
the fundamental distribution pattern of vegetation units in a territory. It is
related to the predominating climatic and geological conditions of that territory,
and can therefore be considered as its genuine geobotanical representation. In it,
the most xerophytic sigmetum (edaphoxerophilous) will occupy the driest and
most eroded crest position, the most hygrophilic (edaphohygrophilous) in the
bottom of the valley on the moister and deeper soils, and the intermediate
stretch encompassing the middle slopes and piedmonts will be occupied by the
climatophilous sigmeta—which can be divided into sub-mesophitic, mesophitic
and tempori-hygrophilous sigmeta. As outlined in Table 2, the middle stretch of
the slope reflects the ombroclimatic (euclimatope) conditions of the area, while
the crest will be under drier and the bottom under moister conditions [64].
This basic model is recognized within a particular vegetation belt and a con-
crete territory (a biogeographical district). It represents the fundamental pattern
of vegetational variability in an area of homogeneous climate, substrate and
history. The topographic geosigmetum is the main geobotanical descriptor of
any territory as it summarizes the great majority of the variability and rela-
tionships of its vegetation. It can also be used as an indicator of paleohistoric
information since edaphoxerophilous sigmeta correspond to past periods of dry
climate in which xerophilous vegetation could invade the area and, vice versa,
the edaphohygrophilous sigmetum to episodes of moister climate. These dry
and wet complexes can serve as a bridgehead for colonization in temporary
episodes of climatic change. The interpretation of thermophilic and cryophilic
vegetation examples in a given territory can similarly correspond to warm and
cold periods in the past. A smart interpretation of the geobotanical conditions in
a given area can provide us with important information about past migrations
driven by climatic changes [62, 64].
B). Cliserial geosigmetum. Cliserial refers to altitudinal zonation and is expressed
at the scale of an entire mountain or mountain range. The cliserial geosigmetum
consists genuinely of the altitudinal vegetation belts on any mountain.
Increasing elevation entails decreasing temperatures and, most commonly,
Dynamic-Catenal Vegetation Mapping ... 51

Fig. 9 Schematic representation of the cliserial geosigmetum. There are two variants,
homogeneous and heterogeneous depending on whether the rock material of the mountain is
homogeneous or heterogeneous.

increasing precipitation. If the mountain is high enough, the summit will bear a
tundra-like vegetation, or even perennial ice, while in the lower level a more
thermophilous vegetation will develop. This altitudinal zonation resembles the
broad-scale geographical latitudinal zonation existing in a continent by which
climatic zones arranged in parallel bands are inhabited by their particular
biomes, giving way to the rule of thumb that 1 m in altitude is equivalent to
1.3 km in latitude [38]. The zonation will be affected depending on whether the
rock material of the mountain is homogeneous or heterogeneous (Fig. 9).
C). Geographic geosigmetum. It is expressed in a geographical scale of an entire
region or continent and consists of latitudinal zonation of biomes or broadly
defined vegetation units arranged according to a gradient of regional climate.
This category overlaps with phytogeography and is not a target for this paper.

GEOSIGMETA UNDER STRESSING CONDITIONS. There are also other types of geosigmeta
present in any territory which at the topographic scale are subjected, in varying
degrees to conditions which partially or entirely mask the influence of the regional
climate.
52 J. Loidi

Table 2 Hydric balance in the three main sigmeta of a topographic geosigmetum

Sigmetum Edaphoxerophile Climatophile Edaphohygrophile
Moisture Input from Input from Input from
Balance rain > Losses and rain = Losses and rain < Losses and
gains from drainage gains from drainage gains from drainage
and ETP; negative and ETP; balance, no and ETP; positive
balance, net loss loss nor gain balance, net gain
Vegetation Drier than climate Equal to climate Moister than climate
in relation Xerophytic vegetation Climatophile Hygrophile vegetation
to climate vegetation

Fig. 10 A riverine geosigmetum, with all the complexity of the vegetation types associated with
the soil water table and flooding gradient. 1 to 13 represent the plant communities existing in the
riverine geosigmetum

Riverine Geosigmetum. This type constitutes the usual vegetation complex we have
called edaphohygrophilous, which is itself not a sigmetum but a geosigmetum. It
includes vegetation of the fluvial and peri-fluvial range, influenced by permanent,
intermittent or sporadic streams (Fig. 10; Fig. 11, C). It does not encompass the belt of
humid soils caused by a high water table unaffected by the stream (edaphohy-
grophilous sigmetum), usually included in the topographic geosigmetum as its most
humid component. Only the fringe directly exposed to the effects of the stream—
erosion, transport and sedimentation—is included in the riverine geosigmetum, i.e. the
space subject to hydric stress in the soils and to longer or shorter flooding by rapidly or
slowly running water. Plants bear adaptations to soil instability (periodical removal
and disposal of materials) in their roots and aerial anatomy as well as in their repro-
ductive and dispersal mechanisms. River types present high diversity all over the
world; from young mountain streams to mature riverine systems or from permanent
streams to those with dry river beds that only carry water when it rains, sometimes very
sporadically. The fluvial regime and the flooding of exposed areas or bands determine
the complexity of this geosigmetum, an essential component in any landscape [79].
Geopermasigmetum and Geominorisigmetum. Under situations of exceptional
stressing conditions the mosaics of microhabitats occupied by permasigmeta and/or
Dynamic-Catenal Vegetation Mapping ... 53

Fig. 11 Schematic representation of a complex of three geosigmeta: climatophilous geosigmetum

(A), riverine geosigmetum (C) and geopermasigmetum (B). A: 1- crest or steep slope
(submesophytic) sigmetum, 2- middle slope (mesophytic) sigmetum, 3- valley
(tempori-hygrophile) sigmetum; B: Rupicolous permaogeosigmetum; C: Riverine geosigmetum

minorisigmeta (Fig. 11, B) are called geo-permasigmeta and geo-minorisigmeta.

Complexes of this kind are usually linked to special edaphic and topographic
circumstances that have a much higher influence on vegetation than climate (e.g.
rock walls, ponds, mires, dunes, etc.).
Permasigmeta have received various names in the literature: microgeosigmeta
by Rivas-Martínez [60, 62], special geoseries by Alcaraz [1], and the term hipo-
geosigmassociation [69, 70] is also partially synonymous. The term geopermasig-
metum was finally established by Rivas-Martínez [64]. The geominorisigmetum has
also been called curtogeosigmetum [39].
We can distinguish the following 11 types of geo-perma-geosigmeta:
• CRIOGEOPERMASIGMETA. Low temperatures are the determining factor.
1. Polar (tundral) geopermasigmeta
Correspond to areas under cold climate (polar), where low temperatures are the
limiting factor in the ecosystems. Sedges, small chamaephytes, bryophytes and
lichens are the most common plant types.
2. High mountain geopermasigmeta
In high mountains (cryo belt) cold also is the limiting factor, similarly as in the
previous type. Alpine grasslands or dwarf shrublands are the most frequent
types of vegetation.
• SAXIGEOPERMASIGMETA. Outcrops of hard rock entail that soils are absent or
shallow; they are particularly frequent in mountains, where glacial and peri-
glacial erosion processes favor the existence of cliffs, wind exposed crests,
cryoturbation and gelifraction phenomena favoring the accumulation of screes.
54 J. Loidi

3. Rocky cliff geopermasigmeta

Vertical cliffs in a rocky mountain have a number of crevices in which some
very specialized stress tolerant species live. Sometimes, the tilt of the cliff is
negative giving way to overhangs. Oozing rocks represent a variant as well as
those cliffs with additional nitrogen nutrient inputs due to animals (bird nests)
or humans.
4. Scree geopermasigmeta
Screes formed by stone accumulation at the foot of cliffs, usually due to
gelifraction, are an unstable substrate which only specialized plants can colonize.
• CHIONOPHILOUS (SNOW) GEOPERMASIGMETA. Accumulated snow limits the
development of the vegetation living below.
5. Snowdrift geopermasigmeta
Zonation is determined by the permanence of snow, which in turn depends on
its degree of accumulation. Hollows and other areas where snow accumulates
are called snowfields and are found in snowy climates such as high mountains
and boreal or polar climates. The catena is determined by the length of the snow
coverage period. In Europe they are generally bryophytic, herbaceous or small
chamaephyte communities, but in other areas they can even be woodlands.
• HYGROGEOPERMASIGMETA. The stress is produced by the waterlogging of the
soil.
6. Freshwater inland geopermasigmeta
Ponds, lakes and swampy areas with still freshwater have a clear hydric gradient
of anoxic stress, from the moist to the permanently flooded soil. The vegetation
is composed by aquatic communities of submersed, floating and emersed plants,
together with the hygrophilous terrestrial vegetation of their borders.
7. Mires geopermasigmeta
Depressions that fill up with water in territories under cold climate and receive
organic matter which decomposition is slowed down or stopped. This organic
matter accumulates in the ponds and forms peat deposits. Such environments
are hostile for many vascular plants, being dominated by bryophytes; the most
frequent group of vascular plants are the Cyperaceae. They are very abundant
in Boreal and Arctic regions but are frequent also in temperate areas. In the rest
of the world, they are found only in high mountains. The influence of stagnant
water soaking up the peat moss establishes the gradient which drives the plant
communities zonation: the low mire rich in sedges develops at the water table
level while the high mire, rich in mosses (Sphagnum) and with some scrubs
(Erica, Vaccinium) rises from that level and depends more on the rainwater
supply. Some reed communities at the edges of the mire and some submersed
plants in the ponds (if present) complete the catena.
• HALOGEOPERMASIGMETA. High soil salinity is the main factor of stress.
Dynamic-Catenal Vegetation Mapping ... 55

Fig. 12 Iberian Atlantic coastal sand dune system. Each of the bands represents an entire
sigmetum (permasigmetum) as it occupies a permatesella. The combined gradient of salinity and
wind exposure is responsible for the zonation

8. Halophilous tidal geopermasigmeta

Coastal salt marshes influenced by tides and sheltered from waves are popu-
lated by vegetation adapted to the high sea water salinity. The tidal floods
determine the gradient. Vascular vegetation is dominated by succulent
Chenopodioideae in the extra-tropical countries and by mangroves in tropical
countries where warm waters dominate.
9. Halophilous inland geopermasigmeta
In the inland endorrheic depressions or basins, soluble salts accumulate in the
soil under certain climatic and geohydrologic conditions, reproducing partially
the salinity levels of the coastal salt marshes. The gradient is given by the saline
concentration in the soil water.
10. Coastal dunes geopermasigmeta
Between the beach and the fixed dunes, the coastal sandy deposits are modelled
by the wind and salinity brought by the sea spray, creating a gradient of
mobility and salinity which clearly results in the zonation of the foredunes,
white or mobile dunes and grey dunes typical of this geopermasigmetum all
over the world. Substrate mobility and coarseness (sand) entail dry edaphic
conditions (almost immediate percolation and no water retention capacity)
which are added to the more or less intense salinity brought by the sea spray (at
least in the fringes nearest to the sea). The wind itself has a strong evaporation
effect and a mechanical abrasive action on the plants (Fig. 12).
56 J. Loidi

Fig. 13 Example of permageosigmetum where the permasigmeta are the vegetation belts arranged
along the rockiness, slope steepness, wind exposure and salinity combined gradients

11. Coastal halo-anemogenous rupicolous geopermasigmeta

Coastal areas with cliffs are influenced by sea winds which impose their
mechanical and desiccating action and bring small drops of sea water spreading
marine salt on the cliffs. They usually occupy a narrow strip in the upper border
of the cliffs, but as the influence of the sea increases during storms, in exposed
windy capes this geopermasigmetum can reach to a level of more than 100 m
high and cover large areas of several kilometres inland. Topography and
rockiness is the second major factor in determining the catena, as it is related to
the capability of building a stable soil (Fig. 13).

Methodological Considerations in Geosynphytosociology

The geosigmetum is the most synthetic concept in dynamic-catenal phytosociology
(or phytosociology sensu lato) and, owing to the extreme variability and complexity
of the situations that occur in nature, an attempt needs to be made in order to
formalize observations and sampling methodology and to facilitate a system for
organizing knowledge. There is no limit to the number of sigmeta to be included in
Dynamic-Catenal Vegetation Mapping ... 57

Fig. 14 Comparison of the zonation (oreogeosigmeta) between two mountains of the same
biogeographical unit but different altitudes. Mountain A has a “longer” catena than B simply
because it is higher. Geobotanically both are equivalent as they respond to a similar pattern of
vegetation distribution and this is determined by a similar set of environmental conditions, climatic
and edaphic

the sample (geosigma relevé) as long as they are defined. Only geography can
impose limitations: the height of mountains, lithological and geomorphological
variability, etc. So it is clear that the number of sigmeta occurring in a territory will
depend on its size and biotope variability: vast but homogeneous flat countries will
contain few sigmeta while mountainous areas with complicated geology and many
local climatic situations will show many as an expression of their differences in
biodiversity. The geosigmeta of each of the two kinds of territory will differ greatly
in the number and definition of their sigmeta.
When using the concept of geosigmetum in field work, it is essential to know
how to carry out the sampling procedure, which in our case means knowing where
the limits of the geosigma relevé are. The comparability of the samples obtained in
different territories must be ensured, as they can be used to assess the geobotanical
differences or similarities between the territories. For example, consider two similar,
neighboring valleys in the Alps, one of which has a higher mountain that rises up to
the alpine belt, and the other with lower altitudes, with only the subalpine belt being
represented at its summits (Fig. 14). The cliserial catena (cliserial geosigmetum) of
both valleys will only differ in their “length”, because the first one will have alpine
grasslands and the second will not, but the rest of the catena is similar because the
conditions, mainly climatic and geological, are the same. The comparison of both
geosigma relevés can give us a misleading idea of dissimilarity but the differences
are due to topographical circumstance, not to climatic, geological or biogeo-
graphical conditions. Both valleys are, using these latter criteria, geobotanically
identical; one simply happens to have a higher mountain than the other.
To overcome difficulties of this kind, as Alcaraz [1] suggests that, in order to
ensure that the samples are comparable, some standardizing criteria should be
followed when sampling. In the case of topographic geosigmeta:
58 J. Loidi

1. By placing the geosigma relevé within a single homogeneous bioclimatic area,

i.e. within one thermotype and ombrotype in the typology of Rivas-Martínez
[64]. For high mountains with a wide climatic gradient from the bottom to the
top, segmentation in bioclimatic belts is required, sampling the topographic
geosigmetum within each belt. This way each of the samples will be comparable
with its bioclimatic homologue from another mountain.
2. By placing the geosigma relevé within a single type of lithological substratum,
at least with regard to its trophic properties. A climatically homogeneous ter-
ritory can support lithological diversity with strong influence in vegetation. The
differences between base-rich and base-poor vegetation inside a homogeneous
climatic region are clear.
3. By placing the geosigma relevé within a unique type of geomorphological unit
in which the catena is due to a single gradient or group of associate gradients.
This permits the separation of riverine geosigmeta (determined by intensity of
flooding, flow of running water, etc.) from adjacent ones not directly influenced
by river water, and also means that salt marsh geosigmetum, where salinity is
the main factor for determining the catena, can be separated from neighboring
geosigmeta.
A complete explanation of the methodology to sample coastal geopermasigmeta
and geominorisigmeta is given by Demartini [20].
Geobotanical Contextualization
To the categories of dynamic-catenal phytosociology described above, can be
added those of bioclimatology (e.g. those by Rivas-Martínez [59, 61, 64]). They
provide the climate typological structure that orders and organizes climatic infor-
mation from the perspective of vascular plants by aiming to adjust the typology to
the distribution patterns of vegetation. This is done by selecting the climatic
parameters with most influence on the distribution of vascular plants on earth. This
bioclimatic typology provides a climatic framework within which plant commu-
nities live and develop describing the climatic scenery of any stretch of landscape, it
offers a sort of description of the climatic context. For that reason, typological
bioclimatology plays an important role in understanding vegetation units better.
This can also be said of typological biogeography (provincialism) and
dynamic-catenal phytodsociology, all of them synthetic disciplines: they summarize
information on vegetation and environmental conditions.
This issue is relevant from a purely theoretical point of view. The synthetic
typology provided by dynamic-catenal phytosociology at any of the integration
levels, together with the typologies provided by bioclimatology, biogeography and
by the earth sciences (edaphology, geomorphology, lithology, etc.), makes geob-
otanical contextualization possible, which is very important in all studies of plant
communities. This is an exercise of considering all the biotic and abiotic conditions
converging on a given area or site in which the plant community we are interested
in lives, so that we can “put it in place” before we start with a survey. Geobotanical
contextualization, in a not so formal way, has been almost routine, especially in
European schools, since the early times of Geobotany [36]; today it can be done
Dynamic-Catenal Vegetation Mapping ... 59

much more accurately thanks to the available typology recently developed.

Geobotanical contextualization is a very convenient preliminary exercise when
planning any study or research on site or area vegetation. The results of the research
on any field of vegetation science should be placed in context in order to calibrate
their validity and potential for extrapolation to other areas.
The aim is to build a universal typological system with explanatory power upon
the distribution patterns of vegetation and valid for any portion of land on Earth. The
system will enable workers to draw the outline of the geobotanical scenery of any
region, using unified criteria, in such a way that results of research or observations on
vegetation obtained in different territories may be correctly compared, by taking into
account the similarities (or differences) of their respective geobotanical sceneries.
As an example of this comparison exercises, consider the results obtained from a
survey on biodiversity of lauroid forest in central eastern China: how can we
compare them to other types of forests in the world? Extrapolating to other phys-
iognomically similar lauroid forests in the world is the initial procedure, i.e. we
would compare it with the lauroid forests of Japan, eastern North America, etc.; but,
what about the mountain lauroid forest in the Canary Islands (Canarian Laurisilva or
Monteverde)? If we know that the first one is found in subtropical climates governed
by the monsoon, which causes maximum rainfall in summer, while the second
survives thank to the persistent fogs which occur on the northern slopes of the islands
caused by the trade winds, in a pronounced Mediterranean context, then the com-
parison can be more accurately modulated. In this case of the lauroid forest, the
bioclimatic study immediately shows that the annual distribution pattern of rainfall
in both areas is diametrically opposed: in China, maximum rainfall occurs in summer
(monsoon) whereas summer is the driest season in the Canary Islands, corresponding
as it does to its Mediterranean climatic model [23]. Continentality is also much
greater in China than in the extremely oceanic Canary archipelago, where frost is
practically absent from the “laurisilva” level [23]. Such differences are certainly
important, as they determine plant reproduction and growth cycles to a great extent.
Then, when analyzing the flora and, consequently, the plant communities, from this
contextual perspective, differences become quite profound. No scientific work can
afford to ignore such circumstances, even if research has nothing to do with syn-
taxonomy or biogeography and focuses on other fields of ecology. When the results
of searches for biodiversity measures are considered in this light, their interpretation,
when compared to the Canarian Monteverde [63], is only relative due to the long
“geobotanical distance” existing between central China and the Canary Islands.
Geobotanical contextualization puts the eventual distances to similarities.
Conversely, China and Japan are closely related areas, biogeographically, bio-
climatically and historically, so the comparisons are easier to interpret from the
perspective of vicinity. If we compare distant territories, not only geographically
but also climatically, biogeographically and historically, although they present a
noticeable morphological and structural convergence in their forest types, the
comparisons have to be done with much more care and taking into account that
distance.
60 J. Loidi

Fig. 15 After Blasi et al. 2000, cartography of sigmeta and geosigmeta over a determined region
can be performed following a deductive landscape analysis in which different attributes are
considered in a hierarchical order, being applied for different scales.

Thus, geobotanical contextualization should be included in the preliminary

designs and planning of any experiment or observation on plant communities. This
will help us to construct hypotheses and to assess the significance of the results in
relation to their potential for general application. For some authors such as Alcaraz
[1] the theoretical model provided by dynamic-catenal phytosociology can be
considered a contribution to the formulation of a “General Theory of Vegetation
Science” with sufficient capacity for general application for it to be considered a
major contribution to the paradigm of Vegetation Science.
Cartography of Sigmeta and Geosigmeta
To produce a map of sigmeta and geosigmeta, in most cases an inductive process is
followed, i.e. initial surveys of the flora and plant communities of the territory give
way to a recognition of the units to be mapped and then the cartography is carried
out. Nevertheless, there is an interesting approach done by some Italian colleagues
[7, 8] in a deductive-inductive way, by which, initially, the landscape is succes-
sively analyzed after several attributes. These classifying attributes are, in hierar-
chical order of space and time: climate, lithology, geomorphology, human
activities, soil, vegetation and fauna [22, 78]. This order of attributes moves from
relatively stable factors controlling larger areas, such as climate or lithology, to
more dynamic factors operating locally, so that fine-scale factors such as human
land use are very variable in space and time (Fig. 15). The tessella would contain
Dynamic-Catenal Vegetation Mapping ... 61

the four upper levels which ensure the ecological homogeneity, and the lowest one,
with the land use, gives way to the existing plant communities.
One of the clearer, more straightforward tasks of this dynamic-catenal phy-
tosociology is cartography. This has been done several times and using several
scales in different countries of Europe (e.g. [55, 58, 47, 70, 43, 14, 7, 20]. In recent
decades this type of mapping has obtained full recognition in the frame of the
European Union documentation for inventories of natural resources [37]. A map
representing sigmeta and geosigmeta is an excellent document summarizing the
biological (vegetation) content of a territory in terms of its ecologic affinities and an
expression of its potentialities. It is a map of the different vocations of the land built
up by the integration of the climatic, topographic, lithologic, geomorphologic,
edaphic, vegetation and human land use information. This ultra-synthetic approach
enables the use of mapped units for targeting ecological restoration, if that is our
intention, because they provide the referential frames for defining the Potential
Natural Vegetation of any piece of territory. They also can be used to evaluate the
ecologic quality of any area and, in general, as basic information for any project in
land management. The interest of such cartography in the design of strategies for
landscape management as well as for the protection of biodiversity and landscape
quality becomes evident.

References

1. Alcaraz F (1996) Fitosociología integrada, paisaje y biogeografla. In: Loidi J (ed) Avances en
Fitosociologia: 59–94. Ed. UPV/EHU
2. Asensi A (1996) Fitosociología y paisaje (Una aproximación histórica). In: Loidi J
(ed) Avances en Fitosociologia. 43–58. Ed. Universidad del País Vasco
3. Béguin C, Géhu JM, Hegg O (1979) La symphytosociologie: une approche nouvelle des
paysages vegetaux. Docum. Phytosociologiques NS 4:49–68
4. Béguin C, Hegg O (1975) Quelques associations d’associations (sigmassociations) sur les
anticlinaux jurassiens recouvertes d’une végétation naturelle potentielle (essai d’analyse
scientifique du paysage). Docum Phytosociologiques 9–14:9–18
5. Biondi E (1996) L'analisi fitosociologica nello studio integrato del paessagio. In: Loidi J
(ed) Avances en Fitosociologia: 13–22. Ed. UPV/EHU
6. Biondi E (2011) Phytosociology today: methodological and conceptual evolution. Plant
Biosystems 145:19–29
7. Blasi C (ed) (2010) La vegetazione d’Italia. Carta delle Serie de Vegetazione in scala
1:500.000. Palombini & Partner, Roma
8. Blasi C, Carranza ML, Frondoni R, Rosati L (2000) Ecosystem classification and mapping: a
proposal for Italian landscapes. Appl Veg Sci 3:233–242
9. Blume H-P, Sukopp H (1976) Ökologische Bedeutung anthropogener Bodenveränderungen.
In: Sukopp H, Trautmann W (eds) Veränderungen der Flora und Fauna in der Bundesrepublik
Deutschland. Schriftenr. Vegetationskd, vol 10, pp 75–89
10. Bolòs O (1962) El paisaje vegetal barcelonés. Ayuntamiento, Barcelona, Cátedra Ciudad de
Barcelona, Excmo
11. Bolòs O (1963) Botánica y geografía. Mem Real Acad Ci Artes Barcelona 34:443–491
62 J. Loidi

12. Braun J, Furrer E (1913) Remarques sur l’étude des groupements de plantes. Soc Languedoc
Geogr Bull 36:20–41
13. Braun-Blanquet J (1928) Pflanzensoziologie. Grundzüge der Vegetationskunde. Biologische
Studienbücher. Berlin, vol 7, p 330
14. Capelo J, Costa JC, Ribeiro S, Arsénio P, Neto C, Monteiro-Henriques T, Aguiar C,
Honrado J, Espirito-Santo D, Lousã M (2007) A methodological approach to potential
vegetation modeling using GIS techniques and phytosociological expert-knowledge: appli-
cation to mainland Portugal. Phytocoenologia 37(3–4):399–415
15. Braun-Blanquet J, Pavillard J (1928) Vocabulaire de sociologie végétale. 3 éme ed.
Montpellier
16. Chiarucci A, Araújo MB, Decocq G, Beierkuhnlein C, Fernández-Palacios JM (2010) The
concept of potential natural vegetation: an epitaph? J Veg Sci 21:1172–1178
17. Clements FE (1905) The development and structure ofvegetation (Studies in the vegetation on
the state 111). Lincoln
18. Clements FE (1916) Plant succession: an analysis of the development of vegetation.
Washington, 515 pp
19. Decoq G (2016) Moving from patterns to processes: a challenge for the phytosociology on the
twenty-first century. In: Box O (ed) Vegetation structure and function at multiple spatial,
temporal and conceptual scales. Geobotanic studies. Springer, pp 407–424
20. Demartini C (2016) Les végétations des côtes Manche-Atlantique françaises: essai de
typologie et de cartographie dynamico-caténales. PhD. Université de Bretagne Occidentale,
Brest
21. Dierschke H (1994) Pflanzensoziologie. UTB für Wissenschaft. Eugen Ulmer Verlag,
Stuttgart
22. Forman R, Godron M (1986) Landscape Ecology. Wiley & Son, New York
23. Gandullo JM (ed) (1991) Estudio ecológico de la laurisilva canaria. ICONA, Madrid
24. Gaussen H (1964) Carte de la végétation de la France 1:200.000, fe u ille 77 Foix. CNRS.
Toulouse
25. Géhu JM (1974) Sur l’emploi de la methode phytosociologique sigmatiste dans l’analyse, la
définition et la cartographie des paysages. CR Acad Se Paris 279:1167–1170
26. Géhu JM (1976) Sur les paysages végétaux ou sigmassociations des prairies salées du
Nord-Ouest de la France. Documents Phytosociologiques 15–18:57–62
27. Géhu JM (1977) Le concept de sigmassociation et son application a l’étude du paysage
végétal del falaises atlantiques françaises. Vegetatio 34(2):117–125
28. Géhu JM (1978a) Les sigmassociations de la xérocére des dunes atlantiques françaises de
Dunkerke a Biarritz. In: Tüxen R (ed.) Assoziationskomplexe (Sigmeten). Bericht Intern.
Symposium 1977 in Rinteln: 77–82. Vaduz
29. Géhu JM (1978b) Premieres éléments pour un sigmasystéme des dunes séches holarctiques.
In: Tüxen R (ed) Assoziationskomplexe (Sigmeten). Bericht Intern. Symposium 1977 in
Rinteln: 267–272. Vaduz
30. Géhu JM (1979) Pour un approche nouvelle des paysages végétaux: la symphytosociologie.
Bull Soco Bot Fr 126:213–223
31. Géhu JM (1986) Des complexes de groupements végétaux a la phytosociologie paysagére
contemporaine. Inform Bot Ital 18:53–83
32. Géhu JM (2006) Dictionnaire de sociologie et synécologie végétales. Cramer. Berlin/
Stuttgart, p 899
33. Géhu JM, Rivas-Martínez S (1981) Notions fondamentales de phytosociologie. In:
Dierschke H (ed) Syntaxonomie. Bericht Intern. Symposium 1977 in Rinteln, pp 5–33
34. Gillet F, Gallandat J-D (1996) Integrated synusial phytosociology: some notes on a new,
multiscalar approach to vegetation analysis. J Veg Sci 7:13–18
35. Hardtle W (1995) On the theoretical concept of potential natural vegetation and proposals for
an up-to-date modification. Folia Geobot. Phytotax. Prague 30:263–276
36. von Humboldt A, Bonpland A (1807) Essai su la géographie des plantes accompagné d’un
tableau physique des régions équinoxiales. Schoell, Paris
Dynamic-Catenal Vegetation Mapping ... 63

37. Ichter J, Evans D, Richard D (eds) (2014) Terrestrial habitat mapping in Europe, an overview.
Luxembourg. MNHN-EEA Technical report, p 153
38. Körner C (2003) Alpine plant life Functional plant ecology of high mountain ecosystems.
Springer, p 344
39. Lazare J-J (2009) Phytosociologie dinamico-catenale et gestión de la biodiversité. Acta Bot
Gallica 156(1):49–61
40. Lazare J-J, Lanniel K (2003) Une sous-association nouvelle de fourrés du Rubo
ulmifolii-Tametum communis du littoral basque. J De Botanique De La Société Botanique
De France 21:33–35
41. Loidi J (1991) Vegetation series: its use for small scale geobotanical mapping. Phytocoenosis
(N.S.) 3:119–122
42. Loidi J, del Arco M, Pérez de Paz PL, Asensi A, Díez Garretas B, Costa M, Diáz González
TE, Fernández-González F, Izco J, Penas A, Rivas-Martínez S, Sánchez-Mata D (2010)
Understanding properly the “potential natural vegetation” concept. J Biogeogr 37:2209–2211
43. Loidi J, Báscones JC (1995) Memoria del mapa de series de vegetación de Navarra. Ed.
Gobierno de Navarra, Depto. de Ordenación del Territorio y Medio Ambiente. Pamplona
44. Loidi J, Biurrun I, Campos JA, García-Mijangos I, Herrera M (2011) La vegetación de la
comunidad autónoma del País Vasco. Leyenda del mapa de series de vegetación a escala
1.50000. University of the Basque Country (UPV/EHU). Ed. Eusko Jaularitza/Gobierno
Vasco, p 197
45. Loidi J, Fernández-González F (2012) Potential natural vegetation: reburying or reboring? J
Veg Sci 23:596–604
46. Matuszkiewicz W (1979) Landscape phytocomplexe and vegetation landscapes, real and
typological landscape units of vegetation. Documents Phytosociologiques N.S. 4:673–688
47. Matuszkiewicz W (1984) Die Karte der potentiellen natürlichen Vegetation von Polen.
Braun-Blanquetia 1:1–99. Camerino
48. Milne G (1935) Some suggested units of classification and mapping particularly for East
African soils. Soil Research 4:183–189
49. Morison CGT, Hoyle AC, Hope-Simpson JF (1948) Tropical soil-vegetation catenas and
mosaic. A study in the south-western part of the Anglo-Egyptian Sudan. J Ecol 36(1):1–84
50. Mucina L (2010) Floristic–phytosociological approach, potential natural vegetation, and
survival of prejudice. Lazaroa 31:173–182
51. Ozenda P (1985) La végétation de la chaine alpine dans l’espace montagnard, européen.
Masson, Paris
52. Ozenda P, Wagner H (1975) Les séries de végétation de la chaine alpine et leurs équivalences
dans les autres systémes phytogéographiques. Doc Cart Ecol 16:49–74
53. Pignatti S (1980) Epharmonische Beziehungen zwischen struktureller und funktioneller
Ordnung in der Vegetation. In: Tüxen R (ed.) Ber. Int. Sympos. Int. Ver. Vegetationskunde
1979. Vaduz, pp 385–396
54. Rivas-Martínez S (1976) Sinfitosociología, una nueva metodología para el estudio del paisaje
vegetal. Anales Inst. Bot. Cavanilles 33:179–188
55. Rivas-Martínez S (1982) Mapa de series de vegetación de la provincia de Madrid. Diputación
de Madrid
56. Rivas-Martínez S (1983) Series de vegetación de la región Eurosiberiana de la Penlnsúla
Ibérica. Lazaroa 4:155–166
57. Rivas-Martínez S (1985) Biogeografia y vegetación. Discurso de ingreso en la Real Academia
de Ci. Exactas, Físicas y Naturales. Madrid
58. Rivas-Martínez S (1987) Memoria del mapa de series de vegetación de España, 1: 200.000.
ICONA. Madrid.
59. Rivas-Martínez S (1993) Bases para una clasificación bioclimática de la Tierra. Folia Botanica
Matritensis 10:1–23
60. Rivas-Martínez S (1994) Dynamic-zonal phytosociology as landscape science.
Phytocoenologia 24:23–25
64 J. Loidi

61. Rivas-Martínez S (1996) La fitosociología en España. Discurso de investidura como Doctor

honoris causa. Universidad del País Vasco: In: Loidi J (ed) Avances en fitosociología. Ed.
UPV/EHU, pp 149–176
62. Rivas-Martínez S (1996) Geobotánica y bioclimatologia. Discurso de investidura como
Doctor honoris causa. Universidad de Granada: 23–98. Granada
63. Rivas-Martínez S, Wildpret W, Arco M del, Rodríguez O, Pérez de Paz PL, García-Gallo A,
Acebes-Ginovés R, Díaz TE, Fernández-González F (1993) Comunidades vegetales de la Isla
de Tenerife (Islas Canarias). ltinera Geobotanica 7:169–374
64. Rivas-Martínez S, et al (2007) Mapa de series, geoseries y geopermaseries de vegetación de
España (Memoria del mapa de vegetación potencial de España,). Parte I. Itinera Geobotanica
17:5–436
65. Rivas-Martínez et al (2011) Mapa de series, geoseries y geopermaseries de vegetación de
España (Memoria del mapa de vegetación potencial de España, 2011). Parte II. Itinera
Geobotanica 18(1,2):5–800
66. Russell EW (1961) Soil conditions and Plant Growth, 9th edn. Longmans, London
67. Schmithüsen J (1959) Allgemeine Vegetationsgeographie I, Ed. De Gruyter. Berlin
68. Schwabe A (1990) Stand und Perspektiven der Vegetationskomplex-Forschung. Berichte der
Reinhold-Tüxen-Gesellschaft 2:45–60
69. Theurillat JP (1992a) L’analyse du paysage végétale en symphytocoenologie: ses niveaux et
leur domains spatiaux. Bull Ecol 23(1–2):83–92
70. Theurillat JP (1992b) Étude et cartographie du paysage végétal (Symphytocoenologie) dans la
région d'Aletsch (Valais, Switzerland). Beitr. Geobot. Landesaufn. Schweiz 68, pp 384.
Geobot. Komm. Schweiz. Akad. Naturwissenschaften. Krypto, Teufen
71. Tüxen R (1956) Die heutige potentielle natürliche Vegetation als Gegenstand der
Vegetationskartierung. Angewandte Pflanzensoziologie 13
72. Tüxen R (1973) Vorschlag zur Aufnahme von Gesellschaftskomplexen in potentiell
natürlichen Vegetationsgebieten. Acta Bol. Acad. Sci. Hungar 19(1–4):379–384
73. Tüxen R (1977) Zur Hornogenität von Sigmassoziationen, ihrer sytaxonomischen Ordnung
und ihrer Verwendung in der Vegetationskartierung. Docum. Phylosociologiques N.S. 1:321–
327
74. Tüxen R (1978a) Bemerkungen zur historischen, begrifflichen und methodischen Grundlagen
der Synsoziologie. In: Tüxen R (ed) Assoziationskomplexe (Sigmeten), Bericht. Intern.
Symposium 1977 in Rinteln: 3–12. Vaduz
75. Tüxen R (1978b) Versuch zur Sigma-Syntaxonomie mitteleuropaischen
Flusstal-Gesellschaften. In: Tüxen R (ed) Assoziationskomplexe (Sigmeten), Bericht.
Intern. Symposium 1977 in Rinteln: 273–286. Vaduz
76. Vigo J (1998) Some reflections on geobotany and vegetation mapping. Acta Bot. Barc.
45:535–556
77. Zerbe S (1998) Potential natural vegetation: validity and applicability in landscape planning
and nature conservation. Appl Veg Sci 1:165–172
78. Zonneveld IS (1995) Land Ecology. SPB Academic Publishing, Amsterdam
79. Schwabe A (1989) Vegetation complexes of flowing water habitats and their importance for
the differentiation of landscape units. Landscape Ecol. 2(4):237–253
80. Schmithüsen J (1948) Fliesen Gefüge der Landschaft und Ökotop. Berndt. Landesk. 5:74–83
Climate Change, Wetland Management
and Alpaca Pastoralism in the Bolivian
High Andes Mountains

Máximo Liberman

Abstract This paper draws on an environmental management study of a com-

munity of alpaca shepherds from a high Andean region in the Apolobamba
mountain range in northwestern Bolivia, at an altitude of 4300–4800 m. In this
region, Aymara indigenous families face the impacts caused by persistent droughts
and climate change derivatives. In the work that was carried out, a series of sci-
entific evidences are presented that demonstrate the dramatic retreat of the glaciers
based on measurements and the multitemporal interpretation of satellite images,
between 1985, 2014 and 2019. The community’s perceptions of climate change
were assembled through interviews with alpaca shepherds and through participatory
meetings with all the members of the community of Cañuma, at the Franz Tamayo
province of the departmental region of La Paz. Through talking schemes and maps,
all the climate-related problems they are going through were revealed, especially
how decreasing water availability negatively affects sites of very high fragility, such
as wetlands, known as the main source of alpaca grazing. Novel techniques for
adapting to climate change are described and mapped, such as water conduction
systems. These canals are built with surprising knowledge in such a way that it has
allowed communities to increase wetland areas for their camelid cattle. These
models of hydraulic management in the high mountains represent sustainable
patterns of adaptation and effective responses to climate change that members of the
community themselves have designed and implemented. These “local environ-
mental engineering” techniques produced an increase in alpaca grazing areas and
greater forage productivity in slope wetlands. This means that today, in times of
climate change, the provision of food for cattle is guaranteed for this community.

Keywords Climate change Wetlands Pastoralism Alpacas High mountains




Andes Bolivia

M. Liberman (&)
Institute of Ecology, Universidad Mayor de San Andrés, Cota Cota - University Campus,
La Paz, Bolivia

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 65

F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_5
66 M. Liberman

1 Introduction

The High Andean region of Bolivia, which lies above 4000 m, is where the Aymara
civilization developed. Its autonomy and originality rest on its ability to use natural
resources through particular socio-cultural strategies and arrangements. Glaciers are
vital for the indigenous communities that live in the high mountains. One of the
skills that characterize the people from the upper part of the Apolobamba Protected
Area, near the glaciers, is their use of grasslands and wet meadows, known locally
as bofedales, for grazing the alpacas and llamas that provide them with highly
valued and productive resources.
Over the past approximately three decades, climate change has begun to modify
rainfall patterns and thus the availability of water in some areas. The bofedal
wetlands regulate the flow of water, which affects access to drinking water and food
for the camelid cattle. Referring to this production level, herders have made a set of
environmental modifications, since they not only take advantage of natural grass-
lands. Peasants from the plains and valleys in the high Apolobamba mountain
range, in the Bautista Saavedra and Franz Tamayo provinces of La Paz department,
through different water-conduction techniques, have managed to modify the natural
environment, in order to have greater areas of wet meadows or additional bofedales
that allow them to feed their camelid herds [25]. Due to the effects of global climate
change, families of camelid breeders in the high mountains are becoming more
vulnerable every year [22].
Livestock populations have been disturbed by the intensifying climate phe-
nomena (rising temperatures that decrease glacier surfaces) and microclimatic
extremes (snowfalls, frosts, hailstorms and strong winds). These climatic changes
have caused a retreat of the glaciers, producing melting ice and greater water runoff.
New lagoons have appeared in many places where there were once glaciers. As a
result there has been a gradual colonization of these sites by a range of bofedal
species.
This study aims to assess climate-change effects on glaciers, wetlands and
camelid grazing in the Apolobamba Mountain Range by evaluating a sequence of
satellite images. Climate-change perceptions and experience of the local population
were evaluated through a participatory workshop with an Aymara indigenous
community. The goal was to propose adaptation mechanisms and strategies to
minimize negative effects of climatic phenomena. Other objectives are the rescue of
local knowledge, its management actions and good management practices for
bofedales and camelid cattle.
Climate Change, Wetland Management and Alpaca Pastoralism … 67

2 Methodology

Based on multi-temporal interpretation of satellite images from 1985, 2014 and

2019, this study presents scientific evidence that demonstrates the dramatic retreat
of the glaciers. Sentinel-2 satellite images were used to identify terrestrial changes
on Earth. This multispectral scanner obtains information in two visible and infra-red
wavelengths, allowing one to monitor changes on the earth and in the vegetation, as
well as to monitor climate change itself. The scanner has a resolution of 10 m.
Three scenes (optical images) from the satellite were used, from 18 July 1985, 15
August 2014, and 29 May 2019. The images cover the entire surface of the
Apolobamba protected area. Each scene has 13 multispectral bands that were
analyzed and combined to highlight the objective of the observation. Subsequently,
a reclassification was carried out for the three different years, by separating the
pixels with positive values, to identify the different glacier surfaces.
A social and biodiversity appraisal of the Cañuma community, located in the
Apolobamba region, is presented. This information was collected in two field trips
during June and July 2016. Several interviews were conducted with camelid her-
ders, and observations on biodiversity management were made. In addition, the
following literature was consulted: Liberman [23], SERNAP (2005), Flores Ochoa
[8] and Hoffmann et al. [20].
A participatory workshop on the effects of climate change and its environmental
impacts and consequences on the management of the wet meadows and the grazing
of camelids in the high Andes was organized with the community. The participatory
method enabled participants to rescue historical elements of peasant planning and
self-assessment. In this case, so-called talking maps were used as methodological
technical instruments that allow adequate communication within the community, by
diagramming scenarios of what happens with natural resources on territorial maps.
On these maps, which were drawn by groups of herders, they collect graphically the
participants’ perceptions of their territory, strengthening their peasant identity.
Thus, the map reflects the most important aspects of its territory, such as glaciers,
bofedales, grasslands, homes, water sources, and the problems that each sector has.
The map making was carried out through a three-step participatory process: a)
Explanation and socialization of the methodology; b) Drawing the maps, with
reference to mountains, roads, paths, houses, water sources, location of wetlands,
grazing areas, etc.; c) Presentation by technicians from the University of San
Andrés of their results on climate and land use analyses. Once the maps were
drawn, the different groups presented talking maps identifying the similarities and
shortcomings of each Then the information presented by the herders was contrasted
with that of the technicians, with the purpose of drawing conclusions and recom-
mendations from the knowledge dialogue that includes the effects of climate change
and the mitigation and adaptation measures developed by the farmers.
The construction of new rustic canals was highlighted, based on the people’s
technology to increase the surface area of bofedales through gravity-driven systems
to optimize irrigation. Climate indicators were identified based upon the results of
68 M. Liberman

technical indicators of aging, such as: frost, hailstorms, rains and droughts versus
mortality, morbidity, birth and fertility, as well as management of the diversity of
colors in the Alpacas. Rotation of grazing areas and use of resting fences, health
management of camelid cattle, fertilization of natural pastures, and establishment of
Q’ochas or small lagoons, water intake and fences with acquired materials were
also considered.
All this information is aimed at generating technical and scientific data
belonging to the Cañuma community, for the optimization of water resources,
pasture management, management of camelid cattle and adaptation of alpaca herder
families to climate change. The purpose is to guarantee sovereignty with food
security and conservation of life systems. Likewise, community activities are
analysed in conservation, protection, restoration and integral management of waters
and bofedales. This concerned especially the irrigation systems (main and sec-
ondary channels), by prioritizing the use of water for communities and grazing of
camelid livestock.

3 Location of the Study Area

The traditional grazing system with camelids in the Apolobamba region, analyzed
in this work, is located in the Ulla-Ulla high Andean plain and glacial valleys of the
Apolobamba Range (Department of La Paz, Franz Tamayo province), in the
Bolivian puna between 15° 00′ and 15° 10′ south latitude and 69° 05′ and 69° 20′
west longitude, above 4300 m (Map 1). The Apolobamba Mountain Range sepa-
rates these plains from the eastern valleys that descend into the Yungas (montane
forests). The Pleistocene glaciers of the Cordillera left formations of moraines, lakes
and lagoons that today cover the Ulla-Ulla plateau and the Apolobamba mountains.
The dominant vegetation in the plains and mountains is puna grasslands of Stipa
spp., Festuca spp and Deyeuxia spp. in different associations, and communities of
Plantago tubulosa and Distichia muscoides, that dominate places saturated with
water, and known locally as bofedales [28]. These wetlands can be natural or,
instead, looks similar to the enormous extents that cover the Apolobamba plains,
produced and maintained through irrigation. Water is channelled from rivers that
descend from lagoons of glacial origin. The intensive study area is located in the
highlands and glacial valleys of the Apolobamba mountains. Studies were con-
centrated in a community called Cañuma, whose members are dedicated exclu-
sively to llama and alpaca breeding, in communal properties, since pre-Hispanic
times.
Table 1 shows characteristics of the region, where the Apolobamba wetlands
(bofedales) occur.
Figure 1 shows a schematic altitudinal profile of the study area and the location
of the community of Cañuma and its relationship with the altitudinal belts and use
on the eastern slope of the Apolobamba mountain range, Bolivia.
Climate Change, Wetland Management and Alpaca Pastoralism … 69

Map 1 Location of the study site Cañuma, in the Apolobamba mountains of the High Andes,
northwest Bolivia

4 Results

4.1 Conditions of the Glaciers from the Andean Mountain

Range of Bolivia

Tropical mountain glaciers are known for their high sensitivity to climate change,
especially rising temperatures. In recent years, climate change has caused the
freezing point to rise. This aspect coincides with the warming of the Andean high
troposphere [11] and leads to reduced albedo and greater surface absorption of solar
radiation. Unlike mid-latitude mountain ranges, as occurs in the Alps, ablation
balances accumulation in the Andes, preventing the development of long-lasting
70 M. Liberman

Table 1 Characteristics of the region where Apolobamba wetlands (bofedales) occur

Biogeographical characteristics and High andean plain in the altiplano and
population of the region mountain range of apolobamba
Geographic location: latitude and longitude High plateau and high mountain area of the
coordinates Apolobamba range: 15° 00′ to 15° 10′ S
latitude; 69° 05′ to 69° 20′ W longitude
Altitude range of the territory 4300 to 5300 m
Climate characteristics: average temperature 4.2° C; 490 mm annual average rainfall.
and annual average precipitation (data of the Strong weather, more than 200 days with frost
Meteorological Station of the locality of per year (4300 m above sea level). Dry and
Ulla Ulla) very cold winters
Wetlands (bofedales) Surface: 105 km2, located in the Ulla Ulla
Altiplano and glacial valleys of the
Apolobamba mountain range
Political administative entity Dept. of La Paz, Bautista Saavedra Province,
Municipalities of Curva and Pelechuco, Franz
Tamayo Province
Protected area Apolobamba. Natural Area of Integrated
Management (ANMI)
Types of vegetation Wetlands (bofedales), lawns, high Andean
grasslands in plains and mountains
Population Aymara indigenous communities organized in
Ayllus (more than 3000 inhabitants)
Economic activity Camelid grazing; wool and meat and fiber
trade of alpacas and llamas; vicuña fiber sale;
trout fishing in lagoons and high Andean rivers
Animal grazing Domesticated camelids: alpacas (grazing in
bofedales) and llamas (grazing in dry
grasslands)
Wild camelid: vicuñas (in all plant formations)
Livestock density Bofedal: 2.7 alpacas/hectare;
Plains: 0.3 alpacas/hectare
Major forage species Distichia muscoides, Oxicloe andina;
Plantago tubulosa, Deyeuxia sp.
Vegetation units with high forage quality Wetlands (bofedales); lawns, grassland
Age of camelid grazing practices (Alpacas Domesticated camelids between 4000 to 3500
and Llamas) BC (*)
Sources Molinillo [28], Liberman and Molinillo [23], Ribera and Liberman [38], Seibert [40],
Meneses et al. [26, 27]; and Wheeler [47]*

seasonal snow cover [45]. Thus, there has been a retreat of the world’s tropical
glaciers over the last decades. In Bolivia, Soruco et al. (2009) estimated that the 367
glaciers of the Cordillera Real had suffered a setback. The average loss was 43% in
volume and 48% of surface area between 1975 and 2006.
In the long term these transformations in glacier volume will cause considerable
changes in the seasonal downstream flow from the glaciers. The most important
changes in flow are expected to occur during the dry season, when glaciers release
Climate Change, Wetland Management and Alpaca Pastoralism … 71

Fig. 1 Profile of vegetation belts on the eastern Andean slope, Apolobamba mountain range. The
community of Cañuma is located in the high Andes, where only alpacas and llamas are bred [38]

meltwater. The snow that falls in the high Andes is stored initially as ice in the
mountain glaciers before it is released after a certain time and adds to the flow
during the dry season [12, 46]. Consequently, glaciers play a very important role as
shock absorbers of seasonal rainfall and provide water during the dry season for
numerous domestic uses. In the case of Cañuma, this flow irrigates the wetlands
(bofedales) that feed camelid cattle (alpacas and llamas).
According to Jordan [21], 80% of the surface of the Andean glaciers has an area
of less than 0.50 km2. The melting of the glaciers began in the early 1980s, and it is
presumed that many of these small glaciers have disappeared in the last 20–25 years.
Some were documented (measured and characterized), but there are no records for
the vast majority [19].
In the long term, glacial retreat will not only impact water for human con-
sumption, but it will also generate stress on pastoral activities in the high moun-
tains, thus affecting alpacas and llamas, because they will not have the necessary
food in the absence of aquifers and bofedales. As stated by Rocha [39], this effect
will occur due to the expected increase in seasonality, which will cause temporary
and spatial decrease of flooded areas, as well as smaller recharge of the aquifers,
where any modification in the hydrology of their basins can significantly alter their
configuration [33]. Disorders in the bofedales are caused not only by the general
drying in the region.
72 M. Liberman

Map 2 Distribution of bofedales, glacier areas and their retreat from 1985 to 2019. Based on
multitemporal analysis of satellite images. Source own elaboration

Map 2 shows the distribution and surfaces of bofedales and glacier areas in the
Apolobamba mountain range since 1985, 2014 and 2019, based on multitemporal
analyses of satellite images, in accordance with the methodology described earlier.
Climate Change, Wetland Management and Alpaca Pastoralism … 73

Photo 1 Ullu Kalli glacier in the northeastern part of the Cañuma community, Apolobamba range

The following map of glaciers shows that, from 1985 to date, their surfaces have
undergone significant reductions. Indeed, over a period of 34 years, there has been
a 45.94% reduction in the surface area of the Apolobamba glaciers, which is
affecting the water supply for the bofedales Photo 1.

4.2 Puna Vegetation in the High Andean Region

of Apolobamba

In the puna, the vegetation is subject to high insolation. The germination and
growth of dominant plants, mainly rosettes and cushions, are hampered by extreme
variations in temperature and low air humidity. This led Troll [43] to declare that
the dry sectors of the Central Andes have the greatest microthermal deficit for plant
growth on the planet.
In these cases plants grow very slowly, and many species function only during
the morning and afternoon, between the phases of nocturnal frosts and midday
overheating. The low productivity transforms the plants into radial cushions and
dwarf rosettes, with small leaves and compact surfaces [6, 10]. The resulting
74 M. Liberman

decrease in transpiring surface area provides protection against low temperatures,

wind and high evaporation [36].
The Cañuma area is located in the puna of the Bolivian Andes, where vegetation
zones are distinguished above the forest limit. The following diverse belts were
identified by Beck and García [16] in the puna: nival-subnival, upper and lower
high Andean, and the puna belt per se. The eastern Andes and the Altiplano present
many different climatic and topographic situations, producing several habitats in
which the vegetation is rich in species and life forms. In this study, we will con-
centrate on the so-called wet bofedales (meadows) that exist in the sub-snowy and
upper Andean belts from 3800 to 5000 m above sea level.
The high Andean vegetation covers mountain slopes up to 5200 m, where the
nival belt and snow cover begin. According to Navarro and Maldonado [31], the
altitudinal ranges for this ecological “floor” are restricted to 3900–4600 m, while
the lower limit of the subnival belt occurs between 4600 and 5200 m. An important
characteristic that differentiates the sub-Andean belt from the high Andean belt is
that the higher presents nightly frosts throughout the year. Instead, below 4600 m,
frosts decrease during the summer months (between one to three months). This area
occupies the high plateaus, as well as glacial valleys, with rocky substrates. The
predominant growth forms are dense bunch grasses, cushions and rosettes. Near the
upper limit, isolated shrubs of Senecio rufescens, Agrostis tolucensis grasses,
Dielsiochloa floribunda, and perennial herbs (Senecio serratifolius) grow on rocks
that accumulate heat during the day and attenuate temperature drops at night [2, 3,
13–16].
In depressions near lagoons or other water sources, the vegetation of wet swards
is characterized by the presence of Plantago tubulosa, Azorella biloba, Gentiana
sedifolia, Oritrophium limnophilum and Mexican Cotula. Between 4300 and
4600 m, on semi-arid soils, a low sward develops. It is composed of isolated
communities of grasses, evergreen dwarf shrubs and perennial herbs in rosettes or
cushions. The height of the plants does not exceed 5 cm, and coverage is estimated
between 5 and 30% [3]. These authors estimate that poorly drained areas promote
the formation of compact herbs and hard cushions that form the bofedales of the
puna. Bofedales are vegetation formations with permanent surface water throughout
the year. In the area of Apolobamba they cover an altitudinal of 4000 to 5000 m
and characterize the high Andean landscape and the puna. The bofedales of the
puna have cushions and swards of Plantago rigida, Werneria pigmaea and Lucilia
tunariensis. These bofedales develop on climatically humid slopes, around ponds,
lagoons, streams and outcrops of water or springs, in the eastern part of the
mountain range. Part of the high Andean vegetation includes bofedales of Distichia
filamentosa and Distichia muscoides (Photos 2 and 3). These are located mainly
around 5000 m but not below 4200 m [40]. These bofedales are different from
those in the puna, which are usually found in depressions, underground water
springs, on moist soil flooded by running or stagnant water, in which they form
hard cushions (García et al. 1995). Oxycloe pads are found in contact with
Distichia, but form separate structures, never overlapping [40].
Climate Change, Wetland Management and Alpaca Pastoralism … 75

Photo 2 Glacial valley in Cañuma, Cordillera de Apolobamba, with wetlands (foreground) and
the glaciers of Ullakalli (background)

Photo 3 Wetland (bofedal) of Distichia muscoides surrounded by water courses and pools
76 M. Liberman

Bofedales typically represent azonal formations. They form a mosaic of elevated

and hard cushions, patches of dense or open sward, with small patches of grasses in
bushes on the edge, where numerous small wells, lagoons and streams appear [10].
Their largest extent is observed at the bottom of U-shaped glacial valleys. There are
also hillside bofedales, just where water sources appear on the slopes [23, 28]. The
determining factor for the formation of bofedales is the surface and underground
water that runs permanently, that gives life to these ecosystems (Photo 3). Steep
stony places are covered by Stipa brachyphylla, with the presence of isolated
Opuntia flocossus. Areas with stagnant water have communities of Plantago
tubulosa, which belongs to the community of Aciachne pulvinata. In open dry
areas, communities of Pynophyllum-Selaginella can be seen, as well as the steppe
of Deyeuxia curvula.
Peasant herders have learned to domesticate plants by converting them into
useful herbs for a number of purposes. Basic relevant, important information has
been generated for the sustainable use and management of biological resources in
the high Andes of Bolivia, at altitudes over 4000 m. Biological resources are crucial
for the economic development of peasants and indigenous people. These people
often depend on plants for their livestock food and to cover their own daily needs.
Peasants suffer more than other groups in society when the environment is degraded
by overgrazing their pastures, by climate change, or by other associated processes.
In the High Andes, most plants and their interactions with the environment are
poorly understood. In many cases the real or potential economic value has not yet
been evaluated.
In the high-mountain community of Cañuma, Distichia muscoides became
useful. It is a typical wet-meadow plant that was turned into a useful plant not only
as a forage source for camelids (Photo 4), but also as a utilitarian means for
the transport of water in the mountains to generate new grazing fields for alpacas
and llamas.
Traditional knowledge of peasant management of species such as Distichia
muscoides, locally called “Khuli Female”, is based on a novel local experience
closely connected to the natural environment where it has been developed. If
preserved and disseminated, it can be widely useful in the design of sustainable
management schemes in other communities. It is necessary to value these examples
to promote a type of development that can take advantage of the knowledge that
was surely implemented over decades of testing by families living in the highly
constraining high Andean ecosystem.

4.3 Alpaca and Llama Grazing Technologies

Flores Ochoa [7, 9] describes herding camelids in the high Andes, above 4500 m,
as an economic system very dependent on environmental conditions. It is limited to
exploiting almost parasitically the resources of nature without making significant
modifications of the natural foundation.
Climate Change, Wetland Management and Alpaca Pastoralism … 77

Photo 4 Alpaca grazing in a bofedal of Distichia muscoides in the community of Cañuma

The presence of glaciers and numerous watersheds in the Apolobamba range

ensures a permanent water supply for the bofedales. This situation has favored and
partly kept the region relieved of forage deficit, even in times of great, painful
droughts [32, 37]. In 2015–2016, which included an “El Niño” event, 10% of
camelid cattle would have died from extreme climatic effects (drought, frost, hail
and snowfall), according to Cañuma residents.
In the Apolobamba mountains there are several communities where pastoralists
have transformed grassland and bofedal ecosystems through innovative irrigation
systems with canals and water reservoirs (Q’ochas), to expand the areas of intensive
grazing, thus creating excellent grasslands that improve grazing environments for
both alpacas and llamas. Grazing is not only a form of exploitation, adapting to, and
making the most of resources in a high Andean ecosystem. It is also a complex,
imaginative world where various management models and uses of resources are
established and proposed.
For the inhabitants in the region, there is no alternative to the work of grazing the
camelids. They are the support for the human populations, which benefit from their
fiber, skins, fresh and dried meat, or trade them in valleys or use them as exchange
goods to acquire food of agricultural origin or other products. Another local product
may be fertilizers, used for grassland regeneration and fuel. Other products are fat
for the manufacture of medicinal ointments, and bones, from which combs and
78 M. Liberman

handicraft instruments are produced. Leather is also transformed into material for
roofing homes.
Physiologically, camelids take better advantage of the high Andean pastures than
do the species introduced from Europe by the Spaniards. Alpacas, for example,
digest the pasture grasses 22% more efficiently than do sheep [5]. Ecologically,
camelids, by their way of grazing, do not degrade the pastures as much as do cattle,
ungulates and horses. The camelids cut the grass, while the others uproot it. Cattle
and horses can be extremely harmful in drier areas, where the vegetation cover
becomes impoverished and the soil degraded. Trampling by sheep, due to the
structure of their hooves, is equally damaging in the high Andean pastures;
camelids, which have plant like pads, do not damage the vegetation cover.
The main characteristics of the herding society in Cañuma are as follows. As in
much of the rural area of Bolivia, family and community are the pillars of social
organization, which determine the daily lives of people and the modes of evaluation
and management of territorial resources. Albó [1] suggests that a community is a
specific group of families that share a defined territory with their own system of
government. It is also an institution that controls space. A community is inseparable
from the space that makes the territory, which is delimited and regulated and by the
group of people which is the institution. In the case of Cañuma, the community
controls the territorial resources (water, grasslands, bofedales) that are common
property, while community members are only beneficiaries. As for family-owned
land, the community guarantees the security of the rights of use. With regard to
grazing activities, it operates and assigns restrictions in some areas. The community
as an entity is sovereign, designates its authorities and dictates the external rules
[44].
Families are responsible for grazing camelids in places where they have tradi-
tionally taken them. There are traditional sectors of intensive grazing in bofedales
where up to six herder families can graze their alpacas. This activity is mostly done
by women and their young children. Males who have not migrated to large cities or
to the mines may graze cattle, but also maintain the bofedal irrigation canals. In
dryland grasslands grazing is free.
Among the different families there is the system of exchanges, assistance or
mutual help and cooperation between individuals and families for the care of
camelid cattle. Cooperation, in fact, is one of the foundations of the High Andean
culture, where there is an exchange of favors. These may include, for example,
requesting the care and monitoring of alpacas when they graze, or communal
solidarity work and internal reciprocity in the maintenance of irrigation canals,
q’ochas (lagoons) and shearing.
There is a process of seasonal transhumance of camelid cattle within the limits of
a community, where herders carry out migrations or changes of lodging so that the
animals have adequate grasslands in each season of the year (dry and rainy). This
means that the inhabitants of Cañuma have homes in different places, usually at
least three homes (one main and two temporary located in other areas of the
Cañuma). Taking advantage of the pastures corresponds to a sectorial fallow system
Climate Change, Wetland Management and Alpaca Pastoralism … 79

within a limited area, without carrying out large-scale or large-distance transfers.

This implies seasonal and spatial rotation.
The technique of handling camelid cattle as domestic animals is one of the great
achievements of Andean cultures. The oldest camelids were apparently not adapted
to great heights. Wheeler [47] affirms that domestication of alpaca and llama took
place in the Central Puna of Peru. Archaeological research reveals that this occurred
between 4000 and 3500 years ago and began the early development of camelid
livestock in the Andean region. As a result of their knowledge of the animals, care
of the herds and other techniques like grazing habits, and the rotation of grassland
areas, herders obtain good yields of Alpaca wool and can maintain and irrigate the
bofedales. Surveys conducted in the area indicate that each family may have
between 100 and 150 alpacas and about 6 to 10 llamas.
Llamas are used as beasts of burden and for meat. The distribution areas of
camelids show that they have specialized in living at great heights, as a conse-
quence of the dependency initiated in the sixteenth century by the domination,
pressure and marginalization exerted on Aymara populations. The association of
shepherds with camelids has brought several benefits, such as pasture supply,
especially in times of cold or environmental degradation; care of weak animals or
neonates; and protection against natural predators. This symbiosis between society
and animals has transformed the landscape for the well-being of both populations
[18]. The relationship between man and animals develops in such a way that llamas
and alpacas are referred as “people of the family.” Herders in Cañuma give names
to their animals, and both animals and people recognize themselves. There is a
permanent concern in the care and breeding of animals, as if they were their own
children.
Herders take care of the animals permanently, so that none get lost in the
mountains or on roads. It has been observed in Cañuma that there is a special
language between the herder and his alpacas, allowing the recognition of alert
sounds in the face of a threat or imminent risk. On one occasion it was observed
that an alpaca was lost and had moved a considerable distance away, and the herder
was desperately searching for the animal, giving sharp, loud screams in amid the
hills and mountain slopes. The entire cattle herd also felt nervous, until finally the
alpaca heard the screams and came in a hurry to join the rest of the cattle. After this
event, we observed tranquility in the group of alpacas. The herder smiled as she met
her lost alpaca again.
Every day herders plan the grazing activities of their alpacas and llamas, early in
the morning before the sun rises, selecting the places where they will take their
animals. They know the places that have been previously grazed and the sites at
different heights that are reserved for the dry or wet season. Before going out,
herders wait for the sun to rise and melt the frost or snow, since wet or frozen grass
produces diarrhea in the animals.
Alpacas are more territorial, so they avoid long walks. They prefer places with
the highest humidity or bofedales that have green, hard, fresh pastures. In most
cases, these are located close to homes that are located in relatively restricted,
contiguous spaces. Llamas, on the other hand, are more resistant, walk long
80 M. Liberman

distances through the steep hills, and prefer places where dry land with yellow,
hard, rough or thorny pastures. They feed themselves on their itinerant way, moving
away from the place where they started their walk. Each family has developed
different grazing strategies according to the season and site, making the most of
what nature offers. Every day herders move to different grazing sites, rotating the
feeding sites, and thus taking advantage of the diversity of forage and production
seasons in an effective way. In many cases, the bofedales are created by the same
herders over many years, building irrigation canals that transport water for the
irrigation of grasslands, in such a way that they have increased the grazing area and
also the areas of water reservoirs.
The animals are slaughtered in the beginning of the dry season, when the cattle
have good weight and there are adequate climatic conditions (solar radiation and
warmth). Camelid meat is used to produce salted dried meat (charqui), which has
good profitability, occupies less space and has a longer storage life. Charqui and
wool, which are surplus products of the herding work, are transported by the llamas
to the valleys of Charazani, where herders exchange food products (corn, coca,
potato, goose, etc.) with farmers from the lowlands. This work is done mainly by
men, who guide the llama caravans. The travelers, merchants or llama herders depart
from Cañuma (4500 m) to the hot valleys of Chari, Lunlaya, Charazani and other
sites that are located at an average height of 3600 m. They move to the eastern slope
of the Andes to exchange their products within the different ecological areas, with
the sense of control of economic verticality and using different ecological spaces that
are exploited interdependently. This theory explains the socio-environmental man-
agement model of the Andean space [29, 30].
Alpaca breeding programs are mainly aimed at gaining and producing fiber,
which has a high commercial value. For this, alpacas are sheared between 5 and 8
times in a life time. The production of fabrics from alpaca and llama fiber is an
artisanal, manual activity carried out by women while grazing their camelid cattle.
Through this they produce extraordinary fabrics that express their culture that
comes from pre-Hispanic periods and that show high levels of quality. They shear
the animals, spin the fiber and weave beautiful clothes such as: dresses, pants,
ponchos, blankets, winches for the head, chullos (hats), sashes, socks, scarves,
blankets, etc. All the garments made on looms have abundant, geometric, symbolic
decoration, both natural and with a great variety of colors made from dyes provided
by native plants of the puna and from the tropics (Quisbert et al. 1984). The most
common representations are alpacas, llamas, birds, flowers and daily grazing
activities, which express extraordinary quality in elaborate fabrics. Teaching girls
how to produce woolen fabrics starts at age 12. It is very common to see mothers
and daughters with their rustic horizontal looms weaving together, while herding
the alpacas.
A characteristic of this peasant unit is the minimization of risk. That is, in
response to the theory of limited goods (scarcity), one could say that pastoralists act
rationally according to their circumstances and levels of information. In this sense
camelid cattle represent a safe source of savings and investment, both for their
convertibility in times of need into meat or money, and for the increase in their
Climate Change, Wetland Management and Alpaca Pastoralism … 81

value over time. This value is not only based on work; animals take advantage of
resources that are not accessible to humans. They also contribute with their manure,
for example, to increase the fertility of bofedales or as fuel (tachy) for cooking [41].

4.4 Management of Bofedales and Climate-Change

Adaptation

Bofedales are high Andean plant formations of great importance for camelids, their
daily activities and their life cycles, because they are food source for alpacas. They
are azonal formations established in an organic environment characterized by
permanent water saturation. They are heterogeneous depending on the physiogra-
phy of the site, as well as the altitude above sea level. In general, they are domi-
nated by the presence of the female Khuli (Distichia muscoides) and the male Khuli
(Oxicloe andina) that form a set of hard, compact cushions (Distichia) and lax,
spiny, hard formations (Oxicloe). They make up a wavy micro-relief that highlights
the presence of small water reservoirs among the Distichia plants. Also common are
Plantago tubulosa, Eleocharis albibracteata, Carex spp. and Werneria pymaea
(Alzerreca et al. 2001, Castro et al., 2014, Coronel et al. 2009). Depending on the
water gradient, these species may be floating, emerged or submerged, up to the edge
of the zonal grassland vegetation [4].
At these sites, plants take advantage easily of the nutrients from the decompo-
sition of organic matter at shallow depth. The bofedales have a great variety of
aquatic plants, some growing their roots in the water and the leaves on the surface.
Prieto et al. [35] report variable of yields of fodder phytomass in bofedales from
Ulla and Huacuchani, ranging from 854.7 to 2,638.8 kg dry matter/ha. These sites
are very close to Cañuma. In Cañuma, bofedal plants form compact cushions
arranged in many layers of dead plants that can be several meters deep (Dangles
et al. 2009). The source of water comes from the glacier and snows of the Ullu
Kalli, as they melt.
The vision of the alpaca herder families is a deep relationship between the land
(Pachamama), the pastures (Jach’u), the alpacas (Allpachus), the water (Uma), the
mountain deities or spirits (Apus) and the herders (pacocheros) [7–9]. One cannot
value one element separately from the others. They are all interrelated, forming a
global total unit.
In general, during the rainy-season grazing period, herders take the alpacas to
areas where they have dry green grass pastures. These are dry, hard, and of low
productivity during the winter season (Palacios 1977) [34]. In winter the alpacas are
taken to bofedales where plants are green. During this period the herders carry out
activities for diverting water from the Ullu kalli glacier to irrigate the bofedales that
have better pastures. Due to its location adjacent to the Ullu Khalli glacier, the
Cañuma valley has a permanent water supply. This is an advantage for grazing, in
contrast to areas where water is scarce during most of the year.
82 M. Liberman

In the region, natural bofedales are located mainly in areas that have a flat or
almost flat topography, at the bottoms of the glacial U-valleys. There are also slope
bofedales where aquifers or springs appear. They are also called ojos de agua
(water springs) and are permanent. Where water flows from a certain point on a
slope, a bofedal develops and opens out downslope like a fan.
As stated earlier, many bofedales are artificial. In many places the wetland area
was enlarged to improve pastures that provide better food for camelids. The con-
struction of these bofedales is a response and an adaptation to the climate change
that has developed in recent years. Herders have a set of climate-change adaptation
strategies to improve grasslands and their forage yields, which ultimately determine
better conditions for alpaca livestock [24].
The strategies that have been created and developed with imagination and skill
by alpaca herders have managed to create ingenious “bioengineering works” and
have generated creative solutions for the transfer and transport of meltwater through
channels that cross steep rocky slopes. Thus, they have been able to create huge
areas of bofedales for the benefit of their alpacas and llamas. All these works
constitute a true legacy to confront climate change. It is true that the Cañuma Valley
herders have historical information on the flows and availability of water
throughout the Ullu Kalli glacier area, throughout the year. Their experience plays
an important role because it has allowed them to manage data, based on knowledge
of situations at other, similar sites. These herders manage climate risk and minimize
losses to their camelid cattle.
For the construction of infrastructure works, labor is required in places where
there are no access roads. When an irrigation channel for the bofedales has to be
built, families ask their members to work within conceptions of reciprocity of
Andean work, known as Ayni. In the Cañuma area, three types of bofedal irrigation
canals are used to extend the borders of wetlands, to maintain their large camelid
herds:
a) rustic canals of Khuli (Distichia muscoides);
b) rustic canals combined with q’ochas;
c) rustic canals in high Andean grasslands.
In these cases, shepherds have developed irrigation techniques to improve the
feeding of alpacas. According to Palacios (1977), alpacas that are bred in a bofedal
with enough soil moisture can provide 5–6 kg of wool in two years, while in dry
places they give only 1.5 to 2 kg of wool during the same period. However, this
type of wool is usually as dry as of that of llamas, which is why shepherds value the
bofedales. In general a bofedal can support about 3 alpacas per hectare per year.
The construction and maintenance methodologies of the irrigation canals in the
high mountains of Cañuma are described below:
Climate Change, Wetland Management and Alpaca Pastoralism … 83

4.4.1 Rustic Canals of Female Khuli (Distichia Muscoides)

In the highest part of Cañuma at about 4850 m, a set of channels was identified that
conduct the melting waters of the Ullu Kalli glacier through main ditches of very
slight slope along the hillsides of the glacial side moraines. From the main ditches,
at certain intervals, there are breaks where the water flows to the lower slopes to
irrigate the hillside bofedales. In this way herders increase the area of bofedal
intensive grazing sites on the slopes of the glacial valley. Periodically, herders carry
out their maintenance activities.
The interesting novelty of this irrigation system is that the walls of artificial
canals are constructed using tepes, which are compact cushions of Distichia mus-
coides (female Khuli), i.e. the same species that forms the wetlands of the valley
bottoms. This is represents extraordinary ecological knowledge, adaptation, and
technological innovation to construct the hillside channels taking advantage of the
cushion life form. Each cushion plant is moved at the same time to the site where a
channel is formed to take water to other sites and the irrigation frontier of the
grazing areas is extended. In this way the negative effects of climate change in the
community are minimized. This is an example of ingenuity and work by a family of
herders to increase the productivity of the vegetation and consequently the greater
yield of their camelid herd. One family member affirms that his family took several
years of work and intense observations of the plant, involving many difficulties for
his parents and grandparents.
All the members of the family of alpaca herders from this area handle the care of
their bofedales and their ditches skillfully, and especially the grazing trade of their
alpacas. This labor and knowledge are transmitted from one generation to the next
through methodical observation since childhood. This helps to perpetuate herder
occupation within families. In addition, this activity is very gratifying spiritually,
for herders experience special moments in the mountain and with their animals.
During childhood these activities mark children’s personalities, which usually bring
two clear consequences. The first is that a child may reject the activity and will not
want to know anything more about it. This possibility is not uncommon, because
more and more children want to study and be “more than their parents.” The second
consequence is that the child loves herding and might never wish to separate from
the camelid herds. Rustic canals were identified on the southern slope of the glacier
valley where the Laguna Verde is located. In this area, there is a large valley-bottom
bofedal that is irrigated with waters of the Ullu Kalli glacier, located at an elevation
from 4690 to 4742 m (52 m difference; Fig. 2). The female Khuli canals were
initially conceived for the provision of water for housing. Herders living at the site
suffered fatigue moving water from the lake to their dwellings at the top of the
hillside.
Building the channel involved several topographic, hydraulic and construction
challenges. Its builders also had to study the Khuli plant’s ecology, physiology and
lifestyle. The channel that was designed moves water from a slope bofedal on the
upper part of the glacial valley. From the bofedal, water intake is achieved through
an excavation.
84 M. Liberman

Fig. 2 Satellite image showing the Khuli canal (Distichia muscoides). The image reveals how the
area of bofedales has increased on the slope of the glacial valley

The water is conducted 1231 m through a firm rustic canal, the walls of which
are lined with cushions of the same Distichia muscoides (Khuli) plant. At the same
time, this guarantees that there is no water shortage, for the survival of the plant
depends on permanent flooding. If it lacks water, the plant inevitably dries
up. According to the herders, building this channel took several years of work by a
father and son. The channel passes along the upper part of the hillside, which in
several areas has a rocky substrate where excavation work was not possible.
For construction of the canal two parallel tasks had to be carried out: con-
struction of the canal itself and permanent provision of water to the bofedales,
especially during the dry season. At certain intervals there are points of water
distribution and regulation, where some holes were made. From these points small
secondary channels are created, allowing flooding to occur on the lower slope. In
this way, the surface area of the slope bofedal is increased. When water is not
provided to the hillside, the holes are covered with stones or champas (lumps of soil
with grasses). When a secondary channel is opened, the main channel is blocked, so
that the entire flow waters a portion of the slope bofedal. When irrigation of that
portion of the bofedal is finished, the secondary channel is blocked and water is
allowed to pass through the main channel.
From the water intake, water runs through a channel with a very smooth average
slope of 4.22%, following the surface contours. When the channel passes through a
rocky area, it is reinforced with large cushions of Khuli, grass clusters and rock,
Climate Change, Wetland Management and Alpaca Pastoralism … 85

Photo 5 A slope canal built with Khuli tepes (Distichia muscoides) for creating artificial
bofedales on hillsides

which support the hillside canal as real water retaining walls. The main channel
decreases progressively in size towards the furthest point in the system (Photos 5, 6,
7 and 8).
To make sure that water does not drag along the plants and substrate, herders
must clean and repair the channels continuously, especially after rainy-season
landslides or obstructions due to excessive growth of Khuli cushions that block
water flow. In general, the water intake in the main channel is maintained during
August, in accordance with circumstances. Channel maintenance activities include:
cleaning out plants that obstruct water flow, arrangement of stones, erosion control,
and opening of parallel or secondary channels.
The result of these processes is that, for many years, the hydraulic and irrigation
works on the unstable mountain slopes function properly, because they are adapted
to and compatible with the environmental conditions as well as the maintenance and
capacity requirements of users. Therefore, a sustainable system has been established
with a continuous flow of water for the bofedales. The area of artificial bofedales in
this case reaches 9.25 ha, which represents an outstanding increase in the area of the
bofedal that provides nourishment to 27 additional alpacas for the herder.
86 M. Liberman

Photo 6 Channel built with Khuli tepes (Distichia muscoides), with secondary channels for
watering slope bofedales enabled by moving the rocks

4.4.2 Rustic Canals Combined with Q’ochas (Lagoons)

This type of rustic canal infrastructure, combined with q’ocha lagoons, is used in
the middle part of the Cañuma Valley (4643–4633 m), in an area where there is no
dominance of Khuli plants (Distichia muscoides) but where herders have built
irrigation channels with a total length of 1040 m.
A major characteristic of this adaptation is that canals run along the slopes of a
series of hills, basal moraines of glacial origin, which in turn are located in the
enormous glacial valley of Cañuma. In this place herders designed their canal to
Climate Change, Wetland Management and Alpaca Pastoralism … 87

Photo 7 Local bioengineering work showing the path of the irrigation canal with Khuli tepes on a
rocky hillside

begin in a small lagoon that is the overflow of the Laguna Verde (Green Lagoon),
which is nourished with the melting waters of the Ullu Kalli glacier. This lagoon
has a permanent water supply. In this case the rustic canals were excavated on the
high mountainsides and have a gentle slope. The walls of the main canal are
composed of stones and champas (grasses with lumps of soil). The canals include
points of water distribution that irrigate the pasture slopes (not the same as in
bofedales). The flow of irrigation water is so well calculated that continuous water
currents on hillsides do not generate serious soil erosion. In other areas, irrigation is
accomplished through overflow of the main channel. Where the canals have sharp
curves, it is interesting how herders design them with a wider radius and more
robust walls, to avoid collapse when there is too much water. Slope adjustments
were made such that the water circulates at more slowly, avoiding useless loss and
at the same time avoiding deposition of solids in the canal.
Another issue that should be highlighted is that this type of canal has improved a
variety of special works that solve technical problems for better water conduction,
such as waterfalls, regulation ponds and obstacle passes. All these works were
undertaken with ingenuity, in an environment with rugged topography, difficult
excavations, and long canals bordering hillsides. The canals are well designed in
such a way that they ensure infrastructure that fulfills its functions over time. In this
system we observed the presence of about three levels of secondary canals located
below and parallel to the main canal. These canals receive water from the main
88 M. Liberman

Fig. 3 Route and altitudinal variation of a rustic canal combined with q’ochas for grassland
irrigation

canal by overflow, and retain the water in order to provide a better water supply for
the hillside pastures.
In the middle part of the canal route, where hillside grasslands are irrigated, the
water arrives at a depression where herders accumulate it to form lagoons. Marsh
vegetation similar to parsley grows here and is craved very much by alpacas.
Animals graze on this area have learnt to submerge their heads in the water to feed
on the herbs that grow in the lagoon. Such lagoons are called q’ochas and are water
reservoirs, like ponds, that regulate the flow of the irrigation canal. This irrigation
system, involving canals and q’ochas, finishes at last part of the segment, where it
reaches an area of reserve pastures or an intensive grazing site, which is fenced with
wire. Here, herders gather gestating alpacas or those weak from lack of fodder,
when extreme droughts occur (Fig. 3 and Photos 7, 8, 9 and 10). In this sector an
additional area of about 13.8 ha was increased, providing grazing for 41 additional
alpacas.

4.4.3 Rustic Canals for Irrigation of High Andean Grasslands

Rustic canals in areas of high Andean grasslands are located in the lower part of the
Cañuma glacier valley, between 4493 an 4499 m, and with a length of 1602 m.
Climate Change, Wetland Management and Alpaca Pastoralism … 89

Photo 8 Start of the irrigation canal that transports water to grazing sites and small lagoons

Photo 9 Irrigation of slope bofedales and grasslands increases the grazing area of camelids
90 M. Liberman

Photo 10 Along the canals there are several lagoons serving as water reservoirs for alpacas and
llamas

(Fig. 4) The area is almost flat. Water in it is captured from the Cañuma River,
which runs through the plain and has an abundant flow throughout the year. One
obstacle in the area is that the Cañuma River eroded the base of the river bed, to
more than 3 m. For this reason there is insufficient water within the irrigation sector.
To solve this problem, herders capture water in areas where surplus water is
available, almost at ground level. From there they built canals with a minimum
slope of 0.35%. In this case, the herders increased the grassland by 7.4 ha, which
can support an additional 22.2 alpacas.
The system consists of multi-flow irrigation canals between mono-flow sectors.
High-Andean grasslands in Bolivia normally receive seasonal rainfall between
November and March, so they do not require additional water (Photos 11, 12, 13
and 14). But when there is an extreme drought, like the one that occurred in 2015–
2016, the need to build irrigation infrastructure to mitigate drought effects began.
The infrastructure consists practically of canals excavated in clay soils to conduct
water along a minimal slope towards dry grasslands.
Climate Change, Wetland Management and Alpaca Pastoralism … 91

Fig. 4 Route and altitudinal variation of the rustic canals for irrigation of high Andean grasslands

Photo 11 Irrigation canals that pass through bofedales

92 M. Liberman

Photo 12 Detail of the main channel that derives its water to a secondary channel for irrigation of
bofedales

Photo 13 Aymara herder children in the lower part of the community

Climate Change, Wetland Management and Alpaca Pastoralism … 93

Photo 14 Cañuma herder

enabling a grassland irrigation
canal

When a herder decides to irrigate his grazing sector, he places stones or a tepe of
the pastures from the area, and in this way generates flooding. Once the grassland
has been watered, the obstacles are removed so that the water flows to other places
that need to be irrigated.

4.5 Perceptions of Climate-Change Effects, from Alpaca

and Llama Herders in Cañuma

Table 2 shows the results of the participatory workshop between technicians from
the Universidad Mayor de San Andrés, La Paz (Bolivia) and members of the high
Andean community of Cañuma, in the Apolobamba mountain range. In that
workshop, alpaca herders presented their perceptions of climate-change effects on
activities related to pastoralism and the measures that are being implemented to
minimize negative effects. In that workshop, so-called talking maps were used,
where farmers wrote and drew the effects of climate change.
94 M. Liberman

Table 2 Perceptions of alpaca and llama peasant herders on climate change

Factors Perceptions of peasant herders from cañuma on climate change
Rainfall • In recent years there is little rain that affects bofedales. Now there are no
pastures, and the alpacas are becoming thin and get sick easily
• With droughts there is less water and fewer bofedales
• The river and water springs have dried up now
• In the past there were many lagoons and puddles that are now dry
• Before there was more water, now there is little water
• With the drought, worms that eat plants appear
• Now there are changes in the rainy season
• The rains before fell from September until April
• Today the river has little water, downstream the river has dried and frozen
Temperature • Now the weather is warmer and the glacier is melting
• The sun is burning very strong
Hailstorms • In recent years hailstorms have become strong
• Hail comes that kills animals
Frosts • Frosts are now more frequent
• Frosts burn pastures
• Water freezes
• A strong frost affects alpacas and llamas
• Abortions occur in our animals
Winds • Now there are more strong winds that affect the eyes of alpacas
Snowfalls • Heavy snowfalls occur
• In 2015 there was a heavy 50 cm snowfall
• The snow lasted a month and a half. It was the death of alpacas, llamas and
vicuñas. We all cried for our animals
Alpacas • Alpacas become thin due to lack of pastures. Frost affects grasses. There is
much animal mortality due to lack of water
• Alpacas get sick easily. They have many external and internal parasites
(Chaymalako and Tenia). They also get fever
• Alpacas get conjunctivitis due to strong winds
• Young alpacas and llamas are attacked by foxes, pumas and condors
• Worms pests emerge and feed themselves on leaves in the bofedales
• In 1998 a hare appeared that eats bofedales. The food of the alpacas is eaten
People • There is migration of peasant families to the cities and gold mines
• People have no profits (in the field) due to lack of resources
• Smoke from burnings affect people and children
• The river freezes and dried up downstream. No water reaches the cabin
• Neighbouring communities suffered from drought

5 Conclusions

• All camelid herders in the high mountains agree that climate conditions in the
region have changed, and recognize that the weather has changed. There is an
irregular distribution of rainfall, drought, hailstorms, snowfall, frost, more wind,
high evaporation by solar irradiation, high temperatures and ultraviolet
radiation.
Climate Change, Wetland Management and Alpaca Pastoralism … 95

• In order to mitigate the retreat of glaciers and the absence of rainfall, alpacas and
llama herders have identified strategies to reduce vulnerabilities and to increase
the resilience of the ecosystems of the alpaca herder community. The adaptation
measures developed are intended to ensure the provision of water for the
grasslands by increasing the surface area of hillside bofedales and the con-
duction of water through canals. In this way they ensure food security and
improve responsiveness to extreme effects.
• Traditional wetland management technologies with innovative irrigation sys-
tems are considered to be forms of wetland expansion, as well as ways of
organization for water management in the Cañuma area. Implementing these
technological innovations for irrigation make the peasants proud, for they value
their identity as alpaca herders in ecosystems of extreme conditions. These
peasants also represent the legacy of herders who have created climate-change
adaptation technologies. These should be rescued and transferred to other
communities in the high Andean mountains.
• New rustic canals, based on innovation and their own technologies to increase of
wetland areas, are constructed by herders through gravity irrigation systems to
irrigate the wetlands. These must be improved with social innovations that allow
the diffusion of ancestral knowledge, complemented with modern technologies.
• It is necessary to monitor the meteorological phenomena of the Cañuma station,
together with the ethno-climatic knowledge through observers. This helps
provide an early warning system of climate phenomena that may be occurring. It
is necessary to have reports of climatological indicators for the results of
technical indicators of alpaca breeding and management of bofedales, under
conditions of: frost, hailstorms, rains and droughts; against mortality, morbidity,
birth, fertility, management of color diversity in alpacas, rotation of grazing
areas and use of resting fences, sanitary management of camelid cattle, fertil-
ization of natural pastures and fences with acquired materials.
• All these climate factors must be acknowledged and investigated more deeply.
This will enable the construction of adaptive processes that respond to the local
deficiencies of risk management of alpaca livestock activity, as well as the
acquisition of solid instruments to minimize the effects of current and future
risks, within the climate-change framework.

References

1. Albo X (1994) Instituciones y alternativas organizativas para la producción. In: Albo X,

Ramón G (eds) Comunidades andinas desde adentro. Dinámicas organizativas y asistencia
técnica. CECI/Abya – Yala, Quito, pp 91–135
2. Anthelme F, Beck S, Ginot P, García Lino MC, Meneses RI (2015) Ecosistemas y plantas
altoandinos de la Cordillera Real; únicos y frágiles. In: Meneses RI, Beck S, Anthelme F (eds) La
Cordillera Real y sus Plantas. IRD & Herbario Nacional de Bolivia, La Paz, pp 17–30
96 M. Liberman

3. Beck S, García E, Zenteno F (2002) Flora del Parque Nacional y Area Natural de Manejo
Integrado Madidi. Proyecto de Apoyo al Parque Nacional Madidi. SERNAP-CARE-HNB-
CE-WCS, La Paz, Bolivia
4. Castro L (2007) Humedales de la Puna del Norte de Chile. In: Lucic MC, Reyes LF (eds) Gestión
Sostenible de Humedales. CYTED, Programa Internacional de Interculturalidad, pp 15-46
5. Comisión de medio ambiente y desarrollo de América Latina y el Caribe (1996) Amanecer en
los Andes. CAF, BID PNUD, Caracas, 177 p
6. Em P, Moreira-Muñoz A, Richter M (2006) El Clima y sus efectos sobre la vegetación. In:
Moraes RM, Ollgaard B, Kvist LP, Borchsenius F, Balslev H (eds) Botánica Económica de
los Andes Centrales. Universidad Mayor de San Andrés, La Paz, pp 11–36
7. Flores Ochoa J (1968) Los pastores de Paratía. Una introducción a su estudio. Serie
antropología social, vol 10. Instituto indigenista interamericano, México
8. Flores Ochoa J (compiler) (1977) Pastores de puna. Uywamichiq Punarunakuna. Instituto de
estudios peruanos. Estudios de la sociedad rural 15, Perú
9. Flores Ochoa J (1997) Pastores de alpacas de los Andes. In: Flores Ochoa J (compiler)
Pastores de la Puna. Instituto de Estudios Peruanos, pp 14–52
10. Flores Y, Villagómez JR, Blanco NM, Salcedo L, Sterner O, Almanza GR (2004) Defense
mechanisms of Bolivia highlands plants. Revista Boliviana De Química 21(1):36–41
11. Francou B, Vuille M, Wagnon P, Mendoza J, Sicart JE (2003) Tropical climate chage
recorded by glacier in the Andes during the last decades of the of twentieth century:
Chacaltaya, Bolivia, 168S. J Geophys Res 108(D5):4154
12. Francou B (2013) El rápido retroceso de los glaciares en los Andes Tropicales: Un desafío
para el estudio de la dinámica de los ecosistemas de alta montaña. Ecología En Bolivia 48
(2):69–71
13. García E., Valenzuela E, López R (1985) Vegetación y Flora. In: CECI/centro de Estudios
Canadienses y Cooperación Internacional. Plan de Manejo Reserva Nacional de Fauna
Andina Ulla Ulla. Informe técnico MDSP (Ministerio de Desarrollo Sostenible y
Palnificación). DNCB (Dirección Nacional de Conservación de la Biodiversidad. La Paz,
Bolivia
14. García E, Beck S, Zenteno F (2002) Plan de Manejo del Parque Nacional y Área Natural de
Manejo Integrado Madidi. Departamento de La Paz, Bolivia. Herbario Nacional de Bolivia
15. García E, Zenteno F, Beck S, Nagashiro N (2003) Identificación y caracterización de especies
de uso forestal del Area Natural de Manejo Integrado Nacional Apolobamba. Estudio de usos
potenciales para un manejo racional de recursos veg et al.es. Herbario Nacional de Bolivia. La
Paz, Bolivia
16. García E, Beck S (2006) Puna. In: Moraes RM, Ollgaard B, Kvist LP, Borchsenius F,
Balslev H (eds) Botánica Económica de los Andes Centrales. Universidad Mayor de San
Andrés, La Paz, pp 51–76
17. Gisbert T, Arce S, Cajias M (1984) Los Textiles de Charazani en su contexto histórico y
cultural. In: Espacio y tiempo en el Mundo Callahuaya
18. Greslou F (1989) Visión y crianza campesina de los animales andinos. In: Crianza de llamas y
Alpacas. PAL – PRATEC. Proyecto Andino, pp 11–44
19. Hoffmann D, Requena C (2012) Bolivia en un mundo 4 grados más caliente – Escenarios
sociopolíticos ante el cambio climático para los años 2030 y 2060 en el Altiplano Norte.
Fundación PIEB, Instituto Boliviano de la Montaña, La Paz, 160 p
20. Hoffmann D, Tarquino R, Corro Ayala J, Lavadenz L (2014) Métodos para caracterizar la
dinámica de los sistemas sociológicos asociados a los bofedales altoandinos (Cordillera Real,
Bolivia). Ecología en Bolivia 49(3):132–140
21. Jordan E (2015) Patrones climáticos de la Cordillera Real: Efectos sobre los glaciares y la
vegetación altoandina. In: Meneses RI, Beck S, Anthelme F (eds) IRD & Herbario Nacional
de Bolivia, La Paz, Bolivia, pp 42–47
22. Liberman M, Baied C (eds) (1997) Desarrollo Sostenible en ecosistemas de montaña. Manejo
de áreas frágiles en los Andes. Artes Gráficas Latina, La Paz, 383 p
Climate Change, Wetland Management and Alpaca Pastoralism … 97

23. Liberman M, Molinillo M (2015) Bofedales de Apolobamba. Estudio de línea base de

bofedales priorizados en el Área Natural de Manejo Integrado Apolobama, Bolivia.
Soluciones Prácticas, La Paz, 160 p
24. Lino Pimentel F, Padilla Callejas V (2015) Gestión de Riesgos a Partir de Saberes
Ancestrales. Instituto de Ecología – Nación Jach’a Suyu Pakajaqi, La Paz, Bolivia, 156 p
25. Maldonado M, Navarro G, Acosta F, Aguilera X, de la Barra N, Cadima M, Coronel J,
Fernández E, Ferreira W, Goitia E (2012) Humedales y cambio climático en los Altos Andes
de Bolivia. UMSS – RUMBOL, 40 p
26. Meneses RI, Beck S, Anthelme F (eds) (2015) La Cordillera Real y sus Plantas. IRD &
Herbario Nacional de Bolivia, La Paz, Bolivia, 364 p
27. Meneses RI, García Lino MC, Beck S, Anthelme F, Ibañez D, Lliully AE, Loza Herrera S,
Palabral-Aguilera AN, Pfanzelt S, Villavicencio X (2015) In: Meneses RI, Beck S,
Anthelme F (eds) La Cordillera Real y sus Plantas. IRD & Herbario Nacional de Bolivia, La
Paz, Bolivia, pp 91–364
28. Molinillo M (2003, Unpublished) Patrones de vegetación y pastoreo en ecosistemas alto
andinos: Una comparación de casos de estudio en Páramos y Punas. Universidad de los Andes
- Merida, Facultad de Ciencias, Postgrado en Ecología Tropical. Tesis, 177 p
29. Murra J (1972) El control vertical de un máximo de pisos ecológicos en la economía de las
sociedades andinas. Ensayo publicado en el tomo II de la visita de la Provincia de León de
Huanuco (1562). Iñigo Ortiz de Zúñiga, Visitador. Universidad Hermilia Valdizan, Huanuco,
Perú
30. Murra J (2002) El mundo andino. Población, medio ambiente y economía. IEP/Pontificia
Universidad Católica del Perú, Lima. Historia Andina, vol 24
31. Navarro G, Maldonado M (2002) Geografía Ecológica de Bolivia: Vegetación y Ambientes
Acuáticos. Centro de Ecología Simón I. Patiño – Departamento de Difusión, Cochabamba,
Bolivia
32. Nina Huanca GS, Loza Herrera S, Gonzales Pomar RK, Rebaudo F (2015) El ser humano: un
actor de la dinámica de los ecosistemas altoandinos de la Cordillera Real. In: Meneses RI,
Beck S, Anthelme F (eds) IRD & Herbario Nacional de Bolivia, La Paz, Bolivia, pp 64–76
33. Olmos C (2005) Résponse des glaciers Boliviens á la variabilité clamátique recente et aux
évenements ENSO – Influence du taux d’englacement et de l’orientation, D.E.A. thesis.
Université de Montpellier II, Montpellier, 61 p
34. Palacios F (1988) Tecnología del pastoreo. In: Ochoa F (ed) Llamichos y Paqocheros:
Pastores de Llamas y Alpacas. Ed. UNSAAC, Cusco, pp 87–100
35. Prieto G, Alzerreca H, Laura J, Luna D, Laguna S (2003) Características y distribución de los
bofedales en el ámbito Boliviano del Sistema T.D.P.S. In: Rocha O, Sáez C (eds) Uso pastoril
en humedales altoandinos. Dirección General de Biodiversidad, La Paz, Bolivia, pp 13–38
36. Rauh W (1978) Die Wuchs-und Lebensformen der tropischen Hochgebirgsregionen und der
Subantarktis – ein Vergleich. In: Troll C, Lauer W (eds) Geoökologische Beziehungen
zwischen der temperierten Zone der Südhalbkugel und den Tropengebirgen. Erdwiss.
Forschung 11:62–92
37. Ribera MO, Rocha O (2003) Uso pastoril en humedales de áreas protegidas altoandinas de
Bolivia. In: Rocha O, Sáez C (eds) Uso pastoril en humedales altoandinos. Talleres de
capacitación para el manejo integrado de los humedales altoandinos de Argentina, Bolivia,
Chile y Perú. Sitio RAMSAR – Lago Titicaca, Huarina 28 de Octubre al 1 de Noviembre
2002. Convención RAMSAR, WCS/Bolivia, La Paz, Bolivia, 135–156 pp
38. Ribera MO, Liberman M (2006) El uso de la tierra y los recursos de la biodiversidad en las
áreas protegidas de Bolivia. Un análisis crítico con propuestas para su conservación y manejo
sostenible. SERNAP – GEF II, La Paz, 520 p
39. Rocha O (2014) Primer Diagnóstico de la Zona Montañosa del Estado Plurinacional de
Bolivia. Ministerio de Relaciones Exteriores, La Paz, 140 p
40. Seibert P (1993) La vegetación de la región de los Kallawaya y del Altiplano de Ulla Ulla en
los Andes Bolivianos. Ecología En Bolivia 20:1–84
98 M. Liberman

41. SERNAP (2016) Plan de manejo del Área Natural de Manejo Integrado (ANMI)
Apolobamba. Servicio Nacional de Áreas Protegidas de Bolivia (SERNAP)
42. Soruco A (2012) Medio Siglo de Fluctuaciones Glaciares en la Cordillera Real y sus efectos
Hidrológicos en la ciudad de La Paz. Institut de Reserche pour le Dévelopment (IRD), La Paz,
Bolivia, 228 p
43. Troll C (1959) Die tropischen Gebirge. Bonner Geogr. Abh. 25:1–93
44. Vassas Toral A (2016) Partir y cultivar. IRD-CIDES-UMSA/Plural Editores, La Paz, 306 p
45. Vuille M, Francou B, Wagnon P, Juen I, Kaser G, Mark BG, Bradley RS (2008) Climate
change and tropical Andean glaciers – past, present, future. Earth Sci Rev 89:79–96
46. Vuille M (2013) El cambio climático y los recursos hídricos en los Andes Tropicales. Banco
Interamericano de Desarrollo (BID). Nota técnica IDB – TN-517, 29 p
47. Wheeler JC (1984) La domesticación de la Alpaca (Lama pacos L.) y la Llama (Lama glama
L.) y el desarrollo de la Ganadería Autóctona en los Andes Centrales. Boletín de Lima 6
(36):74–84
Essays on Geobotanic Mapping
in the Andes of Bolivia, with Particular
Reference to the Conservation Status
of the Vegetation

Franco Pedrotti

Abstract The purposes of this contribution is to present some maps of the vege-
tation of Andes of Bolivia, with particular reference to the conservation status of
vegetation. The study areas are the following: Capinota Valley, Cochabamba (1:
50,000), the Camacho river basin, Tarija (1: 50,000), Navado Sajama, Oruro (1:
130,000) and Isla del Sol in Lake Titicaca, Bolivian plateau (1: 16,000). For each
map there is a description on the environnement and vegetation (phytocoenoses,
vegetation series and vegetation geoseries), with indication of dynamic tendencies
of vegetation: natural fluctuation, regeneration, secondary succession, regression.
The conservation status of vegetation (understood at a phytocoenoses, series and
geoseries) is very different in the four cases studied. In Capinota, Isla del Sol and
Rio Camacho, the forest communities were eliminated and only a few isolated
fragments remain today; in Capinota forests of Schinopsis haenkeana and
Aspidosperma quebracho-blanco; on the Isla del Sol matorrales of Polylepis
incarum; and in the Rio Camacho river forests of Tipuana tipu. On the Nevado
Sajama volcano, despite a notable anthropogenic influence, both phytocoenoses
(matorral of Polylepis tarapacana) and geoseries (including the vegetation series)
have been maintained, although with some areas in regression due to overgrazing
and fire.

Keywords Vegetation mapping Phytocoenoses Vegetation series Vegetation





geoseries Andes of Bolivia Capinota Nevado Sajama Isla del Sol Lake





Titicaca Camacho river Dynamical tendencies of vegetation Conservation

status of vegetation

F. Pedrotti (&)
Università di Camerino, Palazzo Castelli - Via Pontoni 5, 63032 Camerino, MC, Italy
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 99

F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_6
100 F. Pedrotti

1 Introduction

The purpose of this contribution is to present some maps of the vegetation of the
Andes of Bolivia with particular reference to the conservation status of vegetation.
The term “vegetation” is quite generic and refers to the vegetation cover in a broad
sense or plant cover or plant mass, as most botanists refer to it (see, for example, [9,
22, 24, 47] etc.). More specifically, it can be said that the vegetation cover is formed
by the set of plant communities present in a territory at a given time [8]. These
communities (plant associations) are made up of concrete units, the phytocoenoses,
each of which corresponds to an individual of association or elementary population.
The higher levels of synthesis consist of vegetation series and vegetation geoseries
[22, 58]. The conservation status of the vegetation is the result of the influence
exerted by man on the vegetation cover and is evaluated with different method-
ologies depending on the authors. It is proposed here to refer to two parameters: the
phytosociological typology (phytocoenoses, vegetation series, vegetation geoseries)
and dynamic processes affecting vegetation.

2 Methodology

A vegetation map is a topographic map in which the specific units that form the
vegetation of a specific territory are summarized graphically, that is, the phyto-
coenoses and their spatial relationships [48]. It represents the last stage of a research
process that begins with the recognition of phytocoenoses in the field, their
delimitation, their detection and the definition of a typology that may be different
according to the different geobotanical schools. The Braun-Blanquet-Tüxen School
of Phytosociology describes associations of plants and their classification in a
hierarchical system: associations, alliances, orders and classes [22].
A map of vegetation is composed, therefore, of the following elements: topo-
graphic map, cartographic units indicated with different colors and delimited by a line
called the “vegetation limit”, and a list of the vegetation units represented on the map.
Mapping dynamical tendencies involves producing a plant community map and
specifying two labels in each patch, i.e. one referring to the vegetation type and the
other to the dominant dynamical tendency. The first example of such a map was the
Map of Bosco Quarto, Gargano (Italy), printed at the scale 1:10,000 [20]. Although
the dynamical tendencies distinguished have often been reported and used since
then, their nomenclature has not yet been systematised.
Conservation status is evaluated relative to certain units of vegetation, which are
the following: phytocoenoses, that is the concrete units of the associations (iden-
tified by phytosociology); vegetation series (according to synphytososociology);
and vegetation geoseries (according to geosynphytosociology). The dynamic pro-
cesses are the following: fluctuation (of natural, semi-natural and anthropogenic
origin), degeneration, regeneration, primary succession, secondary succession, and
regression [19].
Essays on Geobotanic Mapping in the Andes of Bolivia ... 101

Phytocoenoses are evaluated through intrinsic characteristics of the vegetation,

such as floristic composition, vertical and horizontal structure, and dynamic pro-
cesses. The individual phytocoenoses of the same association are evaluated sepa-
rately, because they can be affected by different ecological processes. When the
associations are not known, reference can be made to vegetation formations. The
starting point for the evaluation of phytocoenoses is the phytosociological map of
the actual natural vegetation. This map must be completed with the map of dynamic
vegetation tendencies. Different levels of conservation [status] are associated with
each dynamic tendency.
Vegetation series are evaluated with reference to the space they occupy, the
homogeneous fragment of the landscape, i.e. the ecotope or “patch” (called tesela in
Spanish). The reference parameter is the relationship forest/territory evaluated in a
concrete form, i.e. within the space of the patch; we also define the current ratio of
actual vegetation to potential vegetation. The series is evaluated in the space in
which it develops, that is its patch (tesela), which can be completely occupied by
the series association or occupied to a decreasing extent, up to that of a patch
formed exclusively by secondary vegetation. With this procedure, it is possible to
calculate the percentage of area covered by forest with respect to secondary veg-
etation. The reference for evaluation of the whole series is the integrated phy-
tosociological map (synphytosociological map).
Vegetation geoseries occupy more or less vast territories, which have undergone
human interference to varying degrees. The reference for evaluation of geoseries is
the geosynphytosociological map. Also, in this case, evaluation consists of quan-
tifying the ratio of primary vegetation to secondary vegetation, but on a broader
scale than in the previous case. This leads to a global assessment of the landscape. It
is thus possible to evaluate wider areas, at the level of the landscape.
For each area of study, a scale of values relative to the state of conservation of
the vegetation has been proposed, without reaching a quantification. The criterion
followed was to place natural vegetation (where existing) at the first place on the
scale, affected by the processes of fluctuation and regeneration, and therefore the
shrub associations resulting from secondary succession, the pastures obtained from
the elimination of primary vegetation and finally the synanthropic units, which are
all affected by the process of anthropic fluctuation, due to periodic intervention by
man. This scale of values, which is based on the series of vegetation (i.e. on the
different stages that compose them) and their dynamic tendencies, can be consid-
ered of general value. On the other hand, the scales proposed for the four areas to be
studied (see below) detected have only a local meaning relative only to themselves.
A map of the conservation status of the vegetation was also made.
Among the many authors who dealt with the conservation status of vegetation,
Carboni et al. [10], Asensi et al. [1, 2], and Pinna et al. [55] refer to phytocoenoses
and landscape; Slater et al. [59], Davis et al. [16] and Miles et al. [42] to a single
type of vegetation (vegetation of river corridors in Wales, coastal scrub in
California and dry forests of the world, respectively), Del Arco Aguilar et al. [17]
refers to actual and potential vegetation; and Jewitt [25] to vegetation types.
102 F. Pedrotti

3 The Study Areas

The study areas are the following: Capinota (Cochabamba), the Camacho River
(Tarija), Nevado Sajama, and Isla del Sol, in Lake Titicaca (Bolivian plateau).
Capinota and the Camacho River involve mesothermic valleys, in which the veg-
etation is tall forest of different types [44]. The Nevado Sajama and Isla del Sol
belong to the puna, a term used to indicate the high-mountain environment and
vegetation of the Andes between 3000 and 4500 m. The emblematic species of the
puna is Puya raimondii, the largest known bromeliad, with a height of 9–10 m [50].
For each map there is a description of the environment and vegetation, indication
of dynamic tendencies with an acronym, a summary of the general characteristics of
the map, and the reproduction of the map (Map of Nevado Sajama) or of a section
of the map (Maps of Capinota, Isla del Sol and the Camacho river basin). The maps
reproduced have been reduced from their original scale, in order to adapt them for
printing. For each map, a legend for the vegetation units is also given.

4 Vegetation Map of the Capinota Valley (Cochabamba)

The study area (Capinota Valley) is located in the lower valley of the Department of
Cochabamba, between the mouth of the Tapacarí River and the town of Orcoma
(2250–3250 m). Orographically, this area belongs to the Cordillera de los Andes
and is part of the ramification of Tunari and Mazo Cruz.
This valley is one of the dry, mesothermic inter-Andean valleys, whose vege-
tation is characterized by the dominant presence of xerophytic, deciduous or
evergreen, microphyllous species. It belongs to the Boliviano-Tucuman province of
the Andean region [45].
The vegetation has been altered significantly by the action of man and is sec-
ondary vegetation, in which there are some elements of primary vegetation, such as
the vegetation of the rivers (Cortaderia atacamensis and Tessaria dodonaeifolia
with the presence of isolated saplings of Salix humboldtiana) and the forests of
Schinopsis haenkeana and Aspidosperma quebracho-blanco. These forests corre-
spond to the series of Carica quercifolia and Schinopsis haenkeana, described by
Navarro [44] for the Cochabamba basin. In the areas along the rivers (alluvial
deposits) there is intensive cultivation of crops, and in the other areas there are
annual dry-land crops.
The vegetation map represents the actual natural vegetation and refers to the
plant communities gathered in physiognomic categories: herbaceous communities
(Neocracca heterantha and Onoseris gnaphalioides, Cortaderia atacamensis,
Viguiera lanceolata), scrubs (Tessaria dodonaeifolia, Dodonaea viscosa), chaparral
(tall matorral) of Acacia macracantha and Prosopis laevigata var. andicola), forests
(Schinopsis haenkeana and Aspidosperma quebracho-blanco), and synanthropic
vegetation (of intensive and annual crops). In most of the Capinota Valley the
Essays on Geobotanic Mapping in the Andes of Bolivia ... 103

potential vegetation is forest of Schinopsis haenkeana and Aspidosperma

quebracho-blanco on the hillsides and riparian forest of Salix humboldtiana along
the rivers.

Fig. 1 Map of communities in a sector of the Capinota Valley with indication of the relative
dynamic tendencies (from [35], modified)
104 F. Pedrotti

General characteristics of the map (Fig. 1)

Size: 78  60 cm
Scale: 1: 50,000
Topographic map: map of the Instituto Geografico Militar (Bolivia).
Vegetation: herbaceous, shrub vegetation, chaparral (tall matorral), matorral, forest,
and synanthropic communities.
Number of communities represented on the map: 11.
Cartographic representation of the vegetation: cartographic units and symbols (the
forests of Schinopsis haenkeana and Aspidosperma quebracho-blanco).
Vegetation limits indicated in black.
Type of map: map of communities, actual natural vegetation.
Bibliography: Pedrotti et al. ([35, 54]),Pedrotti and Venanzoni [51].
Conservation Status of the Vegetation
The map of the actual natural vegetation of the Valley of Capinota (1: 50,000)
served for the detection of dynamic vegetation tendencies. The remaining forests of
Schinopsis haenkeana and Aspidosperma quebracho-blanco are affected by
regeneration processes; the matorrales and chaparrales by the processes of sec-
ondary succession; the riparian vegetation (Cortaderia atacamensis and Tessaria
dodonaeifolia, with isolated trees of Salix humboldtiana) by primary succession;
and the anthropic fluctuation manifests itself in the cultivated areas and in the
ruderal and nitrophilous areas (Fig. 1).
The action of man has caused the following modifications to the vegetation: 1)
elimination of forest vegetation (Schinopsis haenkeana and Aspidosperma
quebracho-blanco), of which only two residual nuclei remain; 2) development of
secondary communities of chaparral and matorral, Prosopis laevigata ssp. andicola
at lower elevations (Fig. 2) and Dodonaea viscosa higher; and 3) transformation of
the most favorable areas into cultivated areas.
The scale of Capinota vegetation values is as follows:
I) Primary vegetation: rocky outcrops (Neocracca heterantha and Onoseris
gnaphalioides) and streambeds (Cortaderia atacamensis and Tessaria
dodonaeifolia)
II) Forests of Schinopsis haenkeana and Aspidosperma quebracho-blanco
III) Matorrales and chaparrales (Prosopis laevigata ssp. andicola; Dodonaea
viscosa)
IV) Agricultural crops.

5 Map of the Vegetation of the Nevado Sajama (Oruro)

The study area is located in the northern part of the Bolivian plateau and covers
approximately 400 km2; this area can be divided into three physiographic units
according to relief: a volcanic cone, glacial valleys and piedmont plains. It belongs
to the plateau province [provincia altiplanica] of the Andean region [45].
Essays on Geobotanic Mapping in the Andes of Bolivia ... 105

The volcanic cone is that of the Nevado Sajama, reaching 6,542 m of altitude.
The upper part of the Nevado is affected by recent glaciations, with large accu-
mulations of snow and snowdrifts that have cracks in the upper part. On the slopes
and around the Nevado Sajama, there are large accumulations of volcanic material
(glacial cirques and glacial valleys). The piedmont plains are between 4,300 and
4,500 m of altitude.
In the Nevado Sajama area we can distinguish the following vegetation for-
mations: high-Andean herbaceous vegetation (Calamagrostis curvula, Festuca
orthophylla, Azorella compacta, Pycnophyllum molle); matorrales (at the upper
limit of distribution) and forests of Polylepis tarapacana; bofedales (wetlands of
Oxychloe andina, Calamagrostis ovata, Werneria pygmaea); grass-shrub steppes
(Parastrephia lepidophylla, Baccharis incarum, Festuca orthophylla, Poa asperi-
flora, Pycnophyllum tetrastichum); grass steppes resulting from burning or natural
fires (Festuca orthophylla, Calamagrostis curvula, Calamagrostis antoniana, Poa
asperifolia, Muehlenbergia peruviana); matorrales of Parastrephia lepidophylla;
and vegetation of saline areas (Distichlis humilis, Puccinellia sp., Salicornia sp.).
The matorrales and open forests of Polylepis tarapacana develop only on the
slopes of the volcanic cone, from the base of the volcano (4,100 m) to 5,200 m,
followed by a strip of herbaceous vegetation of high-Andean steppe. These form a
concentric ring interrupted by lava flows and deposits of volcanic slag. Martín
Cárdenas called them “the highest forests in the world” by altitude [11]. Today
these forests have a degree of coverage between 60 and 80%, but some
better-preserved nuclei reach 100%. The age pyramid of two populations of
Polylepis tarapacana shows the low proportion of seedlings, which have great
difficulties to develop in volcanic waste [29]. More frequent is the renewal of this
species in forest clearings caused by cutting by man and by fires. Today these
clearings are occupied by secondary grasslands of Festuca orthophylla. In this case,
plants of Festuca orthophylla facilitate the vegetative development and germination
and establishment of seeds of Polylepis tarapacana (facilitation phenomenon).
These forests belong to the Mutisio lanigerae-Polylepidetum tarapacanae associ-
ation [45, 46].
Grazing by alpacas and llamas is intense [29, 31, 32, 57].
General characteristics of the map (Fig. 3)
Size: 54  35 cm
Scale: 1: 130,000
Topographic map: map drawn in reduced form with these elements: hydrographic
network, mountains, towns.
Vegetation: plant formations (high-Andean herbaceous vegetation, herbaceous
vegetation, bofedales, matorrales, forests, steppes).
Number of formations represented on the map: 7.
Cartographic representation of the vegetation: cartographic units.
Vegetation limits are indicated in black.
Type of map: map of the formations, actual natural vegetation.
Bibliography: Liberman Cruz [26], Liberman et al. [29, 37].
106 F. Pedrotti

Fig. 2 Capinota, profiles of the Acacia macracantha and Prosopis laevigata var. andicola
chaparral (tall matorral); a: stage with prevalence of Cactaceae, secondary succession; b: stage
with prevalence of Prosopis laevigata var. andicola, secondary succession, more advanced stage;
c: on the right, stage with plants of Schinopsis haenkeana, secondary succession, more advanced
stage; on the left, residual forest of Schinopsis haenkeana, degeneration/regeneration. The arrows
indicate the progressive evolution of the vegetation (from [53] modified)
Essays on Geobotanic Mapping in the Andes of Bolivia ... 107

Fig. 3 Map of formations of the Nevado Sajama with indication of the relative dynamic
tendencies (from [26], modified)

Conservation Status of Vegetation

In the area of the Nevado Sajama grazing is by alpacas and llamas, which has led to
various changes in vegetation, as shown by the map of dynamic tendencies (Fig. 3).
The high-Andean herbaceous vegetation is affected by this fluctuation; the Polylepis
matorrales and forests have different aspects, degeneration in those parts where
grazing takes place, regeneration and even fluctuation in some well-preserved
nuclei. Some slope areas are in regression, due to fires. The slopes of the Nevado
Sajama correspond to a geoseries, composed by 4 vegetation series (Fig. 4); on the
whole, it should be noted that the belt of the matorrales and forests of Polylepis
108 F. Pedrotti

Fig. 4 The northern slope of the Nevado Sajama corresponds to a vegetation geoseries formed by
2 vegetation series: high-Andean herbaceous vegetation, matorrales and forests of Polylepis
tarapacana. For each formation the dynamical tendencies are indicated

Fig. 5 Formations of the plains surrounding the Nevado Sajama, with indication of the relative
dynamic tendencies

tarapacana is well preserved, although regression and degeneration are occurring.

The same applies to the piedmont plain (Fig. 5).
The scale of vegetation values of the Nevado Sajama is as follows:
I) High-Andean vegetation
II) Matorrales and forests of Polylepis tarapacana
III) Bofedales, matorrales of Parastrephia lepidophylla and others
IV) Grass steppe after fire, hyper-nitrophic areas.
Essays on Geobotanic Mapping in the Andes of Bolivia ... 109

6 Map of the Vegetation of Isla del Sol (Lake Titicaca)

The Isla del Sol is located in Lake Titicaca (3810 m); it has a surface area of 20.9
km2 and an irregular shape, with a NW–SE orientation. The island is formed by two
mountain chains with elevations that go up to 4045 and 4076 m. In the past, the
island was a sacred place for the people who lived near the lake, and it is believed
that the Inca Empire originated here [60]. Today it is possible to observe remnants
of the ruins of ancient Inca temples and terraces that are still used by the Aymara
peasants for their crops (Fig. 6). The actual vegetation of the Isla del Sol is the
result of man’s influence on the environment (Figs. 7 and 8). The original forests of
keñua (Polylepis incarum) [21, 23] were completely felled, and today there are only
two small remnants left. This area belongs to the Peruvian Puna province of the
Andean region [45].
Herbaceous Communities. There is a group of three communities, characterized
by the presence of Plantago sericea ssp. polyclada, which develop on rocky slopes,
rounded peaks and on the top of a mountain above 4,000 m. Another community is
characterized by Plantago sericea ssp. sericans and forms a continuous grass cover
on some hills. Flooded areas are characterized by different communities of
Juncaceae (Juncus stipulatus, J. ebracteatus, J. arcticus var. andicola) and other
species (Carex boliviensis, Hydrocotyle ranunculoides).

Fig. 6 Isla del Sol, with slope terraces that date back to the Inca epoch; in the foreground the
remains of the Inca palace, from a drawing from the years 1863–1865 (from [60])
110 F. Pedrotti

Fig. 7 Vegetation profile of

the northeastern slope of
Cerro Palla Khasa (Isla del
Sol), with indication of the
dynamical tendencies of
vegetation

Fig. 8 Vegetation profile of the northeastern slope of Cerro Chequesani (Isla del Sol), with
indication of the dynamical tendencies of vegetation

Shrub Communities. The most frequent community is the “tolares”, which

occupy great extents and involve Baccharis incarum (tola) and Satureja boliviana.
This community also develops on the walls of the terraces and on fallow lands,
forming open shrub thickets. In two areas of this community, Polylepis incarum
was observed, in its shrubby form, with a major renovation. In rocky and humid
places, a community of Pernettya prostrata was observed. Finally, the community
of Puya mollis should be noted, occurring on slopes with NE exposure and a rocky
sandstone substratum.
Synanthropic Communities. These are the communities associated with inten-
sive and extensive crops. The main crops of Isla del Sol are: different varieties of
Solanum sp. (potato), Oxalis tuberosa (oca), Ullucus tuberosus (ullucu),
Chenopodium quinoa (quinoa), Zea mays (maize), Lupinus mutabilis (lupine) and
Hordeum vulgare (barley). When the crop areas are abandoned or are resting for
long periods of time, different shrubs develop, of which Baccharis incarum is
usually dominant, in other cases co-dominant with the Lupinus prostratus. These
matorrales, and their herbaceous companion species, constitute stages of develop-
ment towards the matorral of Baccharis incarum and Satureja boliviana.
Essays on Geobotanic Mapping in the Andes of Bolivia ... 111

Fig. 9 Profile of the vegetation of the shores of Lake Titicaca; A: Lemnetum minusculae-gibbae;
B: Lemnetum valdivianae, with 1 - Myriophyllum elatinoides, 2 - Schoenoplectus californicus
ssp. tatora, and 3 - Chara sp. The vegetation is affected by fluctuation processes (from [35])

Lake Titicaca Communities (Fig. 9). Lake Titicaca is a large, tropical

high-mountain lake with slightly mesotrophic-oligotrophic water [56]. The shallow
water near the shoreline shows eutrophication almost everywhere, due to pollution
from the human settlements [27]. The vegetation on the banks has the following
zonation: shallow water with Myriophyllum elatinoides and Elodea potamogeton,
inland water with Schoenoplectus californicus ssp. tatora (totora), and deep water
with Chara sp. pl. [12, 13, 14]. The floating vegetation involves two associations:
Lemnetum minusculae-gibbae in the shallow water along the shores, with Lemna
gibba, Lemna minuscula and Azolla filiculoides; and Lemnetum valdivianae in the
water far from shorelines, between the totora plants, with Lemna valdiviana, Azolla
filiculoides and Utricularia obtusa [52]. The association Lemnetum minusculae-
gibbae was also found in Lake Uru-Uru, still in the Bolivian plateau south of Lake
Titicaca [36]. Along the shores of the Isla del Sol, the presence of the Schoenoplectus
californicus ssp. tatora, present in only one place, has been mapped with a symbol.
112 F. Pedrotti

Fig. 10 Map of communities of the Isla del Sol, with indication of the relative dynamic tendencies
(from [33] modified)
Essays on Geobotanic Mapping in the Andes of Bolivia ... 113

General characteristics of the map (Fig. 10)

Size: 82  68 cm
Scale: 1: 16,000
Topographic map: map of the Instituto Geografico Militar (Bolivia).
Vegetation: plant communities (herbaceous, scrubland, synanthropic)
Number of communities represented on the map: 23.

Fig. 11 Map of the potential vegetation of the Isla del Sol

114 F. Pedrotti

Fig. 12 Remnants of original vegetation of the Isla del Sol (from [33], integrated with [18] and
[41])

Cartographic representation of vegetation: cartographic units and symbols. Symbols

have been used to indicate less extensive communities (such as the communities of
Carex boliviensis, Hydrocotyle ranunculoides and Schoenoplectus californicus
spp. tatora) and some rare tree species (Polylepis incarum, Buddleja coriacea,
Sambucus peruana, Alnus acuminata ssp. acuminata) and other species (Puya
mollis, Festuca dolichophylla).
Essays on Geobotanic Mapping in the Andes of Bolivia ... 115

Vegetation limits are indicated in black.

Type of map: map of communities, actual natural vegetation.
Bibliography: Liberman Cruz et al. [28, 33, 34]; Pedrotti [49].
Conservation Status of Vegetation
The map of the actual natural vegetation has allowed us to make the map of potential
vegetation, distinct in zonal and azonal vegetation (Fig. 11). In the past the whole
island was covered by Polylepis incarum, today limited to two residual nuclei and to
groups of plants in some patches (Fig. 12). Martínez and Villarte [41] report the
presence in two patches of centenarian trees of Polylepis incarum more than 12 m
high and with a diameter of 69 cm. The azonal vegetation is represented by the
vegetation of humid environments (bofedales and alluvial plains, Lake Titicaca) and
by Pernettya prostrata shrubs limited to rocky stations with northern exposure.
On the map of the actual natural vegetation are indicated the dynamic tendencies
from which the phytocoenoses are affected, which are the following (Fig. 10):
regeneration (Polylepis incarum community); secondary succession (Baccharis
incarum and Satureja boliviana community, Puya mollis community); forced
secondary succession (reforestation with Eucalyptus sp.); natural fluctuation (humid
meadows of bofedales, wet meadows of alluvial plains, aquatic vegetation, lake
vegetation); and fluctuation of anthropogenic origin (pasture communities; agri-
cultural crops - intensive, extensive, at rest, abandoned; and ruderal and nitrophi-
lous vegetation of the villages).
The scale of the vegetation values of the Isla del Sol is as follows:
I) Polylepis incarum scrub, Pernettya prostrata scrub, humid meadows of
bofedales, wet meadows of alluvial plains, aquatic vegetation, lake
vegetation
II) Baccharis incarum and Satureja boliviana scrubs
III) Puya mollis scrubs
IV) Pastures (Plantago sericea ssp. polyclada, P. s. ssp. sericans)
V) Agricultural crops (intensive, extensive, rest, abandoned), ruderal vegetation
of the villages
VI) Reforestation of Eucalyptus sp.
Six categories have been reported on the map of actual natural vegetation, so as
to obtain the map of vegetation conservation status values (Fig. 13). On the basis of
this map it can be seen that, on the Isla del Sol, the vegetation has undergone the
following changes:
1) elimination of Polylepis incarum forests
2) formation of different types of secondary vegetation (matorrales and pastures)
3) large fragmentation of secondary vegetation
4) maintenance of some types of vegetation (bofedales, Pernettya prostrata).
5) transformation of the most favorable areas into agricultural areas.
116 F. Pedrotti

Fig. 13 Map of the values of the communities of the Isla del Sol
Essays on Geobotanic Mapping in the Andes of Bolivia ... 117

7 Map of Forests of the Camacho River Basin (Tarija)

The Camacho river basin is located in an inter-Andean valley in the extreme south
of Bolivia, in the Department of Tarija. The approximate area is 1,050 km2, with
altitudes that go from 1600 m, at the confluence of the Camacho river with the
Guadalquivir river, up to 4500 m, in and near the Taxara mountains. The basin of
the Central Valley of Tarija is a mesothermic valley in which the rainfall is sea-
sonal, concentrated between December and March. In the Camacho river basin,
average annual precipitation varies from 300 mm in the area of the Angustura to
1400 mm in the cordilleran zone in the southwestern part of the basin. This area
belongs to the Boliviano-Tucuman province of the Andean region [45].
The original vegetation has been strongly transformed by man in the past cen-
turies, going back mainly to the arrival of the Spaniards [15, 43]. The few forests in
acceptable condition are found in areas almost inaccessible to man, such as ravines
or wet gorges, hillsides and rocky areas with steep slopes and riverbeds. The
original forests were in many cases replaced by secondary formations of matorral
and shrub scrub. Isolated trees appear in extreme cases, testimony to what the
original vegetation was like.
White Alder Forests (Alnus acuminata ssp. acuminata). Alder forests occupy
the upper part of the humid montane forest belt in the sub-Andean strip of the Andes.
The dominant species is white alder, which occurs from the base of the mountains up
to the subalpine belt, on slopes of southern exposure, between 2300 and 3000 m, in
areas where humid winds and clouds, originating in Argentina.
Keñua Forests (Polylepis crista-galli and P. hieronimi). These correspond to the
areas of semi-humid to semi-dry forests in the mountainous Andes. Forests of
keñua occur on mountains with slopes from 30 to 40°. At the southern end of the
basin, where there are the most mountains and valleys with northern exposure, the
keñua forests reach their highest elevations, due to the Massenerhebung effect.
Pino del Cerro Forests (Podocarpus parlatorei). These belong to the area of
humid high-montane forests, occurring in a vegetation belt below to that of the
alder forests. These podocarp forests develop in areas with variable slopes, from
almost flat to steep (10–15°) to even steeper (30–40°).
Tipa - Jacaranda Forests (Tipuana tipu – Jacaranda mimosifolia). This is a
climax community that has been practically destroyed, because its distribution area
coincides with areas where man has developed intensive agriculture, especially in
areas located on the fluvio - lacustrine sediments of the Camacho river basin.
Currently, there are only some groups of trees along streams or ledges where
humidity accumulates. The tipa trees are the most conspicuous in the whole basin.
These forests correspond to the series of Jacaranda mimosifolia and Tipuana tipu,
described by Navarro [44].
118 F. Pedrotti

Fig. 14 Map of forests in a sector of the Río Camacho basin, with indication of the relative
dynamic tendencies (from [30] modified)

Molle Forests (Schinus molle). These are present as very open formations with a
variable coverage of 30–60%, in areas with much soil moisture, such as the alluvial
plains, and the banks and terraces of rivers and streams.
Areas Without Forest Vegetation. Secondary formations of matorral and shrub
thickets: red alder scrub (Morella chevalieri), chañar matorral (Geoffroea decorti-
cans), tussock matorral (Acacia aroma), scrub (Prosopis alpataco), churqui
matorral (Acacia caven), and others. The churqui scrub in the Camacho basin is the
unit with the greatest geographic extent, colonizing various ecosystems due to its
great capacity for adaptation.
General characteristics of the map (Fig. 14)
Size: 43  43 cm
Scale: 1: 50,000
Topographic map: map drawn in reduced form, with urban areas, towns, roads,
mountain tops, streams.
Vegetation: forests, vegetation series.
Number of formations represented on the map: 5.
Essays on Geobotanic Mapping in the Andes of Bolivia ... 119

Fig. 15 Regression of the forests of Tipuana tipu and Jacaranda mimosifolia in the basin of the
Río Camacho, Puesto Tunal. We observed an isolated plant of tipa (Tipuana tipu) on a soil column
as well as several churqui (Acacia caven) plants: a - original level of the soil; b - intermediate level;
c - level in the year of detection (1988). The arrow indicates the thickness of the eroded soil, about
3.50 m

Cartographic representation of the vegetation: cartographic units.

Vegetation limits indicated in black.
Type of map: partial map, showing only forests (Fig. 14); two maps represent
vegetation series (Figs. 15 and 16).
Bibliography: Beck et al. [3–5]; Liberman Cruz et al. [30, 38].
Conservation Status of Vegetation
For the Camacho river basin, a partial map was surveyed, at a scale of 1: 50,000, on
which only the tree formations listed above appear. The primary forest vegetation is
now strongly fragmented and covers only 4.3% of the entire basin, corresponding to
45 km2. Of this area, 2.9% is covered by Tipuana tipu, 12.4% by Podocarpus
parlatorei, 16.9% by Polylepis crista-galli and P. hieronymi, 20% by Schinus molle
and 45.8% by Alnus acuminata ssp. acuminata. These residual forests are in the
120 F. Pedrotti

processes of degeneration (Tipuana tipu) and regeneration, with areas of fluctuation

(Alnus acuminata ssp. acuminata and Podocarpus parlatorei). The remaining part
of the basin is today subjected to regression, with different types of matorrales.
Deforestation has resulted in a desertification process, originated by destruction and
soil transport (Fig. 15).

Fig. 16 Río Camacho basin, map of the vegetation series; series of Tipuana tipu and Jacaranda
mimosifolia (survey of M. Liberman Cruz and F. Pedrotti)
Essays on Geobotanic Mapping in the Andes of Bolivia ... 121

The Tipuana tipu forests correspond to the series of Jacaranda mimosifolia and
Tipuana tipu, described by Navarro [44]. This series consists of the forest of
Tipuana tipu and Jacaranda mimosifolia (head of the series) and Acacia caven
matorral. Regression of the Tipuana tipu forests led to the formation of churqui
(Acacia caven) scrub, or churquial, which tends to form irregular groups (Fig. 16)
with an almost non-existent herbaceous layer.

Fig. 17 Río Camacho basin, map of the vegetation series; A/P series of Alnus acuminata
ssp. Acuminata and Polylepis crista-galli - P. hieronymi; T series of Tipuana tipu and Jacaranda
mimosifolia (survey of M. Liberman Cruz and F. Pedrotti)
122 F. Pedrotti

The matorrales of churqui (Acacia caven) correspond to a successful formation

in the substitution series of the original forests of tipa and jacaranda, which existed
before the arrival of man in the central valley of Tarija. This is a physiognomic
formation of secondary origin. Since churqui is one of the typical elements
accompanying the Chaco formations, its presence and current distribution in the
Camacho river basin obeys only anthropic influence. The churqui is a species of
great plasticity, which has allowed it to colonize diverse environments, both arid
and humid; to the above qualities one can add its high reproductive capacity and
rapid growth. These shrubs are grouped and constitute the unit with the greatest
geographic extent, colonizing various ecosystems due to their great adaptability. It
grows from 1,600 m to 2,300 m above sea level. The vegetation that accompanies
the churqui involves weeds and grasslands that, in one way or another, have been
favored by agricultural activities, such as the genus Paspalum, a grass that always
appears with important coverage.
Regression of the forests of Alnus acuminata ssp. acuminata leads to the
matorrales of red alder - aliso rojo (Morella chevalieri). The red alder matorral
corresponds to a stage in the substitution series of the forests of pino del cerro
(Podocarpus parlatorei) and, to a greater extent, of white alder (Alnus acuminata
ssp. acuminata) (Fig. 17). Red alder is found in the shrub layer of the podocarp
forests on hills and white alder, with little significant coverage. When the forests are
cut, devastated by uncontrolled overgrazing and fire, red alder develops secon-
darily, covering areas with remnants of the original forest. These are located on
south-facing slopes, forming dense pure scrub or pure cases that can reach up to a
height of 3 m; in other cases, they form small isolated shrub patches.
Red-alder matorrales are substitution formations of the former forests, occurring
in areas where white alders previously existed and, in some cases, hill podocarps.
The replacement is more evident near the lower distributional limits of the former
forests. Altitudinally, these grow between 2,000 and 2,500 m. Red-alder matorrales
occur in areas with south to southwest-facing slopes of 15° and 45°, with significant
coverage ranging from 95 to 100%.

8 Discussion

The four vegetation mapping trials shown here have different characteristics. The
maps of Capinota and Isla del Sol were made with complete topographic maps
(contour lines, hydrographic network, urban centers, roads, toponymy, etc.), which
undoubtedly facilitate reading and understanding. The maps of Nevado Sajama and
Río Camacho have, on the other hand, a simplified topography, limited to the
hydrographic network and some other elements, but nevertheless they are equally
legible. In addition, the maps differ because they refer to different units of vege-
tation: Capinota and Isla del Sol to plant communities, Nevado Sajama to plant
formations and Camacho river only to forests.
Essays on Geobotanic Mapping in the Andes of Bolivia ... 123

From these maps of the vegetation of the Andes of Bolivia, and from other
personal experiences, it seems that a vegetation map constitutes the graphic
response to a scientific problem placed a priori and related to the vegetation cover
of a given territory (formations, associations, series, geoseries, etc.). Therefore, a
map must have two fundamental requirements: appropriate scientific content,
according to a determined geo-botanical conception, and adequate graphic repre-
sentation, according to the prescriptions recommended by the general science of
symbols, the graphic semiology [6, 7, 40], which deals with the formal aspects of
the use of symbols in cartography.
The conservation status of vegetation (understood as phytocoenoses, series and
geoseries) is very different in the four cases studied. In Capinota, Isla del Sol and
Río Camacho, the forest communities were eliminated and only a few isolated
fragments remain today: in Capinota, forests of Schinopsis haenkeana and
Aspidosperma quebracho-blanco; on the Isla del Sol forests of Polylepis incarum;
and in the Río Camacho forests of Tipuana tipu. In terms of the landscape (geo-
series), the elimination of vegetation has triggered extensive erosion in the Río
Camacho basin, while in Capinota and Isla del Sol, secondary matorral vegetation
has developed.
On the Nevado Sajama volcano, despite a notable anthropogenic influence, both
phytocoenoses and geoseries (including the vegetation series) have been main-
tained, although with some areas in regression due to overgrazing and fire.
The results obtained with the proposed methodology are qualitative: only for the
Río Camacho basin was quantification relative to the residual forest present today
possible, which is only 4.3% of the entire basin, corresponding to 45 km2 out of
1,050 km2. In addition, the percentage of the species making up these residual
woods was also calculated.
Further evaluation of the conservation status of the vegetation cover could be
carried out by applying to the units chosen here (phytocoenoses, series, geoseries)
the indices proposed each time by the various authors mentioned above.

References

1. Asensi A, Díez-Garretas B (2010) Empleo de los hábitats naturales y seminaturales en la

evaluación de las Sierras del Levante almeriense. Braun-Blanquetia 46:103–106
2. Asensi A, Díez-Garretas B, Comino O (2017) Cartografìa y evaluación de la vegetación actual
en el Parque Natural Cabo de Gata-Nìjar (Almerìa, Espana). Collana Natura Aree Protette
Camerino 32:62–78
3. Beck S, Liberman Cruz M, Pedrotti F, Venanzoni R (1990) Estudio de las formaciones
leñosas de un valle interandino de Bolivia como base para la recuperación de áreas
degradadas. Resumenes V Congreso Americano de Botanica (Palacio de las Convenciones,
Ciudad de la Habana, 24–29 de junio de 1990). La Habana, p 122
4. Beck S, Liberman Cruz M, Pedrotti F, Venanzoni R (1992) Estado actual de los bosques en la
cuenca del Río Camacho (Departamento de Tarija, Bolivia). Studi Geologici Camerti vol
spec, pp 41–61
124 F. Pedrotti

5. Beck S, Paniagua Zambrana N, Yevara GarÁte M, Liberman M (2001) La vegetación y uso

de la tierra del altiplano de los valles en el oeste del Departamento de Tarija, Bolivia. In:
Beck S, Paniagua N, Preston D (eds), Historia, ambiente y sociedad en Tarija, Bolivia. La
Paz, Instituto de Ecología, UMSA – Leeds, School of Geography, University of Leeds,
pp 47–93
6. Bertin J (1967) Sémiologie graphique: les diagrammes, les réseaux, les cartes. París
Gauthier-Villars
7. Bertin J (1977) La graphique et le traitement graphique de l'information. París Flammarion
8. Bioret F, Estève R, Sturbois A (2009) Dictionnaire de la protection de la nature Rennes
Presses Universitaires
9. Borza A, Boşcaiu N (1985) Introducere in studiul covoruli vegetal Bucarest, Acad Rep Soc
Române
10. Carboni M, Carranza ML, Acosta A (2009) Assessing conservation status on coastal dunes: a
multiscale approach. Landscape Urban Plan 91:17–25
11. Cárdenas M (1969) Disertaciones botánicas y amenidades biológicas. Cochabamba Los
amigos del libro
12. Collot DD (1978–1980) Les Macrophytes de quelques lacs andins (Lacs Titicaca, Lac Poopó,
Lac de la vallée d’Hichu Khala et d’Ovejhuyo). Paris ORSTOM
13. Collot DD (1982) Mapa de vegetación de la bahia de Puno. Ecología En Bolivia 2:49–63
14. Collot DD, Koriyama F, Garcia E (1983) Répartitions, biomasses et productions des
macrophytes du Lac Titicaca. Rev Hydrobiol Trop 16(3):241–261
15. Corrado M (1884) El colegio franciscano de Tarija y sus misiones. Noticias historicas,
Quaracchi (Florenze) tipografia del Colegio San Buena Ventura
16. Davis FW, Stine PA, Stoms DM (1994) Distribution and conservation status of coastal sage
scrub in southwestern California. J Veg Sci 5:743–756
17. Del Arco Aguilar M-J, Gonzáles Gonzáles R, Garzón-Machado V, Pizarro-Hernández B
(2010) Actual and potential natural vegetation on the Canary Islands and its conservation
status. Biodivers Conserv 19:3089–3140
18. Domic AI, Palabral-Aguiklera N, Gomez MI, Hurtado R, Ortuno N, Liberman M (2017)
Polylepis incarum (Rosaceae) una especie en peligro critico en Bolivia: propuesta de
reclasificación en base al área de ocupación y estructura poblacional. Ecologia En Bolivía 52
(2):116–131
19. Faliński JB (1986) Vegetation dynamics in temperate lowland forests. Ecological studies in
Bialowieza forest. Geobotany 8:1–537
20. Faliński JB, Pedrotti F (1990) The vegetation and dynamical tendencies in the vegetation of
Bosco Quarto, Promontorio del Gargano, Italy (maps at scale 1:10,000 and a comment on the
maps). Braun-Blanquetia 5:1–31
21. Fjeldså J, Kessler M (2004) Conservación de la biodiversidad de los bosques de Polylepis de
las tierras altas de Bolivia. Santa Cruz de la Sierra Editorial FaN
22. Géhu JM (2006) Dictionnaire de sociologie et synécologie végétales. Stuttgart Cramer
23. Kessler M, Schmidt-Lebuhn AN (2006) Taxonomical and distributional notes on Polylepis
(Rosaceae). Org Divers Evol 6, Electric Suppl 1:1–10
24. Ivan D (1979) Fitocenologie şi vegetaţia Republicii socialiste România. Bucarest ed didact
pedagogica
25. Jewitt D (2011) Conservation target and status for vegetation types. Ecosystem ecologist
Ezemvelo KZN wildlife, Cascades South Africa
26. Liberman Cruz M (1986) Microclima y distribución de Polylepis tarapacana en el Parque
Nacional del Nevado Sajama, Bolivia. Doc Phytosoc X(ll):235–272
27. Cruz L (1987) Uso de la tierra en el altiplano norte de Bolivia como base por la evaluación del
impacto ambiental de un proyecto de desarrollo rural. Riv Agricolt Subtrop Trop 81(1–
2):207–235
28. Liberman Cruz M, Beck S, Pedrotti F, Venanzoni R (1986a) La vegetación de la Isla del Sol
en el Lago Titicaca. Resumenes IV Congreso Latino-Americano de Botanica (Medellin,
Colombia, 1986). Medellin, p 122
Essays on Geobotanic Mapping in the Andes of Bolivia ... 125

29. Liberman Cruz M, Gafta D, Pedrotti F (1997) Estructura de la población de Polylepis

tarapacana en el Nevado Sajama, Bolivia. Desarrollo sostenible de ecosistemas de montaña:
manejo de áreas frágiles en los Andes, pp 59–70. La Paz the United Nations University
30. Liberman Cruz M, Pedrotti F (2006) Woody formations in a mesothermic valley of Tarija
Province, Bolivia. In: Gafta D, Akeroyd J (eds) Nature conservation. Concepts and practice,
pp 75–86. Springer, Heidelberg
31. Liberman Cruz M, Pedrotti F, Gafta D (1995a) Los camelidos del Parque Nacional Sajama,
Bolivia: distribución, pastos y problemas ambientales. In: 2nd European symposium on South
American Camelids (September, 1995). Göttingen – Camerino, p 82
32. Liberman Cruz M, Pedrotti F, Molinillo M (1995b) Aspectos naturales y culturales del
pastoralismo de alpacas en una region altoandina de Bolivia. In: 2nd European symposium on
South American Camelids (September, 1995). Göttingen – Camerino, p 75
33. Liberman Cruz M, Pedrotti F, Venanzoni R (1986) La vegetación de la Isla del Sol en el lago
Titicaca – Bolivia (Informe preliminar). Giorn Bot Ital. 120(suppl 2):159
34. Liberman Cruz M, Pedrotti F, Venanzoni R (1986c) Mapa de la vegetación de la Isla del Sol
(Lago Titicaca - Bolivia). Camerino, Dipartimento Botanica Ecologia
35. Liberman Cruz M, Pedrotti F, Venanzoni R (1988) La vegetación de la Isla del Sol en el lago
Titicaca – Bolivia. Ecologia En Bolivia 11:69–70
36. Liberman Cruz M, Pedrotti F, Venanzoni R (1991) La vegetación flotante (Clase Lemnetea)
del Lago Uru-Uru, Depto. Oruro, Bolivia. Ecologia en Bolivia 17:47–51
37. Liberman Cruz M, Pedrotti F, Venanzoni R (1995) La Isla del Sol en el Lago Titicaca
(Bolivia). Descripción de la vegetación. In: II international symposium on sustainable
mountain development: managing fragile ecosystems in the Andes (Huarina - Bolivia), 2–12
April 1995, pp 71–78
38. Liberman Cruz M, Pedrotti F (2003) Deforestation and desertification in Camacho river basin
(Bolivia). Abstracts 46th Symposium. I.A.V.S. (8–14 June 2003, Napoli, Italia). Camerino,
Università degli Studi, p 143
39. Liberman Cruz M, Gafta D, Pedrotti F (1995) La estructura de la población de Polylepis
tarapacana en el Parque Nacional del Nevado Sajama, Bolivia. In: II international symposium
on sustainable mountain development: managing fragile ecosystems in the Andes (Huarina -
Bolivia), 2–12 April 1995, pp 17–32
40. Martinelli M (2014) Mapas da geografia e cartografia temática. São Paulo Contexto
41. Martínez O, Villarte F (2009) Estructura dasométrica de las plantas de un parche de Polylepis
incarum besseri y avifauna asociada en la Isla del Sol (Lago Titicaca, La Paz – Bolivia).
Ecología En Bolivia 44(1):36–49
42. Miles L, Newton AC, DeFries RS, Ravilious C, May I, Blyth S, Kapos V, Gordon JE (2006)
A global overview of the conservation status of tropical dry forests. J Biogeogr 33:491–505
43. Mingo M (1981) Historia de las Misiones Franciscanas de Tarija entre los Chiriguanos. Tarija
Universidad Juan Misael Caracho
44. Navarro G (1997) Contribución a la clasificación ecológica y florística de los bosques de
Bolivia. Rev Bol Ecol 2:3–37
45. Navarro G, Maldonado M (2006) Geografía ecológica de Bolivia. Cochabamba Centro
Ecología Simon I Patiño
46. Navarro G, Molina J, De la Barra N (2005) Classification of the high-Andean Polylepis
forests in Bolivia. Plant Ecol 176:113–130
47. Pavillard J (1935) Elements de sociologie végétale (Phytosociologie) Paris Hermann et C
(new edition 2019 Paris Hermann et C by Bioret F and Pedrotti F)
48. Pedrotti F (2013) Plant and vegetation mapping. Springer, Heidelberg
49. Pedrotti F (2013) Vegetation map of the Isla del Sol. In: Plant and vegetation mapping,
pp 194–198. Springer, Heidelberg
50. Pedrotti F, Venanzoni R (1986) Observaciones fitogeográficas sobre Puya raimondii en
Bolivia. Doc Phytosoc X(II):183–288
51. Pedrotti F, Venanzoni R (1990) Ricerche geobotaniche in Bolivia. Inform Bot Ital 22(3):281–
286
126 F. Pedrotti

52. Pedrotti F, Liberman CM, Venanzoni R (1988) Le associazioni della classe Lemnetea del lago
Titicaca (Bolivia). Riv Idrobiol 27(2–3):377–388
53. Pedrotti F, Venanzoni R, Suarez Tapia E (1988) Comunidades vegetales del Valle de
Capinota (Cochabamba - Bolivia). Ecología En Bolivia 11:25–45
54. Pedrotti F, Suarez Tapia E, Venanzoni R, Crespo Callaù V (1986) Las comunidades vegetales
de un valle mesotermico del interior de los Andes en el Departamento de Cochabamba -
Bolivia. Giorn Bot Ital 120(suppl 2):162
55. Pinna MS, Gogoni D, Fenu G, Bacchetta G (2015) The conservation status and anthropogenic
impacts assesments of mediterranean coastal dunes. Estuar Coast Shelf Sci 167:25–31
56. Quintanilla Aguirre J (1986) Sinopsis de las características fisico-químicas del lago Titicaca y
Poopó (Bolivia). Riv Idrobiol 24(3):87–139
57. Ribeira MO, Liberman M (2006) Parque Nacional Sajama. In: El uso de la tierra y los
recursos de la biodiversidad en la áreas protegidas de Bolivia. La Paz, Servicio Nacional
Áreas Protegidas SERNaP – Proyecto Gef – ii Banco Mundial, pp 31–49
58. Rivas-Martínez S (1987) Memoria del mapa de series de vegetaciòn de España. ICONA
Madrid
59. Slater FM, Curry P, Chadwell C (1987) A practical approach to the evaluation of the
conservation status of vegetation in river corridors in Wales. Biol Cons 40:53–68
60. Squier EG (1974) Un viaje por tierras incaicas. Crónica de una expedición arqueológica
(1863–1865). Cochabamba Los amigos del libro
Using Geobotanical Tools
to Operationalize Mapping
and Assessment of Ecosystem Services
(MAES) in Southern Portugal

S. Mesquita, J. Capelo, I. Gama, C. Marta-Pedroso, M. Reis,

and T. Domingos

Abstract As Action 5 of the EU Biodiversity Strategy for 2020 called on EU

member states to map and assess ecosystems and evaluate their services (MAES) in
national territories, a pilot-project was created to set up a straightforward method
for survey and evaluation of ecosystem types in the south of Portugal. The pro-
ject’s primary aim was to produce ecosystem maps wherefrom spatial estimations
of ecosystem services could be drawn, as well as biodiversity-related indicators of
ecosystem condition. The resulting ecosystem typology and map were also used by
others to estimate supporting, provisioning and regulating services.

M. Reis—Died 10 August 2017.

S. Mesquita (&)
CEABN - InBio, Instituto Superior de Agronomia, Universidade de Lisboa, Tapada da Ajuda,
1349-017 Lisbon, Portugal
J. Capelo
Instituto Nacional de Investigação Agrária e Veterinária, Herbarium, Av. da República,
Quinta do Marquês, 2784-505 Oeiras, Portugal
e-mail: [email protected]
J. Capelo
LEAF Research Centre – Linking Landscape, Environment, Agriculture and Food, University
of Lisbon, Tapada da Ajuda, 1349-017 Lisbon, Portugal
I. Gama
Terraprima – Serviços Ambientais, Quinta da França, Caria, 6200-710 Teixoso, Portugal
e-mail: [email protected]
C. Marta-Pedroso
T. Domingos
MARETEC – Marine, Environment and Technology Centre, Instituto Superior Técnico,
Universidade de Lisboa, 1049-001 Lisbon, Portugal
e-mail: [email protected]
T. Domingos
e-mail: [email protected]
M. Reis
CONEXA - Tecnologias e Sistemas de Informação, 2890-010 Alcochete, Portugal

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 127
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_7
128 S. Mesquita et al.

The proposed typology distinguishes 12 main ecosystem types based on the

higher EUNIS Habitat Classification levels. EUNIS types were obtained by cor-
responding a Vegetation Series Map (Capelo et al. in Phytocoenologia 37:399–415,
2007) to land-cover/land-use datasets (COS’07—land cover map with 192 classes
and MMU = 1 ha), and refined by using elevation, geology and satellite imagery
(NDWI) as ancillary information. For most units, correspondence between seral
stages in vegetation series and land use types allowed overall one-to-one identifi-
cation of candidate groups of the corresponding syntaxa as EUNIS types.
Auxiliary NDWI, geology and elevation allowed discriminating intrazonal special
habitats such as mesic and dry grasslands. The ecological value of plant commu-
nities and phytodiversity were modelled as relevant contributions for the assessment
of ecosystem condition.

Keywords Ecosystem services
Vegetation series
Mapping
Remote sensing

Portugal

1 Introduction

Action 5 of the EU Biodiversity Strategy for 2020 calls on the Member States to
Map and Assess Ecosystems and their Services (MAES) in their national territories
with the European Commission’s assistance.
To address this call for mainland Portugal, a pilot-project was developed for
mapping agricultural, agroforest and forest ecosystems and their services in the
South of Portugal (Alentejo), using the best available information and resources to
bring forth and test a method to accomplish the task. The project’s primary aim was
to produce an ecosystem map for the area wherefrom spatial estimations of
ecosystem services could be drawn, as well as biodiversity-related indicators of
ecosystem condition. The resulting ecosystem typology and map could also be used
as a base for estimates of supporting, provisioning and regulating services.
Establishing a coherent ecosystem typology is of paramount importance to
MAES. A necessary condition for MAES is that mapped ecosystem types should be
unambiguous, i.e. clearly spatially circumscribed and expressing recurrent ecolog-
ical identity as to composition, structure and function. This condition implies that
ecosystem types may be recognizable as ecological entities per se, irrespective of
changes in utilitarian approaches such as MAES. Surely, operational ecosystem
attributes used for MAES should also be present or inferable from such types. In
practice, a consensus seems to exist in the current literature on how to obtain
ecosystem types: MAES may be primarily derived from land use [9, 31–33, 51].
According to these authors, a sufficient relationship between land-use and ecosystem
types will be obtainable in most circumstances. As this assertion is debatable, Banko
et al. [6] came to recommend ‘thematic refinement’ of land use types to approach
more detailed and coherent ecosystem types. To resolve this relationship as both
unique and informative, additional data must be integrated using ancillary datasets
Using Geobotanical Tools to Operationalize Mapping … 129

such as terrain variables, environmental regions, potential natural vegetation,

satellite imagery, soil and geology [6]. Furthermore, UN et al. [51] came to redis-
cover the definition of ‘biocoenosis’ by Karl Möbius (1877) in the context of MAES.
They so defined the ‘land-cover/ecosystem functional unit’ (LCEU) as a spatial unit,
along with its ecological content, that is characterized by a unique combination of
‘land-cover type, water resources, climate, altitude and soil type’ (…). And they
proceed: ‘it should be possible consistently to differentiate an LCEU from neigh-
bouring LCEUs based on differences in its ecosystem characteristics’. Apparently,
these important calls were subsequently ignored, as the thematic refinement of land
use data is seldom practised, and land-use types are straightforwardly taken as
ecosystem types. Here we argue that, in most cases, land-cover and ecosystem types
are not equivalent. Many of the available land-use classifications and maps are
significantly too coarse to be meaningful as ecosystem types, leading to overgen-
eralizations and significant information loss in subsequent MAES assessments.
The second set of properties relates to ecosystem condition and expresses, in the
MAES perspective, the real capacity of an ecosystem to produce services when
compared to the maximum best qualitative condition possible it could be on. As
an example, an ecologically mature forest can provide more and better
biodiversity-related services than a degraded forest, due to disturbance and sim-
plification of composition, structure and function in the latter case. UN et al. [51]
propose that a set of key features, such as water, soil, vegetation, biodiversity,
carbon, and nutrient flows, be evaluated in comparison to a reference condition. The
reference condition relates either to typical, well-conserved, ecologically mature or
optimal integration in landscape functions, sometimes including human activities
such as traditional agricultural systems. In general, an assessment of the degree of
human-induced disturbance, i.e. hemeroby, is taken as a surrogate of how deviant
from the maximum integrity state the ecosystem is. Other features as functional
stability or resilience could be considered, expressing the ecosystem’s recovery
capacity if human disturbance ceases, changes or reduces. Several proposals of how
to find a subset of key ecological indicators to act as a surrogate of overall
ecosystem condition exist. Bordt [9] recognizes that the indicators may be
ecosystem-specific, and so the challenge is to find a set of key characteristics for
each ecosystem type. Bordt [9] states that ‘ideally, one would have a general
measure of ecosystem condition that would capture the ongoing functioning of the
ecosystem with respect to its capacity to generate services’. In executing MAES,
few have attempted to go beyond land-use features. Most methodological proposals
can be exemplified by Maes et al. [31] that propose the use of straightforward
easy-to-measure indicators of ecosystem condition regarding vegetation: leaf-area
index, biomass, annual net productivity (mean annual biomass increment), species
richness or diversity for a given period—the accounting period—and in spatially
well-delimited unit: the EAU—Ecosystem Accounting Unit [51]. Maes et al. [31]
even advocate a more straightforward approach by solely assessing general phys-
iognomic types of ecosystems: forest, lakes, grassland, crops, wetlands etc. This
means not going beyond land-use features and not distinguishing between pro-
duction agrosystems and natural and semi-natural ecosystems.
130 S. Mesquita et al.

Conversely, as we find the latter approaches incapable of producing necessary and

sufficient ecosystem types and assessing their condition, we emphasize that both are
scientific problems for Ecology, namely for Vegetation Science [27, 48]. A wealth of
regional or global (European and even worldwide) ecosystem typologies have been
made available, namely the UICN ecosystem classification scheme [28], the EUNIS
habitat classification [20, 27], the floristic vegetation classification of Europe
(EUROVEG) of Mucina et al. (2016) and the crosswalk between the latter and EUNIS
habitats by Schaminée et al. [48]. The assessment of ecosystem condition, namely
expressing the degree of human disturbance, may follow the geobotanical concept of
Natural Potential Vegetation (NPV) [50] and the inherent concept of ‘vegetation
series’ [12, 43]. A vegetation series is the set of stages, including potential (NPV) and
subseral, that may be found along a vegetation succession occurring on a homoge-
neous biotope. For details on concept operationalization in ecosystem condition
assessment, please see Loidi [30] and our own proposal in ‘Materials and Methods’.
It should be noted that a habitat survey is often not available or even feasible, so
the thematic refinement of land-use types proposed by Banko et al. [6] will be the
only least minimum standard to establish an ecosystem typology for MAES. In the
latter, land use is used mostly as baseline spatial information for physiognomical
types.Subsequently, many other ecological information levels are specifically
associated: VNP, soil, terrain, satellite and floristic data, etc. Also, filtering and
reclassification of land use may be issued by using supplementary information.
And then, what indicators should be looked at as good proxies for ecosystem
services associated with biodiversity? As vegetation is ecologically dominant (more
than 90% of biomass) in most terrestrial ecosystems and relatively fixed in space,
the identification of habitat types resorts to vegetation features as a core proxy
indicator. Plant communities, within their phytosociological framework, have been
used since the early 1980s as indicators of ecosystem value in land use planning and
to support the definition of nature conservation policies [3, 5, 23]. Plant commu-
nities reflect the environmental variables of the ecosystems: they integrate and
summarize them in their floristic diversity, as well as the ecological relationships
between organisms in the ecosystem. In the context of MAES, it should be noted
that regulation & maintenance ecosystem services in the CICES classification [8]
depend ultimately on biodiversity and attempting to estimate it through plant
richness may be a sensible approach.
Phytodiversity comprises the plant component of an ecosystem’s biodiversity. It
is usually addressed using diversity indexes such as the Shannon Index, as sug-
gested in Maes et al. [31], or the Simpson Index [21]. However, the use of such
biodiversity or phytodiversity indicators is only possible when detailed information
is available for the mapped ecosystem units, which is rarely the case. Moreover,
these indices do not discriminate native species, exotic species, invasive species,
rare species, common species, etc., which results in an overvaluation of disturbed
ecosystems and an undervaluation of less diverse plant communities but where
endemic or rare species of high conservation value occur. Furthermore, we propose
a method of processing land-use data through several levels of floristic and vege-
tation information yielding an estimation of plant richness issuing from phytoso-
ciological plot databases (see Materials and Methods).
Using Geobotanical Tools to Operationalize Mapping … 131

In this work, we propose the assumption of the ecosystem’s plant component as a

proxy for its total diversity since it reflects the entire ecosystem’s complexity.
Furthermore, accounting for plant communities’ successional position in vegetation
series reflecting their successional maturity and phytodiversity, one should obtain
good proxies for ecosystem condition. We believe that the phytosociological
approach to plant communities, with all its accumulated body of knowledge, is a solid
methodological basis for solving ecosystem mapping and assessment of condition.

2 Materials and Methods

2.1 Study Area

The study area for the pilot-project is the NUTS II region Alentejo, one of the
Territorial Units for Statistics developed by Eurostat and employed in Portugal for
statistical purposes; it covers around 31 500 km2, about one-third of mainland
Portugal (Fig. 1).

2.2 Map Sources

Information available for the area encompassed maps and non-spatial information.
Cartographic data relevant for the task was mainly about land cover and abiotic
factors influencing vegetation development:
• Land cover map (COS’07, with 192 classes and MMU = 1 ha) [24];
• Vegetation Series map [15],
• Bioclimatic maps [34],
• Geological map [38],
• Distance to coastline, derived from the Official Portuguese Administrative Map1
(CAOP) (Aviso n.º 2298/2018, de 7 de março de 2018);
• NDWI (Normalized Difference Water Index), derived from satellite imagery
Landsat-8 RNIR and RSWIR [26].

2.3 Geobotanical and Phytosociological Data Sources

• Base sources for the geobotanical model for mainland Portugal:

1
https://www.dgterritorio.pt/cartografia_e_geodesia/cartografia/carta_administrativa_oficial_de_
portugal_caop_/.
132 S. Mesquita et al.

Fig. 1 Study area: NUTS II Alentejo, in Mainland Portugal

– Biogeographic model [18],

– Bioclimatic and geobotanical model of the Iberian Peninsula [41, 43, 46, 47],
– Syntaxa checklist [17],
– Vegetation Series model [15].
• Phytosociological relevés2 on-line databases:
– SIVIM (https://www.sivim.info/sivi/);
– European Vegetation Archive (EVA; https://euroveg.org/eva-database).
• Other relevant bibliography:
– Regional geobotanical models [1, 10, 11, 13, 14, 16, 17, 36, 37, 39, 40].

2.4 Ecosystem Mapping

The proposed typology of ecosystems distinguishes 12 main ecosystem types based

on the higher levels of the EUNIS Habitat Classification,3 as recommended by most

2
Lists of taxa present in a sample plot, each affected by an abundance value.
3
https://eunis.eea.europa.eu/.
Using Geobotanical Tools to Operationalize Mapping … 133

authors [48] for mapping ecosystems and their services in European countries.
EUNIS is a comprehensive Pan-European typology, including natural and artificial
habitats and common vegetation. It has a hierarchical structure, allowing the use of
the level which is more adequate to the detail of available information within each
main type. The EUNIS typology encompasses 10 habitat classes at the first hier-
archical level (plus one class of mixed habitats), subdivided into as many subclasses
as necessary, sometimes up to the eighth level. Moreover, it is being revised, and
crosswalks have been developed between level 3 EUNIS terrestrial habitat types
and vegetation syntaxa, which are well established for most of the Portuguese
native vegetation [17].
The adopted methodology was implemented in two separate steps, generally
following the approach proposed in Banko et al. [6]: first, a crosswalk between
EUNIS and COS’07 land cover was performed, second, thematic refinement was
carried out in a top-down approach. Since ecosystem maps were produced at a large
scale, specific problems arose due to increased thematic and spatial detail, so new
strategies and solutions were developed.

2.4.1 EUNIS – COS’07 Crosswalk

A first attempt to establish a formal relation between the COS’2007 land-use types
and UNIS habitat types of level 2 or higher was made to detect direct 1:1 corre-
spondence. This was successfully performed for 103 of the 142 agricultural,
agroforest and forest land-use types, covering 75.3% of the study area (see Table 1,
in Annexe I).
Although the main goal was to map ecosystems down to EUNIS level 2, it was
considered advantageous to descend in the hierarchy as much as possible within the
scope of each COS’07 class. Thanks to this conservative option in which available
information is used in all its detail, it was possible to reach levels 3 or 4 in most
cases, and sometimes even level 5. It is always possible to quickly revert any
EUNIS class to level 2 of this hierarchy, if necessary. This detail was only possible
due to the high detail of the COS’07 land use map, produced with a reference scale
of 1:25,000, based on the visual interpretation of orthorectified aerial images from
July to October of 2007, with the help of diverse ancillary information; it has a
positional accuracy greater than or equal to 5.5 m and an overall thematic accuracy
of 85.13% with a 2.00% error for a 95% confidence level.4

2.4.2 Thematic Refinement

The remaining 24.7% of the area corresponds to land use classes having a 1:n
relation to EUNIS habitat types. These situations were solved using ancillary

4
https://www.dgterritorio.pt/cartografia_e_geodesia/cartografia/cartografia_tematica.
134 S. Mesquita et al.

spatially explicit datasets, which provided the information needed for addressing
four distinct ecological issues.
2.4.2.1. Differentiation According to Natural Potential Vegetation
Natural potential vegetation, mapped as vegetation series [15], was used to separate
different types of native heathland, scrub and woodland.
The assumed paradigm for natural potential vegetation was that of
Rivas-Martinez [42] and Rivas-Martinez et al. [45], which reasons a univocal
correspondence between a uniform combination of bioclimate, lithology and bio-
geographical context and a set of plant communities that occur successively and
organize themselves in mosaics throughout the process of ecological succession.
NUTS II Alentejo vegetation series are listed in the Table 2 (Annexe II).
Knowing a place's vegetation series and the vegetal cover physiognomic type that
corresponds to its land-use at a particular moment, it is possible to determine the
dominant plant community. Thus, for areas of undetermined native heathland, scrub
and woodland, crossing current land use information with spatial explicit vegetation
series data, the dominant plant community in each site can be inferred. The
crosswalk between EUNIS types and COS’07 land use/Natural potential vegetation
types for native heathland, scrub and woodland is presented in Table 3 (Annex II).
The EUNIS typology is currently under review by the European Vegetation
Survey work group within the International Association for Vegetation Science
(IAVS, see https://euroveg.org/). Since native Mediterranean scrub, mostly included
in EUNIS class F, is still poorly discriminated in this typology (see [49], the most
updated document at the time available from this review was used as basis for a
possible crosswalk between the phytosociological typology for shrub communities
and EUNIS types [27], detailed information for each type is available at https://
forum.eionet.europa.eu/european-red-list-habitats/library/terrestrial-habitats/).
This task proved to be more complex than initially thought, requiring compu-
tational capacities beyond standard operator software packages. Therefore, specific
programming procedures were implemented for dealing with the great amount of
geographical data and topological complexity of operations for combining geo-
graphic information layers.
2.4.2.2. Differentiation According to Geology
Maritime pine and stone pine forests correspond to EUNIS level 2 class ‘G3:
Coniferous woodland’. However, the distinction between disclimactic pinewoods
of high natural value (G3.714, G3.7312, G4.B) and productive planted pine forests
(G3.F22, G4.F) is important, since these are distinct ecosystems, providing different
ecosystem services.
Using Geobotanical Tools to Operationalize Mapping … 135

The criteria for separating these two groups were created by Aguiar and Capelo
[2]: old undisturbed pinewoods occurring in dune systems, despite their possible
anthropic origin, mimic the floristic combination of the pinewoods from the
Atlantic period and can therefore be taken as native forests.
For the specific purpose of mapping EUNIS types, only geological substrate was
considered (as mapped in [38]), since there was no available information con-
cerning stands' age and undercover management. Therefore, all pine forests on
dunes and coastal paleodunes were considered natural forests (see Table 3 in
Annexe III).
2.4.2.3. Differentiation According to Distance to the Ocean
EUNIS classification separates coastal ecosystems from all others as high as the first
level. However, the available land-use maps distinguish such ecosystems only in
specific conditions, such as beaches, inter-tidal zones and marshes.
Thus, an efficient methodology was conceived to separate the coastal cliffs and
sparsely vegetated dunes from interior rocky outcrops and other sparsely vegetated
areas. After a systematic visual inspection of the coastline, it was decided to assume
as coastal all areas located less than 100 m from the coastline (see Table 5, in
Annexe IV). Distance to the coast-line was obtained and computed directly from
the Official Portuguese Administrative Map (CAOP), which records the state offi-
cial demarcation of Portugal's administrative districts and borders.
2.4.2.4. Differentiation According to NDWI
The EUNIS classification separates natural herbaceous communities at level 2 in
dry, mesic, wet and alpine grasslands, as well as woodland fringes and clearings,
inland salt steppes and sparsely wooded grasslands. Except for dehesas, delimited
separately in great detail in land use maps due to their economic importance in
Portugal, all grasslands were mapped as permanent pasture or natural herbaceous
vegetation (see Table 6, Annexe V).
Alpine grasslands and inland salt steppes are absent from the study area and
woodland fringes and clearings are too small to be represented in land use maps, so
grasslands included dry, mesic and wet subtypes (E1, E2 and E3), which should be
separated resorting to criteria used in the definition of these classes, such as tro-
phy degree, intra-annual variation of soil moisture and productivity (see Table 7,
Annexe V).
The distinction between mesic and dry grasslands would require fieldwork, since
these typologies can only be separated knowing the floristic composition of plant
communities. Likewise, agriculturally improved and reseeded pastures can only be
distinguished from other rainfed meadows conducting fieldwork or surveys to
farmers. So, it was not possible to map E2 meadows, whose areas are included in
the areas classified as E1.
E3 meadows in the study area are water-logged meadows and rush communities,
which occur in alluvial areas with high soil water levels throughout the Summer.
Satellite imagery obtained in the Summer was used to acquire information about
the water content of vegetation during the dry season and, therefore, to map those
136 S. Mesquita et al.

meadows which retain high levels of water during the Summer. All areas of natural
herbaceous vegetation located near water lines and with high water content were
selected as E3 grasslands.
The Normalized Difference Water Index (NDWI) was used to evaluate liquid
water content of vegetation, since this is a particularly effective index, less sensitive
to atmospheric effects than other vegetation indexes [22]. It is calculated as
(RNIR − RSWIR)/(RNIR + RSWIR), where RNIR represents the reflectance in a
near-infrared band and RSWIR is the reflectance in a middle-infrared band. Since
Landsat-8 imagery was used, RNIR and RSWIR correspond, respectively, to bands 5
(0.851–0.879 lm) and 6 (1.566–1.651 lm) [26].
Cut-off values of NDWI were empirically set for separating wet meadows from
dry meadows, using representative samples for each type previously known to the
team. All meadow polygons with an NDWI value greater than the estimated cut
value in most of its area were classified as E3 grasslands. A single cut-off value was
considered unsatisfactory for the whole territory. Bioclimatic variables [34] were
tested to separate the study area into two zones, corresponding to distinct areas
regarding the amount of water available to plants in the Summer months. For this
separation, Summer Compensated Ombrothermic Index (Ios3) maps were used.
This index is part of the “Worldwide Bioclimatic Classification System” and was
created by Rivas-Martínez [44] to account for the Summer dryness in
Mediterranean areas such as Alentejo, integrating information on Summer precip-
itation and temperature.5
Satellite images from 2013 were used because older images were of poor quality,
so changes in land use can be a source of error, namely from grassland to scrub.
Nevertheless, the results thus obtained were satisfactory.

2.5 Ecological Value of Plant Communities

Plant communities are highly visible and measurable and reflect the ecosys-
tem's environmental variables, therefore can be assumed as reliable proxies for their
diversity and condition [7, 29].
The ecological value of plant communities present in each type of ecosystem
was assessed using a methodology adapted from Loidi [30]. The original method
was developed in the context of detailed surveys, assuming the availability of
detailed floristic information about plant communities occurring in all evaluated
plots. Additionally, the method uses information about the general characteristics of
the phytosociological association to which each plant community can be ascer-
tained, as well as the region where it is located. It was later adapted to apply to less

5
Ios3 = (Pp3/Tp3), where Tp3 is the sum of the monthly average temperature of June, July and
August (in tenths of degrees Celsius) and Pp3 is the total rainfall of June, July and August (in mm).
Using Geobotanical Tools to Operationalize Mapping … 137

detailed studies, particularly when field surveys are not available, as is the present
case (e.g. [4, 35]).
‘Ecological value’ was achieved as the arithmetic mean of five assessed
parameters, all ranging from 1 to 5: naturalness, replaceability, threat, rarity and
hemeroby [4, 25, 30].
Naturalness: Expresses the intrinsic degree of human influence, reflecting, on the
one hand, floristic impoverishment due to disturbance and, on the other hand, the
degree to which the plant community depends on active management. The scale
used for evaluating this parameter is as follows:
1. Maximum disturbance: exotic and cultivated vegetation, agricultural areas,
agriculturally improved and reseeded pastures, productive forests.
2. Dehesas, natural pastureland.
3. Sparse scrub and brush (not in coastal dunes).
4. Coastal dune sparse scrub, heaths and meadows.
5. Minimal disturbance: dense native woodland, dense thickets and bushes, bare
rock.
Replaceability: Capability of a vegetation type to recover itself after destruction
by an external agent. It is very low in climacic communities, usually woodlands,
and increases along the vegetation types' ecological succession sequence. An
inverse scale of values was used to attribute a higher value to less resilient com-
munities, as follow:
1. Exotic vegetation, agricultural land, plantations.
2. Pioneering communities: natural herbaceous vegetation, brambles.
3. Scrub, forests with exotic species, pastures.
4. Thickets, dehesas, chestnut woodland, coastal dune scrub and heaths, bare rock.
5. Climacic vegetation: native woodlands.
Threat: Reflects the community's attractiveness or its biotope for destructive
human activities and is influenced by ease of access, presence of species of eco-
nomic value, population density, etc.; it also reflects the threat posed by ceasing of
active management in low-intensity agroforestry habitats. This criterion is highly
dependable on regional specificities, particularly socio-economic context, which
varies greatly throughout history. The assessment scale used was as follows:
1. Communities in biotopes of difficult to access, with no apparent economic gains
and low attractiveness, therefore less threatened.
2. Brambles, thickets, pastures with permanent crops.
3. Vegetation types highly dependent on active management: dehesas, natural
pastureland.
4. Vegetation types which include plants traditionally harvested for commercial
exploitation.
138 S. Mesquita et al.

5. Plant communities in biotopes of high economic value or high attractiveness and

easy access: beaches, coastal areas6 and areas close to urban centers.7
Rarity: This parameter was assessed using the typology established in Izco
(1998), which integrates three distinct attributes: the size of the community's dis-
tribution area, average community size, and community occurrence frequence
within its distribution area. Different combinations of the three types of rarity
defined by these attributes result in seven distinct rarity types, summarized in
Table 1 below.
For the pilot-project study area, geographical rarity was considered when the
potential area of the community is less than 1.5% of mainland Portugal (about
130,000 hectares); small size when the average size of polygons is less than 100 ha;
and low frequency of occurrence when the typology is present in the study area in
less than 100 polygons.
These rarity types were translated on a scale of 1 to 5 as follows:
Without vegetation or with exotic, cultivated or cosmopolitan types of
vegetation.
1. R1, R2.
2. R3.
3. R4, R5.
4. R6, R7.

Table 1 Rarity types, according to Izco (1998)

Small geographical Low frequency of Small size of plant
range occurrence communities
Not
rare
R1 ✓
R2 ✓
R3 ✓ ✓
R4 ✓
R5 ✓ ✓
R6 ✓ ✓
R7 ✓ ✓ ✓

6
Areas located at less than 2 km from coastline.
7
Areas located within a circle around urban centers with a radius equal to 2/3 of their population
(in meters), from the central point of its center, for all urban areas with more than 10 000
inhabitants.
Using Geobotanical Tools to Operationalize Mapping … 139

Hemeroby: A measure of the distance that separates a vegetation type from its
undisturbed state, namely concerning the proportion of synanthropic vegetation.
The sub-parameters used for estimating this indicator were defined for each group
of ecosystems in the territory according to the optimal characteristics of each,
resulting in the following scale:
1. Maximum disturbance within the community, with exotic and cultivated veg-
etation: crops, dehesa with irrigated or permanent crops.
2. Open forests with exotic vegetation, sparse scrub, sparse herbaceous vegetation,
pastures with permanent crops, bare rock.
3. Dense forests with exotic species, dense scrub, herbaceous vegetation, dehesa
with rainfed crops, beaches, coastal dune sparse scrub.
4. Open forests, rush communities, chestnut woods.
5. Little or undisturbed vegetation types: dense native woodlands with no exotic
species, dense scrub and thickets, dehesa with natural meadows.
The ecological value of plant communities for each mapped unit is the average
of these five parameters, rounded to the nearest integer value. This semi-
quantitative valuation should have the following interpretation (Table 2):

2.6 Phytodiversity

This indicator aims tomeasure the value of plant diversity occurring in a community
using available information and avoid the disadvantages of diversity indexes com-
monly used. The proposed methodology assumes that EUNIS vegetation types can
be further detailed using information from Vegetation Series maps; then, knowing
the plant community which occurs at each location, it is possible to estimate its
average or characteristic floristic composition by consulting a phytosociological
table of relevés of similar communities, which reflects its floristic richness and
originality, as well as the presence of endemic or threatened plant species.
The adopted approach consisted primarily of establishing a correspondence
between each type of EUNIS ecosystem and the corresponding phytosociological
unit, using the information provided by a Potential Natural Vegetation map [15].
For each vegetation series, correspondence between EUNIS vegetation types and
the most common successional stages (woodlands, thickets, scrub, perennial

Table 2 Interpretation of ecological value

Ecological value of plant communities Range Interpretation
1 [1, 1.5[ Very low ecological value
2 [1.5, 2.5[ Low ecological value
3 [2.5, 3.5[ Average ecological value
4 [3.5, 4.5[ High ecological value
5 [4.5, 5] Very high ecological value
140 S. Mesquita et al.

meadows) was established using phytosociological criteria and expert knowledge,

including some types of azonal vegetation, such as riparian, chasmophytic and
ruderal.
Next, an evaluation of the phytodiversity for each phytosociological unit was
conducted, using its floristic composition from available bibliography and infor-
mation about the distribution of flora species produced in the framework of the
report for 2013 under Article 17 of the Habitats Directive, on the conservation
status of Natura 2000 Network in Portugal. Around 3,500 phytosociological relevés
were analysed. For calculating phytodiversity in each plant community, estimates
were made of mean values of: i) protected plant species; ii) other endemic species;
iii) other rare plant species; iv) characteristic species. The final value of phytodi-
versity is a weighted average of these parameters using, respectively, values of 4, 3,
2 and 1. Estimations were carried out as follows:
1. For each phytosociological unit, a significant set of phytosociological relevés
was compiled from phytosociological literature, online databases, and the
authors' archives, including, whenever possible, the protologue (original table)
of the association.
2. Within the set of phytosociological relevés obtained for each vegetation unit, all
taxa legally protected in Portugal (including those in Annexes II and IV of
Directive 92/43/EEC) were counted. This number corresponds not to protected
taxa potentially occurring in the habitat type, but to those that were effectively
observed in the plant communities floristically described in the relevés. The total
number of taxa endemic to Portugal but not legally protected was also obtained,
and the total number of rare species (not endemic or legally protected).
3. Finally, the average number of characteristic species was estimated. These cor-
respond to a statistically reliable combination of diagnostic species or bioindi-
cators that uniquely identify the plant community (type of vegetation) and its
environmental conditions. For this purpose, companion species in the relevés
were discarded, since they are interpreted as corresponding to a set of plants
occasionally present, with an ecological optimum in other types of vegetation, in
adjacent biotopes or mosaics (either due to catenal or successional phenomena).
The average number of characteristic species for a vegetation type representats its
average floristic richness and, indirectly, of its ecological integrity.
Whenever one particular stage of a vegetation series corresponds not to a single
community but several territorial variants, the total set of relevés of these com-
munities was considered. In such cases, the maximum value was considered for
parameters i), ii) and iii) and its arithmetic mean in the case of parameter iv).
For major azonal vegetation types, analogous procedures were implemented.
However, as these vegetation types correspond to permanent complexes along
topographically explicit environmental gradients without serial dynamics, the
number of characteristic species considered was the sum of characteristics of all
plant associations present. Likewise, parameters i) to iii) were calculated as the
maximum values of all plant associations involved.
Using Geobotanical Tools to Operationalize Mapping … 141

Table 3 Interpretation of phytodiversity value

Phytodiversity value of plant communities Range Interpretation
0 0 No value
1 ]0, 5,5] Very low value
2 ]5,5, 16] Low value
3 ]16, 26] Average value
4 ]26, 49] High value
5 ]49, 111] Very high value

For semi-natural habitat types resulting from anthropic disturbance, pioneer

vegetation types and degraded or sparse vegetation, relevés corresponding to the
pioneer vegetation were used.
Phytodiversity value was calculated as the weighted arithmetic mean of these
parameters, assigning different weights to each one: 1 for the total number of
characteristic species, 4 for protected species, 2 for other endemic species and 3 for
other rare plants. In agricultural and forestry ecosystems, where phytodiversity is
unknown but generally expected low, a value corresponding to 5% of the param-
eter’s range was assumed (5.5). Finally, the value 0 was assigned to areas without
vegetation.
Phytodiversity for each mapped unit is the weighted average of these five
parameters, rounded to the nearest integer value. This semi-quantitative valuation
obtained should have the following interpretation (Table 3):
Since the dominant vegetation is zonal in most of the territory, and accounting
for the major types of azonal vegetation, estimates of phytodiversity produced have
a high degree of reliability and translate, in relative terms, the floristic and phy-
tochoenotic importance of mapped EUNIS habitats.

3 Results

Ecosystem Map
An Ecosystems map of Alentejo was produced, with great detail, both thematic
(EUNIS types mapped at levels 2 to 6) and spatially (scale 1:25,000), as presented
in Fig. 2.
This map reproduces the thematic heterogeneity of land-use types in COS’07,
with great detail in forest and agricultural typologies contrasting with very low
detail in herbaceous and scrub vegetation types.
All 68 forest types are directly attributable to level 2 EUNIS types, but greater
discrimination was developed within these types, using ancillary information.
19 EUNIS classes were obtained, from levels 2 to 6, most at levels 3 and 4.
142 S. Mesquita et al.

Fig. 2 Ecosystem map

Likewise, 64 agricultural land-use types were crosswalked to 14 EUNIS

ecosystem types at levels 2 to 4. Correspondence between grasslands in land use
map and EUNIS typology at level 2 could only be achieved using ancillary data.
Finally, a correspondence between the remaining 10 land-use types—encom-
passing scrub, sclerophyle and herbaceous vegetation, as well as sparsely vegetated
surfaces and bare rock—and EUNIS ecosystem types was achieved using ancillary
information and 13 classes were obtained, at levels 2 and 3.
Thematic detail in the produced Ecosystems map is very satisfactory, with
spatial detail appropriate to the 1:25,000 scale. This map is adequate to support
further analysis, since mapped units are complex entities containing a high amount
of information, helpful in evaluating ecosystem services and for landscape planning
and management. Thematic detail in some ecosystem types is of paramount
importance for assessing ecosystem services related to biodiversity. Since EUNIS is
Using Geobotanical Tools to Operationalize Mapping … 143

a hierarchical classification system, working at less detailed levels is always pos-

sible and easy to perform.
The production of this map in a relatively short time was possible due to the
accomplished methodological robustness and the good operational integration of
ecological, GIS and computational expertise. This map was produced using an
explicit, replicable methodology, allowing updates, at any time, if in the presence of
updated or additional georeferenced information and the necessary processing
capacity. This results in a significant ability to update information and correct it if
new data is available, incorporating minor methodological rectifications if necessary.
Ecological Value of Plant Communities
A map of the ecological value of plant communities was produced for the study
area, using the described methodology (Fig. 3). The chart in Fig. 4 presents the
contribution of each type of ecosystem to the ecological value map, considering
EUNIS classification at level 2. Some patterns stand out in this analysis:
i. High ecological value corresponds to protected and successionally evolved
vegetation stands in dunes, paleodunes and coastal cliffs; and relict woods, of
pristine character.

Fig. 3 Map of ecological value of plant communities

144 S. Mesquita et al.

B1
B3
E1
E3
E7
F3
F4
F5
F6
FB
G1
G2
G3
G4
G5
H3
H5
I1
X07

0 1000 2000 3000 4000 5000 6000 7000

(x1000 ha)
1 2 3 4 5

Fig. 4 Areas of ecosystem types at level 2 in the study area and their ecological value (thousands
of ha)

ii. Low ecological values occur in agricultural areas, particularly in Southern

Alentejo.
iii. Intermediate ecological value is associated with areas of dehesa, with
low-intensity agricultural practices, usually in poor soils or steep slopes.
iv. Areas with cork oak woods as climacic vegetation differ considerably from
those with holm oak climacic woods, having higher ecological value due to
occupying a richer biotope—with higher precipitation and dynamical relief—
facilitating the development of richer vegetation mosaics (i.e., with higher
beta-diversity).
‘Ecological value’ integrates information on plant community structural quality,
phytocoenotic integrity and successional maturity. Geobotanic models reflect the
essence of this variability and the variations in functional diversity with an impact
on the community's capacity to provide ecosystem services.
Phytodiversity
Applying the described methodology to the study area resulted in a map of phy-
todiversity (Fig. 5). Total variability of phytodiversity is presented in 5 classes,
defined as ranges of increasing amplitude. Figure 6 presents a chart showing, for
each level 2 EUNIS ecosystem, the areas corresponding to each phytodiversity
class. The most interesting patterns in this analysis are:
i. The lowest phytodiversity values occur in agricultural ecosystems on siliceous
soils with annual crops, even if integrated with dehesas. The same agricultural
systems on limestone soils tend to have higher phytodiversity.
Using Geobotanical Tools to Operationalize Mapping … 145

Fig. 5 Map of phytodiversity

ii. Dehesas have medium to high phytodiversity, primarily when associated with
low-intensity management and steep slopes that allow for a richer vegetation
mosaic and flora.
iii. Higher phytodiversity values occur in agricultural or forest ecosystems on
sandy soils (arenosols and podzols on paleodunes in river Sado water basin),
higher than those of the same land-use typology on siliceous soils (those
derived from schist, greywacke, granite and syenite). In this particular case,
biogeography is crucial, for this unit has a considerable pool of regional
endemisms.
iv. High phytodiversity values do not necessarily correlate to known distribution
areas of rare and protected plant species. These plants often occupy very
specific biotopes with low biodiversity, to which they are narrowly special-
ized, such as active dunes and coastal cliffs.
146 S. Mesquita et al.

B1
B3
E1
E3
E7
F3
F4
F5
F6
FB
G1
G2
G3
G4
H3
H5
I1
X07

0 1000 2000 3000 4000 5000 6000 7000

(x1000 ha)
1 2 3 4 5

Fig. 6 Areas of ecosystem types at level 2 in the study area and their phytodiversity (thousands of
ha; zero values not shown)

v. Areas of successively mature vegetation, complex mosaics or annual pastures

on limestone have high phytodiversity values due to the soils' mesotrophic
character. However, most limestone areas are occupied by intensive agricul-
ture, with low phytodiversity values.
‘Phytodiversity’ was estimated using an entirely original methodology devel-
oped specifically for this task. It is based on the phytosociological concept of
characteristic species—a set of plants that have their ecological optimum in a given
community—which can be assumed as proxies for global phytodiversity. Species
that, although present, have a stochastic nature are excluded since there is no point
in considering them as members of the plant community.

4 Discussion and Conclusions

The EUNIS habitats typology is particularly suitable for ecosystem mapping in the
context of ecosystem services assessment. It is a comprehensive hierarchical
pan-European system established for harmonising habitat identification and
description across Europe, covering all types of habitats from natural to artificial.
Using Geobotanical Tools to Operationalize Mapping … 147

‘Habitat type’ is defined for the purposes of EUNIS classification as ‘Plant and
animal communities as the characterising elements of the biotic environment,
together with abiotic factors operating together at a particular scale.’8
The EUNIS classification is suitable for this task thanks to other characteristics
worth mentioning:
i. The adopted definition of ‘habitat type’ is, in fact, the standard definition of
ecosystem.
ii. EUNIS description of types indicates a vegetation type, the abiotic conditions
in which it thrives and associated succession phenomena. This is no surprise,
since plant communities are actually most of an ecosystem’s biomass and,
therefore, its most easily assessed and mappable component.
iii. Although covering all types of habitats, from natural to artificial, this typology
is focused on defining criteria for the identification and classification of natural
ecosystems. These criteria are fundamentally ecological: dominant physiog-
nomy, biotope characteristics, biogeographic units, bioclimatic limits.
The methodology used for ecosystem mapping was based on interpreting
land-use units as territorial units of ecological succession, an innovative approach in
Portugal although in line with procedures developed in other EU countries [19].
The core methodological innovation of the approach is the use of phytosocio-
logical typologies as a basis for interpreting and establishing a crosswalk between
vegetation types and EUNIS habitat types. A similar procedure, although using the
Habitats Directive Annexe I typology, was used in France (CarHAB9) and Spain
(Hábitat (Directiva 92/43/CEE)10). It is a ‘rule-based’ modelling procedure that
integrates successional phytosociological models and land use data, incorporating
ancillary information of potential natural vegetation, water indexes applied to
satellite imagery and geology.
EUNIS habitat type mapping of Alentejo at levels 2 to 5 and a detailed 1:25,000
scale in a relatively short time was possible thanks to methodological robustness
and optimized integration of a team of ecology, GIS and computer science experts.
The methodology may efficiently be applied to all mainland Portugal and adapted to
other areas, depending on the availability of spatially explicit data.
Ecological value integrates indicators of structural quality, phytocoenotic
integrity and successional maturity of plant communities and reflects the pattern of
territorial representativeness, or rarity, of the plant communities.
This parameter is particularly useful in managing protected areas, for instance,
conveying information beyond the mere presence or absence of rare species, such
as functional and structural ecological values, which are not evident when con-
ventional biodiversity indexes and parameters are considered. Ecological value is a

8
https://eunis.eea.europa.eu/about.
9
https://www.cbn-alpin.fr/actions/habitats/carhab.html.
10
https://www.miteco.gob.es/es/biodiversidad/servicios/banco-datos-naturaleza/informacion-
disponible/habitat.aspx.
148 S. Mesquita et al.

relevant indicator for detecting ecosystems of high ecological value but low bio-
diversity, such as native cork oak woodlands.
This is an efficient methodology, easy to interpret and extrapolate. Its main
difficulty lies in adapting the parameter ranges defined by Loidi [30] and Izco [25]
to the available information and different scales of approach. However, both pub-
lications are quite explicit in the definition of sub -parameters so, whenever a
phytosociological successional model is available, the adaptation of this method-
ology to other study-cases is entirely possible.
Phytodiversity estimates encompass two distinct components of plant diversity.
On the one hand, intrinsic species richness of ecosystems is estimated as a function
of their set of characteristic species (i.e., with their ecological optimum in this
ecosystem), reflecting its ability to provide ecological niches for a variable number
of taxa. This variability is a function of the ecosystem's structural complexity, its
trophy, bioclimatic regime, etc. and of the regional pool of species.
On the other hand, phytodiversity also reflects the ecosystem’s richness in
regional endemisms and legally protected plants, which are extrinsic characteristics,
representing a political and societal perspective on the plants' conservation value.
An endemism is defined in relation to a political frontier and is not, therefore, a
strictly biogeographical characteristic, although, in some cases, these two concepts
may coincide.
Together, these two components may well express the habitat’s ability to provide
ecosystem services associated with biodiversity.
As stated before, an original methodology was developed specifically for this
task, based on the phytosociological concept of characteristic species, which can be
used as proxies for the global plant diversity of an ecosystem.
The main difficulty in applying the methodology is choosing which vegetation
unit (i.e. plant association) corresponds to each type of vegetation, when dealing
with low naturality ecosystems. It may be a time-consuming task, since it requires
the compilation of a large number of relevés, which must necessarily be repre-
sentative of all types of vegetation present.
The quality of estimated values can only be assessed by qualitatively and sub-
jectively comparing the map produced with available information on species dis-
tribution, considering the various types of errors that may be associated with them.
Globally, the results achieved were considered very satisfactory. Again, this is an
objective methodology that, as such, can be easily replicated and extended to other
parts of the country and to other ecosystems.

Acknowledgements The authors wish to express their gratitude to all team members of the Pilot
Project “Mapping and Assessment of Ecosystem Services” at IST, for constant support, incentive,
knowledge sharing and excellent companionship. Finally, we want to remember the departed
Marco Reis, whose memory we all cherish, for his irreplaceable in computer science expertise and
for having been essential to completing the project.
This work was carried out under the National Authority for Biodiversity and Forest
Conservation (ICNF) financial support under contract AD 288/2014/ICNF/SEDE.
Using Geobotanical Tools to Operationalize Mapping … 149

Annex I

Table 1 Crosswalk EUNIS – COS’07

COS’07 EUNIS
2.1.1.01.1 Rainfed crops I1.3 Arable land with unmixed crops
grown by low-intensity agricultural
methods
2.1.1.02.1 Greenhouses and Nurseries J2.43 Greenhouses
2.1.2.01.1 Irrigated crops I1 Arable land and market gardens
2.1.3.01.1 Rice fields I1.4 Inundated or inundatable croplands,
including rice fields
2.2.1.01.1 Vineyards FB.4 Vineyards
2.2.1.02.1 Orchards with vineyards FB.4 Vineyards
2.2.1.03.1 Vineyards with olive groves FB.4 Vineyards
2.2.2.01.1 Fresh fruit orchards G1.D4 Fruit orchards
2.2.2.01.2 Almond orchards G1.D3 Prunus amygdalus groves
2.2.2.01.3 Orchards of chestnut trees G1.D1 Castanea sativa plantations
2.2.2.01.5 Orchards of citrus fruits G2.92 Citrus orchards
2.2.2.01.6 Other orchards G1.D5 Other high-stem orchards
2.2.2.02.1 Orchards of fresh fruit with G1.D4 Fruit orchards
vineyards
2.2.2.02.6 Other orchards with vineyards G1.D5 Other high-stem orchards
2.2.2.03.1 Fresh fruit orchards with olive G1.D4 Fruit orchards
trees
2.2.2.03.3 Orchards of chestnut with olive G1.D1 Castanea sativa plantations
trees
2.2.2.03.5 Orchards of citrus with olive trees G2.92 Citrus orchards
2.2.2.03.6 Other orchards with olive groves G1.D5 Other high-stem orchards
2.2.3.01.1 Olive groves G2.91 Olea europaea groves
2.2.3.02.1 Olive groves with vineyards G2.91 Olea europaea groves
2.2.3.03.1 Olive groves with other fruit trees G2.91 Olea europaea groves
2.4.1.01.1 Rainfed crops associated with I1.3 Arable land with unmixed crops
vines grown by low-intensity agricultural
methods
2.4.1.01.2 Rainfed crops associated with I1.3 Arable land with unmixed crops
orchard grown by low-intensity agricultural
methods
2.4.1.01.3 Rainfed crops associated with I1.3 Arable land with unmixed crops
olive groves grown by low-intensity agricultural
methods
2.4.1.02.1 Irrigated crops associated with I1.3 Arable land with unmixed crops
vines grown by low-intensity agricultural
methods
(continued)
150 S. Mesquita et al.

Table 1 (continued)
COS’07 EUNIS
2.4.1.02.2 Irrigated crops associated with I1.3 Arable land with unmixed crops
orchards grown by low-intensity agricultural
methods
2.4.1.02.3 Irrigated crops associated with I1.3 Arable land with unmixed crops
olive groves grown by low-intensity agricultural
methods
2.4.2.01.1 Complex cultural systems I1.22 Small-scale market gardens and
horticulture, including allotments
2.4.3.01.1 Agriculture with natural and X07 Intensively-farmed crops interspersed
semi-natural areas with strips of natural and/or semi-natural
vegetation
2.4.4.01.1 Agroforestry systems of cork oak E7.3 Dehesa
with rainfed crops
2.4.4.01.2 Agroforestry systems of holm oak E7.3 Dehesa
with rainfed crops
2.4.4.01.3 Agroforestry systems of other E7.3 Dehesa
oaks with rainfed crops
2.4.4.01.4 Agroforestry systems of other tree E7.3 Dehesa
species with rainfed crops
2.4.4.01.5 Agroforestry systems with cork E7.3 Dehesa
oak, holm oak and rainfed crops
2.4.4.01.6 Agroforestry systems of other E7.3 Dehesa
mixed trees with rainfed crops
2.4.4.02.1 Agroforestry systems of cork oak E7.3 Dehesa
with irrigated crops
2.4.4.02.2 Agroforestry systems of holm oak E7.3 Dehesa
with irrigated crops
2.4.4.02.4 Agroforestry systems of other tree E7.3 Dehesa
species with irrigated crops
2.4.4.02.5 Agroforestry systems with cork E7.3 Dehesa
oak, holm oak and irrigated crops
2.4.4.02.6 Agroforestry systems from other E7.3 Dehesa
mixed trees with temporary
irrigated crops
2.4.4.03.1 Agroforestry systems of cork oak E7.3 Dehesa
with pastures
2.4.4.03.2 Agroforestry systems of holm oak E7.3 Dehesa
with pastures
2.4.4.03.3 Agroforestry systems of other E7.3 Dehesa
oaks with pasture
2.4.4.03.4 Agroforestry systems of other tree E7.3 Dehesa
species with pastures
2.4.4.03.5 Agroforestry systems of cork oak E7.3 Dehesa
with holm oak and pastures
(continued)
Using Geobotanical Tools to Operationalize Mapping … 151

Table 1 (continued)
COS’07 EUNIS
2.4.4.03.6 Agroforestry systems of other E7.3 Dehesa
mixed trees with pastures
2.4.4.04.1 Agroforestry systems of cork oak E7.3 Dehesa
with permanent crops
2.4.4.04.2 Agroforestry systems of holm oak E7.3 Dehesa
with permanent crops
2.4.4.04.3 Agroforestry systems of other E7.3 Dehesa
oaks with permanent crops
2.4.4.04.4 Agroforestry systems of other tree E7.3 Dehesa
species with permanent crops
2.4.4.04.5 Agroforestry systems of cork oak E7.3 Dehesa
with holm oak and permanent
crops
2.4.4.04.6 Agroforestry systems of other E7.3 Dehesa
mixed trees with permanent
cultures
3.1.1.01.1 Cork oak forests G2.112 Southwestern Iberian Quercus
suber forests
3.1.1.01.2 Holm oak forests G2.124 Quercus rotundifolia woodland
3.1.1.01.4 Chestnut forests G1.7D Castanea sativa woodland
3.1.1.01.5 Eucalyptus forests G2.81 Eucalyptus plantations
3.1.1.01.6 Invasive species forests G2.83 Other evergreen broadleaved tree
plantations
3.1.1.01.7 Forests of other hardwoods G1 Broadleaved deciduous woodland
3.1.1.02.1 Cork oak forests with other G2.112 Southwestern Iberian Quercus
hardwoods suber forests
3.1.1.02.2 Holm oak forests with other G2.124 Quercus rotundifolia woodland
hardwoods
3.1.1.02.4 Chestnut forests with other G1.7D Castanea sativa woodland
hardwoods
3.1.1.02.5 Eucalyptus forests with other G2.81 Eucalyptus plantations
hardwoods
3.1.1.02.6 Invasive species forests with other G2.83 Other evergreen broadleaved tree
hardwoods plantations
3.1.1.02.7 Other hardwoods forests G1 Broadleaved deciduous woodland
3.1.2.01.3 Forests of other softwoods G3.F2 Exotic conifer plantations
3.1.2.02.1 Maritime pine forests with other G3.F2 Exotic conifer plantations
softwoods
3.1.2.02.3 Forests of other softwoods G3.F2 Exotic conifer plantations
3.1.3.01.2 Holm oak forests with softwoods G4.F Mixed forestry plantations
3.1.3.01.3 Other oaks forests with softwoods G4.F Mixed forestry plantations
3.1.3.01.4 Chestnut forests with softwoods G4.F Mixed forestry plantations
3.1.3.01.5 Eucalyptus forests with softwoods G4.F Mixed forestry plantations
(continued)
152 S. Mesquita et al.

Table 1 (continued)
COS’07 EUNIS
3.1.3.01.7 Other hardwood forests with G4.F Mixed forestry plantations
softwoods
3.1.3.01.8 Other hardwoods and softwoods G4.F Mixed forestry plantations
mixed forests
3.1.3.02.3 Other softwood forests with G4.F Mixed forestry plantations
hardwoods
3.1.3.02.4 Other softwoods and hardwoods G4.F Mixed forestry plantations
mixed forests
3.2.4.01.1 Open cork oak forests G2.112 Southwestern Iberian Quercus
suber forests
3.2.4.01.2 Open holm oak forests G2.124 Quercus rotundifolia woodland
3.2.4.01.4 Open chestnut forests G1.7D Castanea sativa woodland
3.2.4.01.5 Open eucalyptus forests G2.81 Eucalyptus plantations
3.2.4.01.6 Open invasive species forests G2.83 Other evergreen broadleaved tree
plantations
3.2.4.01.7 Open forests of other hardwoods G1 Broadleaved deciduous woodland
3.2.4.02.1 Open cork oak forests with other G2.112 Southwestern Iberian Quercus
hardwoods suber forests
3.2.4.02.2 Open holm oak forests with other G2.124 Quercus rotundifolia woodland
hardwoods
3.2.4.02.4 Open chestnut forests with other G1.7D Castanea sativa woodland
hardwoods
3.2.4.02.5 Open eucalyptus forests with G2.81 Eucalyptus plantations
other hardwoods
3.2.4.02.6 Open invasive species forests G2.83 Other evergreen broadleaved tree
with other hardwoods plantations
3.2.4.02.7 Open other hardwoods forests G1 Broadleaved deciduous woodland
3.2.4.03.3 Open forests of other softwoods G3.F2 Exotic conifer plantations
3.2.4.04.1 Open maritime pine forests with G3.F2 Exotic conifer plantations
other softwoods
3.2.4.05.2 Open holm oak forests with G4.F Mixed forestry plantations
softwoods
3.2.4.05.4 Open chestnut forests with G4.F Mixed forestry plantations
softwoods
3.2.4.05.5 Open eucalyptus forests with G4.F Mixed forestry plantations
softwoods
3.2.4.05.7 Open forests of other hardwood G4.F Mixed forestry plantations
with softwoods
3.2.4.05.8 Open mixed forests of other G4.F Mixed forestry plantations
hardwoods and softwoods
3.2.4.06.3 Open forests of other softwood G4.F Mixed forestry plantations
with hardwoods
3.2.4.06.4 Open mixed forests of other G4.F Mixed forestry plantations
softwoods and hardwoods
(continued)
Using Geobotanical Tools to Operationalize Mapping … 153

Table 1 (continued)
COS’07 EUNIS
3.2.4.08.1 Clear cuts G5.8 Recently felled areas
3.2.4.08.2 New plantations G5.7 Coppice and early-stage plantations
3.2.4.09.1 Tree nurseries G4.F Mixed forestry plantations
3.2.4.10.1 Firebreaks H5.61 Unsurfaced pathways
3.3.1.01.1 Inland beaches and dunes H5.3 Sparsely- or un-vegetated habitats on
mineral substrates not resulting from
recent ice activity
3.3.1.02.1 Coastal beaches and dunes B1 Coastal dunes and sandy shores
3.3.4.01.1 Burnt areas H5.5 Burnt areas with very sparse or no
vegetation

Annex II

Table 2 Vegetation series present in NUTS II Alentejo

Code Vegetation Series
1 Arbuto unedonis-Querco pyrenaicae Sigmetum
2 Arisaro clusi-Querco broteroi Sigmetum
3 Asparago aphylli-Querco suberis Sigmetum
4 Daphno gnidi-Junipero navicularis Sigmetum
10 Lonicero implexae-Querco rotundifoliae Sigmetum
14 Myrto-Querco rotundifoliae Sigmetum
15 Oleo sylvestris-Querco suberis Sigmetum
16 Osyrio quadripartitae-Junipero turbinatae Sigmetum
28 Phlomido-Juniperetum turbinatae Sigmetum
19 Pyro bourgaeanae-Querco rotundifoliae Sigmetum e Pistacio terebinthi-Querco
broteroi Sigmetum
21 Rhamno olioidis-Querco rotundifoliae Sigmetum
23 Sanguisorbo agrimonioidis-Querco suberis Sigmetum
29 Smilaco-Quercetum suberis Sigmetum e Pistacio-Juniperetum badiae Sigmetum
24 Teucrio baetici-Querco suberis Sigmetum
25 Viburno tini-Oleo sylvestris Sigmetum
34 Riparian geosseries
154 S. Mesquita et al.

Table 3 Crosswalk COS’07 + VS – EUNIS for some types of native heathland, scrub and
woodland
COS’07 Vegetation Series (VS) code EUNIS
3.1.1.01.3 Forests of Mosaic of 1 and 23 G1.7 Thermophilous
other oaks deciduous woodland
3.1.1.02.3 Forests of 19 G1.772 Portuguese
other oaks with other [Quercus faginea] forests
hardwoods
2, 3, 10, 19, 23, mosaic of 2 and 3 G1.772 Portuguese
3.2.4.01.3 Open forests
[Quercus faginea] forests
of other oaks
3.2.4.02.3 Open forests 1 G1.7B [Quercus
of other oaks with pyrenaica] woodland
other hardwoods 14 G2.1 Mediterranean
evergreen [Quercus]
woodland
3.2.2.01.1 Dense scrub 16, mosaic of 4 and 15 B1.6 Coastal dune scrub
3.2.2.02.1 Sparse scrub 34 F3.1 Temperate thickets
and scrub
1, 3, 23, mosaic of 1 and 23 F4.2 Dry heaths
2, 10, 14, 15, 19, 21, 24, 25, 28, 29, F6.1 Western garrigues
mosaic of 19 and 15, mosaic of 2 and
3, mosaic of 3 and 15
3.2.3.01.1 Dense 16 B1.6 Coastal dune scrub
sclerophyte vegetation 2, mosaic of 19 and 15, mosaic of 2 F5 Maquis, arborescent
3.2.3.02.1 Sparse and 3, mosaic of 3 and 15 matorral and
sclerophyte vegetation thermo-Mediterranean
3.2.4.07.1 Other brushes
woody formations
1, 3, 15, 23, 24, mosaic of 24 and 15, F5.2 Maquis
mosaic of 4 and 15
10, 14, 19, 25, 28, 29 F5.5
Thermo-Mediterranean
scrub
34 G1.3 Mediterranean
riparian woodland
Using Geobotanical Tools to Operationalize Mapping … 155

Annex III

Table 4 Crosswalk COS’07 + Geology – EUNIS for pine forests

COS’07 EUNIS in dune systems EUNIS in other
substrata
3.1.2.01.1 Maritime pine G3.714 Iberian Pinus pinaster G3.F22 Exotic
forests ssp. atlantica forests pine plantations
3.2.4.03.1 Open maritime pine G3.714 Iberian Pinus pinaster G3.F22 Exotic
forests ssp. atlantica forests pine plantations
3.1.2.01.2 Stone pine forests G3.7312 Lusitanian stone pine forests G3.F22 Exotic
pine plantations
3.1.2.02.2 Stone pine forests G3.7312 Lusitanian stone pine forests G3.F22 Exotic
with other softwoods pine plantations
3.2.4.03.2 Open stone pine G3.7312 Lusitanian stone pine forests G3.F22 Exotic
forests pine plantations
3.2.4.04.2 Open stone pine G3.7312 Lusitanian stone pine forests G3.F22 Exotic
forests with other softwoods pine plantations
3.1.3.01.1 Cork oak forests G4.B Mixed mediterranean Pinus - G4.F Mixed
with softwoods thermophilous Quercus woodland forestry
plantations
3.1.3.02.1 Maritime pine G4.B Mixed mediterranean Pinus - G4.F Mixed
forests with hardwoods thermophilous Quercus woodland forestry
plantations
3.1.3.02.2 Stone pine forests G4.B Mixed mediterranean Pinus - G4.F Mixed
with hardwoods thermophilous Quercus woodland forestry
plantations
3.2.4.05.1 Open cork oak G4.B Mixed mediterranean Pinus - G4.F Mixed
forests with softwoods thermophilous Quercus woodland forestry
plantations
3.2.4.06.1 Open maritime pine G4.B Mixed mediterranean Pinus - G4.F Mixed
forests with hardwoods thermophilous Quercus woodland forestry
plantations
3.2.4.06.2 Open stone pine G4.B Mixed mediterranean Pinus - G4.F Mixed
forests with hardwoods thermophilous Quercus woodland forestry
plantations
156 S. Mesquita et al.

Annex IV

Table 5 Crosswalk COS’07 + Distance to coastline – EUNIS for sparsely- or un-vegetated areas
COS’07 EUNIS in coastal areas EUNIS in inland areas (>100 m from
(<100 m from coastline) coastline)
3.3.2.01.1 B3 Rock cliffs, ledges and H3 Inland cliffs, rock pavements and
Rock shores, including the outcrops
outcrops supralittoral
3.3.3.01.1 B1 Coastal dunes and sandy H5.3 Sparsely- or un-vegetated habitats on
Sparse shores mineral substrates not resulting from recent
vegetation ice activity

Annex V

Table 6 Grassland types COS’07

included in land use map
COS’07 2.3.1.01.1 Permanent pastures
2.4.1.03.1 Pastures associated with vineyards
2.4.1.03.2 Pastures associated with orchards
2.4.1.03.3 Pastures associated with olive groves
3.2.1.01.1 Natural herbaceous vegetation

Table 7 EUNIS grasslands types present in the study area

EUNIS Definition according to https://eunis.eea. Meadows present in the study area
europa.eu/habitats/
E1 Dry Well-drained or dry lands dominated by Annual non-nitrophile meadows
grasslands grass or herbs, mostly not fertilized and (Tuberarietea guttatae); annual
with low productivity nitrophile communities (Stellarietea
mediae); perennial or biennial
sub-nitrophile communities (Artemisetea
vulgaris); biennial and perennial tall
grasslands (Stipo-Agrostietea,
Festuco-Brometea, Lygeo-Stipetea)
E2 Mesic Lowland and montane mesotrophic and Hygrophilous meadows
grasslands eutrophic pastures and hay meadows of (Arrhenatheretalia); rainfed reseeded
the boreal, nemoral, warm-temperate pastures
humid and mediterranean zones. They
are generally more fertile than dry
grasslands (E1) and include sports fields
and agriculturally improved and
reseeded pastures
E3 Seasonally Unimproved or lightly improved wet Water-logged meadows and rush
wet and wet meadows and tall herb communities of communities (Holoschoenetalia
grasslands the boreal, nemoral, warm-temperate australis); irrigated reseeded pastures
humid, steppic and mediterranean zones
Using Geobotanical Tools to Operationalize Mapping … 157

References

1. Aguiar C (2001) Flora e vegetação da Serra de Nogueira e do Parque Natural de Montesinho.

Ph.D. dissertation. Technical University of Lisbon, 659 p.
2. Aguiar C, Capelo J (2004) Anexo às fichas dos habitats de pinhal: 2180, 2270 e 9540.
In ALFA (2004). Tipos de Habitats Naturais do Anexo I da Directiva 92/43/CEE (Portugal
Continental): Fichas de caracterização Ecológica e de Gestão para o Plano Sectorial da Rede
Natura 2000. Relatório. Lisboa
3. Arnáiz C (1981) Utilización de criterios ecológicos para la valoración del territorio. In: Géhu
JM, Pelt JM (eds) L’evaluation biologique du territoire par la méthode des indices
biocenotiques. Inst. Européen d’ Ecologie, Metz, pp 119–126
4. Arsénio P (2011) Qualidade da paisagem e fitodiversidade. Contributo para o ordenamento e
gestão de áreas costeiras de elevado valor natural. Lisboa: ISA, 426 p.
5. Asensi A (1990) Aplicación de la fitosociología a la evaluación del territorio. Monografías
Flora y Vegetación Béticas 4/5: 91-100
6. Banko G, Weiss M, Moser D, Ubach R, Abdul Malak D, HALADA L, Roerink GJ,
Hazeu GW, Mucher CA, Hennekens SM, Schaminee JHJ, Brodsky L (2013) Developing
Conceptual Framework for Ecosystem Mapping. Kopenhagen, EEA - European Environment
Agency
7. Barbour MG, Burk JH, Pitts WD (1987) Terrestrial plant ecology. Benjamin Cummings,
Menlo Park
8. BISE (2016) https://biodiversity.europa.eu/maes/common-international-classification-of-
ecosystem-services-cices-classification-version-4.3. Acceseed 20 Nov 2018
9. Bordt M (2015) Land and Ecosystem Condition and Capacity - DRAFT v. 1.0. Technical
note. SEEA – EEA
10. Braun-Blanquet J, Pinto da Silva AR, Rozeira A (1964) Résultats de deux excursions
géobotaniques à travers le Portugal septentrional & moyen III. Landes à Cistes et Ericacées
(Cisto-Lavanduletea et Calluno-Ulicetea). Agronomia Lusitana 23(4):229–313
11. Capelo J (2007) Nemorum Transtaganae Descriptio. Sintaxonomia das comunidades florestais
e pré-florestais do Baixo Alentejo. Ph.D. dissertation. Technical University of Lisbon, 528 p.
12. Capelo J (2018) Evidence of a unique association between single forest vegetation-types and
seral sequences: praise for the concept of ‘vegetation series’. In: Greller A, Fujiwara K,
Pedrotti F (eds) Geographical changes in vegetation and plant functional types. Geobotany
studies (basics, methods and case studies). Springer, Cham
13. Capelo J (1996) Esboço da Paisagem Vegetal da Bacia Portuguesa do Rio Guadiana. Silva
Lusitana. nº esp, pp 13–64
14. Capelo J (2007) Nemorum Transtaganae Descriptio. Sintaxonomia das comunidades florestais
e pré-florestais do Baixo Alentejo. Tese de doutoramento. ISA. 528 p. Instituto Superior de
Agronomia
15. Capelo J, Mesquita S, Costa JC, Ribeiro S, Arsénio P, Neto C, Monteiro T, Aguiar C,
Honrado J, Espírito-Santo MD, Lousã M (2007) A methodological approach to potential
vegetation modeling using GIS techniques and phytosociological expert-knowledge: appli-
cation to mainland Portugal. Phytocoenologia 37(3-4):399–415
16. Costa JC, Alvarez ML, Díaz T, Lousã M, Espírito-Santo MD, Vasconcelos T, Monteiro A,
Amor A (1993) Vegetação da Serra de Sintra. Guia Geobotânico da Excursão das XIII
Jornadas de Fitossociologia. XIII Jornadas de Fitossociologia, Lisboa, pp 1–98
17. Costa JC, Neto C, Aguiar C, Capelo J, Espírito-Santo D, Honrado J, Pinto-Gomes C,
Sequeira M, Monteiro-Henriques T, Lousã M (2012) Vascular plant communities in Portugal
(continental, the Azores and Madeira). Global Geobotany 2:1–180
18. Costa JC, Aguiar C, Capelo J, Lousã M, Neto C (1998) Biogeografia de Portugal Continental.
Quercetea 0:1–56
19. EEA (2014) Terrestrial habitat mapping in Europe: An Overview 2014 (EEA Technicalreport
nº 1/2014), MNHN-EEA. European Environment Agency, Luxembourg (ISSN1725-2237)
158 S. Mesquita et al.

20. EUNIS (2007) EUNIS Habitat Classification. https://www.eea.europa.eu/themes/biodiversity/

eunis/eunis-habitat-classification
21. Felipe-Lucia MR, Soliveres S, Penone C, Manning P, Van Der Plas F, Boch S, Prati D,
Ammer C, Schall P, Gossner MM, Bauhus J, Buscot F, Blaser S, Blüthgen N, De Frutos A,
Ehbrecht M, Frank K, Goldmann K, Hänsel F, Jung K, Kahl T, Nauss T, Oelmann Y, Pena R ,
Polle A, Renner S, Schloter M, Schöning I, Schrumpf M, Schulze ED, Solly E, Sorkau E,
Stempfhuber B, Tschapka M, Weisser W, Wubet T, Fischer M, Allan E (2018) Multiple forest
attributes underpin the supply of multiple ecosystem services. Nature Commun 9:4839
22. Gao B (1996) NDWI — a normalized difference water index for remote sensing of vegetation
liquid water from space. Remote Sens Environ 58:257–266
23. Géhu JM, Géhu-Frank J (1991) Essai d’évaluation phytocoenotique de l’artificialisation des
paysages. Coll Phytosoc 17:479–515. https://doi.org/10.1038/s41467-018-07082-4
24. IGP (2011) COS’ 07. https://www.igeo.pt/produtos/CEGIG/Cos2007.htm
25. Izco J (1998) Types of rarity of plant communities. J Veg Sci 9:641–646
26. Jackson TJ, Chen D, Cosh M, Li F, Anderson M, Walthall C, Doriaswamy P, Hunt ER (2003)
Vegetation water content mapping using Landsat data derived normalized difference water
index for corn and soybeans. Remote Sens Environ 92(4):475–482
27. Janssen J, Rodwell J, Bergmeier E, Chytrý M, Schaminée JHJ (2014) Terrestrial and
freshwater typology for European Red List of Ecosystem project. Draft Version, 15 Septemer
2014. Unpublished
28. Keith DA, Rodríguez JP, Rodríguez-Clark KM, Nicholson E, Aapala K, Alonso A,
Asmussen M, Bachman S, Basset A, Barrow EG, Benson JS, Bishop MJ, Bonifacio R,
Brooks TM, Burgman MA, Comer P, Comín FA, Essl F, Faber-Langendoen D,
Fairweather PG, Holdaway RJ, Jennings M, Kingsford RT, Lester RE, Mac Nally R,
Mccarthy MA, Moat J, Oliveira-Miranda MA, Pisanu P, Poulin B, Regan TJ, Riecken U,
Spalding MD, Zambrano-Martínez S (2013). Scientific foundations for an IUCN red list of
ecosystems. PLoS ONE 8(5):e62111. https://doi.org/10.1371/journal.pone.0062111
29. Kent M (2012) Vegetation description and analysis: a practical approach. Wiley-Blackwell,
Oxford
30. Loidi J (1994) Phytosociology applied to Nature Conservation and Land Management. In:
Song Y, Dierschke H, Wang X (eds) Applied vegetation ecology. East China Normal
University Press. Shanghai, pp 17–30
31. Maes J, Teller A, Erhard M, Murphy P, Paracchini Ml, Barredo JI, et al (2014) Mapping and
Assessment of Ecosystems and their Services: Indicators for ecosystem assessments under
Action 5 of the EU Biodiversity Strategy to 2020. (European Union Technical Report;
No. 2014-080). Luxembourg: Publications Office of the European Union. https://doi.org/10.
2779/75203
32. Maes J, Teller A, Erhard M, Grizzetti B, Barredo JI, Paracchini Ml, Condé S, Somma F,
Orgiazzi A, Jones A, Zulian A, Vallecilo S, Petersen JE, Marquardt D, Kovacevic V, Abdul
Malak D, Marin Ai, Czúcz B, Mauri A, Loffler P, Bastrup Birk A, Biala K, Christiansen T,
Werner B (2018) Mapping and Assessment of Ecosystems and their Services: an analytical
framework for ecosystem condition. Publications office of the European Union, Luxembourg
33. Maes J, Zulian G, Thijssen M, Castell C, Baro F, Ferreira Am, Melo J, Garrett Cp, David N,
Alzetta C, Geneletti D, Cortinovis C, Zwierzchowska I, Louro Alves F, Souto Cruz C,
Blasi C, Alos Orti MM, Attorre F, Azzella MM, Capotorti G, Copiz R, Fusaro L, Manes F,
Marando F, Marchetti M, Mollo B, Salvatori E, Zavattero L, Zingari PC, Giarratano MC,
Bianchi E, Dupre E, Barton D, Stange E, Perez-Soba M, Van Eupen M, Verweij P, De
Vries A, Kruse H, Polce C, Cugny-Seguin M, Erhard M, Nicolau R, Fonseca A, Fritz M,
Teller A (2016) Mapping and Assessment of Ecosystems and their Services. Urban
Ecosystems. Publications Office of the European Union, Luxembourg
34. Mesquita S, Sousa AJ (2009) Bioclimatic mapping using geostatistical approaches:
application to mainland Portugal. Int J Climatol 29(14):2156–2170
Using Geobotanical Tools to Operationalize Mapping … 159

35. Mesquita S (2013) Subsistemas da Estrutura Ecologica Nacional – Vegetacao. Magalhaes MR

(coord.) Estrutura Ecologica Nacional – uma proposta de delimitacao e regulamentacao:
Centro de Estudos de Arquitectura Paisagista Professor Caldeira Cabral. ISA Press. Lisboa,
pp 105–120
36. NETO C (2002) A Flora e a Vegetação do superdistrito Sadense (Portugal). Guineana. 8:1–
269
37. Neto C, Arsénio P, COSTA JC, (2009) Flora e vegetação do sudoeste de Portugal
Continental. Quercetea 9:43–144
38. Oliveira, Pereira E, Ramalho M, Antunes MT, Monteiro JH (1992) Cartas Geológicas de
Portugal Continental 1:500 000
39. Pereira EP (2013) Análise geobotânica dos bosques e galerias ripícolas da bacia hidrográfica
do Tejo em Portugal. Ph.D. dissertation. IGOT, 438 p.
40. Ribeiro S (2013) Padrão de diversidade florística de comunidades herbáceas sujeitas a
diferentes usos no território interior de Portugal Continental. Ph.D. dissertation, Technical
University of Lisbon. 361 p.
41. Rivas-Martínez S, Díaz TE, Fernández-González F, Izco J, Lousã M, Penas A (2002)
Vascular Plant Communities of Spain and Portugal. Addenda to the Syntaxonomical checklist
of 2001. Itinera Geobotanica 15(1):5–432
42. Rivas-Martínez S (1976) Sinfitosociología, una nueva metodología para el estudio del paisage
vegetal. Anal Instit Bot Cavanilles 33:179–188
43. Rivas-Martínez S (1987) Memoria del Mapa de Series de Vegetación de España 1:400000.
ICONA. Madrid, 268 p.
44. Rivas-Martínez S (1996) Geobotánica y bioclimatologia. Discursos pronunciados en el acto
de investidura de doctor “honoris causa” de excelentísima señor S. Salvador Rivas-Martínez.
Universidad de Granada, 98 p.
45. Rivas-Martínez S, Sánchez-Mata D, Costa M (1999) North american boreal and western
temperate forest vegetation (syntaxonomical synopsis of the potential natural plant
communities of North America II). Itinera Geobotanica 12:5–316
46. Rivas-Martínez S, Lousã M, Díaz Te, Fernández-González F, Costa JC (1990)La Vegetación
del Sur de Portugal (Sado, Alentejo e Algarbe). Itinera Geobotanica 3:5–126
47. Rivas-Martínez S, Penas A, Díaz T (2001) Bioclimatic Map of Europe-Thermoclimatic Belts.
Cartographic Service, University of Léon, Spain
48. Schaminée JHJ, Chytrý M, Hennekens SM, Janssen Jam, Jiménezalfaro B, Knollová I,
Mucina L, Rodwell JS, Tichý L (2014) Vegetation analysis and distribution maps for EUNIS
habitats. Report for the European Environmental Agency (EEA/NSV/14/006), Copenhagen
49. Schaminée JHJ, Chytrý M, Hennekens SM, Mucina L, Rodwell JS, Tichý L (2012)
Development of vegetation syntaxa crosswalks to EUNIS habitat classification and related
data sets. Final report EEA/NSV/12/001, Alterra, Wageningen, NL, 134 p.
50. Tüxen R (1956) Die heutige potentielle natürliche Vegetation als Gegenstand der
Vegetationskartierung. Angewandte Pflanzensoziologie 13:5–55
51. United Nations, European Union, Food and Agriculture Organization of the United Nations,
Organisation for Economic Co-Operation and Development, World Bank Group (2014)
System of Environmental-Economic Accounting 2012—Experimental Ecosystem
Accounting. UN, New York, US, 177 p.
Spatial Representation of Plant Diversity
at Geographical Scale: The Italian
Experience

Francesco Maria Raimondo

Abstract The first attempts to draw up plant biodiversity maps in Italy are iden-
tified, some of which represent re-workings of data collected for chorological
atlases of geographic areas of different extents. These were based particularly on the
UTM grid, introduced and codified in Italy by Pignatti [6]. The limited examples
made in Italy concern regional, provincial or district areas, some also in protected
natural areas. As examples considered original, the maps developed at the
Phytotaxonomy and Phytogeography Laboratory of the University of Palermo are
presented. These involve the province of Palermo, the Madonie Natural Park in
Palermo province, and the eastern Sicani mountains. The map of this last differs
from the first two only in the type of reticule adopted for the smaller extent of the
base quadrant. The three maps presented constitute complements of more complex
maps that focus on the plant landscape of the mapped areas. In these, plant diversity
attributed to the individual quadrants of the grid is represented by quantitative data
for the specific and infra-specific taxa surveyed, expressed in different color classes.
The quantitative incidence of endemic taxa is shown on each map with different
symbols (red circles of different diameters), also expressed here by diametric rather
than color classes. For the Madonie Natural Park, plant diversity has been mapped
with the same quantitative/color criteria, also showing the phytocoenotic richness of
this important plant area.

Keywords Map
Plant biodiversity
Italy
Palermo province

F. M. Raimondo (&)
Department STEBICEF/Section of Botany and Plant Ecology, University of Palermo,
Via Archirafi 28, 90123 Palermo, Italy
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 161
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_8
162 F. M. Raimondo

1 Introduction

As can be inferred from existing literature, several authors have been engaged in
representing plant diversity of a territory, especially in studies of limited extent. On
the contrary, only a few concern wider areas reaching over territorial boundaries
such as a small island or mountain area. This can be explained easily, considering
the difficulty of acting over wider areas, even without considering the necessary
economic and professional resources. The political and administrative value which
knowledge of biodiversity assumes, for management of a territory and its natural
resources, often leads researchers to analyse and then to represent plant diversity
within territorial and administrative limits, even municipal or provincial limits, but
in some cases also regional or national.
Municipal scales are easier to carry out and they do not stir the researcher much,
while intermediate scales are more satisfying and give the opportunity to find funds
to produce maps quickly, if the flora of that territory is already well explored
taxonomically and geographically. In Italy, numerous efforts involved areas of
large, medium and restricted extent. The national context has been highlighted,
since 1990, by various authors ( [13] ) and by others under the auspices of the
INTERREG IIC MEDOCC, Project “Base de données et cartographie de la
biodiversité” [5]. In Sicily, the Systematics, Phytogeography and Plant Ecology
Laboratory of the Department of Botanical Sciences, University of Palermo,
obtained original results that can be extended to other geographical contexts (4, 8,
11, 12). For central Italy, important contributions have been provided recently by
Lucchese [1, 2].
Italy is a territory rich in plant biodiversity and notoriously complex both cli-
matically and orographically. Apart from methodological aspects, experiences and
maps in Italy were produced for defined geographical areas, such as protected areas
(parks or natural reserves) and administrative territories (municipalities and pro-
vinces). The regional scale is missing, except for an example limited to taxa of
conservation interest in the Lazio region [1, 2].
In this paper three case studies involving our experience are presented. The first
—the most important also for developing the methodology—concerns the province
of Palermo in Sicily [8]. The second concerns the area of a natural park included in
the same provincial territory [11]. The third concerns a representative area in the
Sicani Mountains [4]. In these three cases, the approach is similar but not the same:
the first is more synthetic, the last two more detailed. What changes is therefore the
scale of reference and, in the case of Sicani, also the grid.
Spatial Representation of Plant Diversity … 163

2 Italian Examples

In Italy, summary cartographic documents based on floristic richness, calculated for

areas based on the IGM or UTM grid, and therefore referable to the official car-
tography, are rather limited. The great work of Poldini [7] for the entire
Friuli-Venezia Giulia region, that could have generated a regional map of floristic
biodiversity, does not achieve a summary by basic areas of accumulated quanti-
tative data, and therefore remains only what the author had set for himself, a
chorological atlas, from which one can draw a beautiful regional map of plant
biodiversity. In continental Italy, Lucchese [2] produced a map of plant biodiversity
for the Lazio region, but it concerns only a portion of the regional biodiversity, i.e.
the taxa of greatest protectionist interest (Fig. 1). Unlike the maps of plant biodi-
versity of representative areas of Sicily, which will be presented below [4, 10, 11],
the Lazio map was developed following a different approach and the results are not
comparable. In fact, it was obtained by means of Kriging-like spatial analysis,
which is nothing more than a spatial interpolation of points.

Fig. 1 Plant biodiversity map of Latium region (C Italy) (from [2])

164 F. M. Raimondo

3 Experience of the Phytotaxonomic-Phytogeographic

Laboratory in the University of Palermo

The first of the three cases identified above concerned the whole province of
Palermo (Sicily), about 500,000 hectares; the second case concerned only a part of
it, namely the territory of the Madonie Regional Natural Park, about 45,000 hec-
tares. The third case covered the area of the eastern Sicani Mountains (SW Sicily),
rising in Agrigento province, about 36,000 hectares overall. In all three cases the
territories are very different, in geomorphology and climate. The soils are mainly
carbonate, but siliceous and clayey soils are also frequent. In the first and second
cases, the altitude varies from sea level (approximately) to 1979 m; but in the third
case from 140 to 1578 m. In the three areas, both semi-natural and anthropogenic
environments prevail. Nevertheless, nearly natural and even some very restricted
natural environments do occur [9].

3.1 Plant Biodiversity of Palermo Province

The Province of Palermo lies in Northwestern Sicily (Fig. 2), in a mainly moun-
tainous and hilly area including different environments with various habitats. These
give birth to and shelter biological formations and biotic communities of great
scientific and natural value. The volcanic islet Ustica is also included in the pro-
vince territory. Data acquired from former floristic and vegetation investigations
have been processed in order to produce a landscape and plant-biodiversity
map. From this a map of naturalness has been deduced. The vegetation data (Figs. 5
and 6) have been integrated—for each single unit (quadrant)—with data on floristic
richness as well as distribution and incidence of endemic taxa (Fig. 7). This work
was realised by a stepwise methodology. The starting point was the interpretation of
a set of photographs. Then, 29 vegetation types were defined (Fig. 5), delimited and
mapped in ten 1:50,000 tables (Figs. 3 and 4). All are 20’ longitude by 15’ latitude,
except no. 1 (Balestrate-Partinico) and no. 3 (Camporeale-Corleone) which are 25’
longitude by 15’ and 12’ latitude, respectively. Subsequently, this work has been
verified and supplemented with appropriate field investigations leading both to
recording further data and verifying the borders of each vegetation type, as pre-
viously defined. The resulting map was a combination of up-to-date computer
graphics software (CAD) and computer tools analytically correlating data from
different sources. The management of the data on the identified typologies was
performed by using a GIS (Geographical Information System). Accordingly, hard
questions on the available data were formulated.
The distribution of the records was digitized based on the geographical grid used
in the Central European floristic cartography [6], then applied by Poldini [7] for the
Friuli-Venezia Giulia region and by Lucchese and Lattanzi [3] for the Ausoni
Spatial Representation of Plant Diversity … 165

Fig. 2 Placement of the province of Palermo area (NW Sicily) (from [8])

Fig. 3 On the left, geographical framework of the 19 IGM sheets (1:50000) where the territory of
the province of Palermo (NW Sicily) is placed; on the right, geographical framework of the 10
tables accomplished where the territory of the province of Palermo is placed (from [8])

mountains in Lazio, and more recently by Lucchese [1, 2] for the entire Lazio
region. The grid is 6’ latitude by 10’ longitude.
The derived unit (base area) is divided into four 3’ latitude and 5’ longitude
quadrants with a 40.5 km2 average extent: in total, 167 elementary quadrats were
obtained. The best part of the Palermo province area is cultivated as follows: crops
39.57%; traditional extensive tree cultivation about 11% (olive-groves, to a lesser
extent ash and almond-groves, etc.); and intensive shrub and tree cultivation, on the
whole over 11%. The wooded areas, covering about 24,000 hectares, are mainly in
166 F. M. Raimondo

Fig. 4 Area grid of the province of Palermo (NW Sicily) divided in 167 quadrants. For each of
them are shown the sequence number, the code according to Pignatti [6] and the name,
corresponding to the same name of the C.T.R. (Regional Technic Map) sections in scale 1:10,000
(from [8])

Fig. 5 A synthetic representation of the potential natural vegetation of the province of Palermo
area (from [8])
Spatial Representation of Plant Diversity … 167

Fig. 6 A synthetic representation of the plant landscape of the province of Palermo area (NW
Sicily) (from [10])

Fig. 7 Plant biodiversity map of the province of Palermo area (NW Sicily) (from [8])
168 F. M. Raimondo

the eastern sector, mostly in the Madonie Mountains. Smaller patchy formations,
such as the nature reserves “Bosco della Favara and Bosco Granza”, “Monte San
Calogero and Pizzo Cane, Pizzo Trigna” and “Grotta Mazzamuto”, are in the
central part of the area. The woods in the western sector are uninterrupted within the
“Bosco della Ficuzza, Rocca Busambra, Bosco del Cappelliere e Gorgo del Drago”
natural reserve but become smaller and patchy on the mountains surrounding
Palermo. In the southwestern sector woody formations are located on the Sicani
range, mainly within the “Monte Carcaci”, “Monte Genuardo and Santa Maria del
Bosco” and “Monti di Palazzo Adriano and Valle del Sosio” natural reserves.
Roughly 16,000 hectares of maquis, including shrublands and strongly degraded
woods, occur in the whole area, with a higher incidence on the quadrants Contessa
Entellina-Bisacquino and Marineo-Caccamo.
A complementary map of naturalness status, at 1:125,000—a synthesis of high
biological-environmental value showing the plant communities of the area under a
different degree of human pressure in time—is associated with the map of plant
diversity. It is a map derived from the plant-landscape map on the basis of artificial
elements occurring both in the structure and in the floristic components of the
vegetation types surveyed [12]. The artificiality degrees have been digitized through
a chromatic range. The map shows their incidence and distribution, allowing a
comparison of the present vegetation status and the corresponding natural status.
The ecological systems have been defined as: natural systems, nearly-natural sys-
tems, partly natural systems, rural systems, and highly exploited systems. Over 50%
of Palermo province is rural systems mainly occurring in the hilly hinterland and
lower mountain. Systems under intensive human exploitation (14%) are common in
the littoral and hilly areas. The partly natural systems (29.67%) affect the whole
province territory, while the nearly-natural systems (5.52%), occurring mainly in
the nature park of the Madonie and in some natural reserves, are of low impact. The
natural (sensu stricto) systems are very small (0.1%) and occur exclusively on
inaccessible vertical rocky reliefs, discontinuously located in the province area.

Plant Diversity
Naturally, Palermo province is one of the most interesting areas in Sicily and even
in the Mediterranean basin. Due to its well diversified habitats, it shows an
extraordinary biodiversity both in quantity and quality. With about 2145 specific
and infra-specific taxa (1811 species, 285 subspecies, 39 varieties, 2 forms, 8
nothotaxa), the area (19.47% of Sicily) boasts a conspicuous biological patrimony,
about 81% of all Sicily. There are 127 families, in which the richest in genera are
Asteraceae, Fabaceae, Poaceae, Brassicaceae, Apiaceae and Caryophyllaceae,
with 253, 225, 219, 98, 96 and 89 specific and infraspecific taxa, respectively.
Within the native flora, 224 taxa are endemic, i.e. 5.6% of the vascular flora of the
province and 4.5% of the Sicilian vascular flora [12]. Endemism is particularly
high, both on the Madonie calcareous reliefs and on the mountains south of
Palermo, such as the Rocca Busambra.
Incidence and distribution of the vascular flora in the province is shown by seven
color classes (Fig. 7), each of which corresponds to a different number of
Spatial Representation of Plant Diversity … 169

taxa: <400, 401–500, 501–600, 601–700, 701–800, 801–900, 901–1000, >1000.

Incidence and distribution of endemics are displayed within each quadrant through
five red dots of different diameter. Each dot represents an incidence of endemic taxa
as follows: 1–25, 26–50, 51–75, 77 100, >100. Each quadrant includes 588 taxa on
average.
Within the 167 quadrants into which the Palermo province was divided, plant
biodiversity is generally high in the less populated areas. This is especially
noticeable in the quadrants of the Madonie orographic system, with a taxon number
between 822 (Petralia Sottana) and 1174 (Pizzo Carbonara-Piano Battaglia), as well
as 1093 in Ficuzza-Rocca Busambra, 933 in Monte Pizzuta, and 822 in Pizzo
Trigna. Biodiversity is still appreciable in the quadrants that include the promon-
tories of Monte Gallo and Monte Pellegrino, and in those of Belmonte Mezzagno,
Tommaso Natale, Montelepre, Altofonte, Lago Scanzano, Mezzojuso, Palazzo
Adriano and Filaga. There is a generally close qualitative correlation between
endemism and plant diversity. In the quadrants of the Madonie mountains, endemic
taxa oscillate from 30 (Petralia Sottana) to 118 (Pizzo Carbonara-Piano Battaglia).
Endemism is also well represented in the northern and northwestern parts of the
provincial territory, with highest values in the quadrants of Monte San Calogero
(49), Capo Zafferano (41), Monte Pellegrino (50), Piana degli Albanesi (61); and in
the southwestern mountainous reliefs of Palazzo Adriano (40). The lowest
numerical values are recorded in the areas subjected to greater human impact.
The quadrant no. 82 (Pizzo Carbonara-Piano Battaglia) shows a significant peak
with 1174 taxa, while the quadrant no. 158 (Borgo Callea) has the fewest, at 103
taxa. Plant biodiversity is generally higher in the area under lighter anthropogenic
pressure. Qualitatively, there is a strong correlation between endemism and bio-
logical diversity. In the quadrants including the Madonie mountains (as will be
shown subsequently) the incidence of endemism is at its peak with 62 endemic taxa
in the quadrant no. 82 (Quacella), with the highest number of taxa.

4 Plant Biodiversity Map of the Madonie Natural Park

The landscape and plant biodiversity map of the territory of the Madonie Natural
Park (scale 1:50,000) was also made by GIS, based on previous floristic and
vegetation investigations, air-photo analyses, and appropriate field verifications. It
was drawn up based on data obtained from the study presented earlier for the
territory of the province of Palermo, which includes the area of the Madonie
(Fig. 8). However, this map presents for the first time an attempt to represent not
only the floristic biodiversity calculated from species richness and infraspecific
taxa, but also the phytocenotic diversity calculated from richness in plant associ-
ations, as found in literature and in the field, quadrat by quadrat.
This territory has very varied geology and geo-morphology, with maximum
altitude (Pizzo Carbonara) of 1979 m. The plant landscape, defined according to
physiognomic and structural criteria, is expressed by 24 type units. Units related to
170 F. M. Raimondo

Fig. 8 Placement of the Madonie Natural Park in the province of Palermo (NW Sicily) (from
Raimondo et al. 2004)

pasture are remarkable (about 13,000 ha), as well as to widespread agricultural

crops (9,160 ha) and to forest formations (9,600 ha). Lesser extents are occupied by
Mediterranean maquis cores, shrublands and very degraded woods (about
3500 ha), submontane and mountain grasslands and artificial forest plantations, as
well as peculiar aspects of rock vegetation of high altitude and of lower screes,
which find their maximum expression on the “Serre” of Quacella.
As for environmental integrity, it was possible to distinguish 10 natural levels,
then distributed in five classes on the basis of the degree of human impact and the
structure and quality of the plant systems.

Plant Diversity (Fig. 9)

Apart from the high phyto-coenotic diversity (Fig. 10), the territory of the Madonie
Park also shows high biologic diversity. Within a surface of 1.55% of the entire
administrative Region, about 1500 specific and infra-specific vascular taxa are
found. This is a conspicuous and diversified biologic patrimony equivalent to about
57% of that occurring in the whole region. There are 170 endemic entities, which is
about 46% of the endemism of the whole island. The distribution of the floristic
richness of Palermo province is represented with grids having an average area of
40.5 km2 (3’ latitude by 5’ longitude); this richness is divided into eight classes, the
same as proposed for the map of Palermo province presented before.
Each quadrant hosts 753 entities on average, with a maximum of 1175 in Pizzo
Carbonara-Piano Battaglia. Floristic biodiversity is, in general, very remarkable in
the few humanized areas and in those lacking forest vegetation, where the strong
relationship between endemism and floristic richness is shown.
Spatial Representation of Plant Diversity … 171

Fig. 9 Plant biodiversity map of the Madonie Natural Park (NW Sicily) (from Raimondo et al.
2004)

Plant biodiversity of the peculiar mountain area of Sicily, valued also on a

phytocoenotic basis, is expressed by 127 units between syntaxonomically defined
associations and floristically and ecologically well characterized groups. The richest
quadrants are those of East Polizzi Generosa and Petralia Sottana (Fig. 9).
172 F. M. Raimondo

Fig. 10 Phytocoenotic
biodiversity map of the
Madonie Natural Park (NW
Sicily) (from Raimondo et al.
2004)

5 Plant Biodiversity Map of the Southeastern Sicani Mts

The Sicani Mountains are a mostly calcareous system in southwestern Sicily, rising
in the hinterland of Palermo and Agrigento provinces (Fig. 11). The map shows the
eastern part of the Sicani geographic range, which is entirely in Agrigento province
(N latitude 37° 32’ 26” to 37° 41’ 35”; E longitude 13° 27’ 02” to 13° 50’ 00”). Its
maximum altitude is 1578 m (Monte Cammarata).
The map of the landscape and plant biodiversity of this area was made by
Marino et al. [4] through a Territorial Information System (S.I.T.) in a GIS 3-D
environment. Analysis of this document, which contains 32 vegetation types
gathered into 10 groups, and the data reported by the authors in Table 20 of their
article [4], shows that the autochtonous forest community suffered a remarkable
reduction; they point out several aspects of degradation. Nevertheless, remaining
nuclei of particular importance occur, frequent especially inside the territory of the
natural reserve “Monte Cammarata” and that of SIC “M. Cammarata-Contrada
Salaci”, “Pizzo della Rondine-Bosco di S. Stefano Quisquina”, “La Montagnola e
Acqua Fitusa”, and ZPS “Serra del Leone e M. Stagnataro” (Fig. 12).
These are mainly holm-oak (Quercus ilex) woods or thermophilous deciduous
oak-woods which account for about 11% of the complex. The shrublands and the
brushwoods occur at the margins of wooded formations, while a few edges of
Mediterranean maquis of Oleo-Euphorbietum dendroidis occur on semi-rupestrian
calcareous sites and cover 5% as a whole. Pasture communities (14%) are of major
Spatial Representation of Plant Diversity … 173

Fig. 11 Placement of the eastern Sicani Mts area (SW Sicily) (from [4])

Fig. 12 Plant biodiversity map of eastern Sicani Mts area (SW Sicily) (from [4])
174 F. M. Raimondo

extent, mostly on Mt. Cammarata, and host several endemic or otherwise rare plant
species. Smaller extents are evident for plant communities along water courses
(3%), in marshy meadows and cane fields (3.63%), on calanques (0.9%) and sed-
iment layers (0.55%), and on cliffs (0.33%). Remarkable is the importance of
agricultural and ornamental crops (58.68%), distributed throughout almost all the
territory except for mountains and characterizing the hill landscape between 140
and 900 m. Artificial forest plantations, in the hill and mountain portions of the
central part of the territory, cover about 7% of the total area.
The map of the level of naturalness shows that the territory, exploited since
ancient times with agro-forestry-pastoral activities, is dominated by human-rural
systems (55.53%) and those having intensive use (8.46%). Semi-natural systems
account for 29.07%, nearly-natural systems for 6.6%, and natural systems, sensu
strictu, for only 0.33% of the area, only in almost inaccessible areas.
The map of potential natural vegetation suggests a trend from the present phy-
tocoenoses toward communities of thermophilous deciduous oaks.

Plant Diversity
The floristic richness of this area includes 956 specific and intraspecific taxa, of
which there are 853 species, 88 subspecies and 15 varieties, belonging to 449
genera of 95 families. The taxon-richest families are: Asteraceae (108), Poaceae
(100), Fabaceae (90), Apiaceae (51), Brassicaceae (48), Lamiaceae (44), Liliaceae
(41), Orchidaceae (32), Caryophyllaceae (31) and Scrophulariaceae (25).
The floristic biodiversity increases in the territory of Mt. Cammarata, Serra
Quisquina and Montagnola. In these territories endemism for each base unit
(quadrant) is on average between 40 and 50 specific and infra-specific taxa. The
endemic vascular flora is represented overall by 90 taxa (68 species, 20 subspecies
and 2 varieties), corresponding to 9.43% of the whole flora attributed to the area.
Moreover, there are several rare taxa and/or taxa of particular phyto-geographic
interest, including Celtis tournefortii ssp. asperrima (Lojac.) Raimondo & Schicchi
and Salvia argentea L.
The results of this study contribute to fill gaps in information on plant biodi-
versity of the area. Moreover, these results could be used by the local
Administration in charge of planning and management of the territory, which is
very rich floristic and phytocoenotically, to prepare strategies for conservation and
evaluation of the autochthonous plant patrimony. This investigation also contributes
to the solution of problems connected with recovery and environmental restoration
with particular regard to the extensive artificial forest plantations. Those areas, in
fact, could be reconverted based on up-to-date criteria of natural silviculture,
considering the potentialities of the territory and of the native forest formations,
well preserved, which are optimal role models.
Spatial Representation of Plant Diversity … 175

6 Discussion and Conclusions

The examples considered highlight how it is possible to create maps of plant bio-
diversity based on the floristic and phytocenotic richness of well-explored territories,
using models tested widely in Europe for geographic grids compatible with official
cartography available at different scales in the different countries. In the cases pre-
sented, the maps of floristic diversity do not represent the main purpose, but rather a
complement of more complex maps that focus attention on the representation of the
plant landscape and plant biodiversity more generally. They are therefore the result
of much more complex research that has also attempted cartographic representation
of the vegetation, then typified both on a physiognomic-structural and also a phy-
tosociological basis. Putting together landscape, phytocenotic and plant diversity has
not been easy. For plant diversity, associating the quantitative data with qualitative
data (endemism rate) referred to elementary areas (quadrants) of different extent,
based also on different grids, represents a result that is considered not only original
but above all of remarkable interest for the management of the environmental and
landscape heritage and therefore of biodiversity.

Ackowledgements Thanks to Franco Pedrotti for the decisive stimulus for publishing this article.
Thanks to Roberta Orlando, Laura Bertini, and Edimir Murrja for their collaboration in the editing
of text and figures.

References

1. Lucchese F (2017) Atlante della flora vascolare del Lazio, cartografia, ecologia, biogeografia.
Parte generale e flora alloctona. Regione Lazio, Direzione Capitale naturale, Parchi e Aree
protette, Roma. Digitiliab 1:352
2. Lucchese F (2018) Atlante della Flora vascolare del Lazio, cartografia, ecologia, biogeografia.
La flora di maggiore interesse conservazionistico. Regione Lazio, Direzione Capitale naturale,
Parchi e Aree protette, Roma. Ed. Digitiliab 2:400
3. Lucchese F, Lattanzi E (2000) Atlante della Flora dei Monti Ausoni. Reg. Lazio, Assess.
Utilizzaz. e Valorizzaz. Risorse Ambientali e Univ. Del Molise, Fac. Agraria, Dip. S.A.V.A.
Roma, p 461
4. Marino P, Castellano G, Bazan G, Schicchi R (2005) Carta del paesaggio e della biodiversità
vegetale dei Monti Sicani sud-orientali (Sicilia centro-occidentale). Quad Bot Amb Appl
16:3–60 [text] + map 1:50,000 (Tables 1–2)
5. Margiocco C, Mariotti M (eds.) (2001) Progetto “Basi di Dati e Cartografia della
Biodiversità”. Rapporto finale parte italiana. Iniziativa Comunitaria INTERREG II C/
Programma operativo MEDOCC – Università della Calabria, Centro Editoriale e Librario,
p 215
6. Pignatti S (1978) Dieci anni di cartografia floristica nell’Italia di Nord-Est. Inform Bot Ital
10:212–217
7. Poldini L (1991) Atlante corologico delle piante vascolari nel Friuli-Venezia Giulia. Regione
Autonoma Friuli-Venezia Giulia and Università degli Studi di Trieste, Udine, p 890
8. Raimondo FM (2000a) Carta del paesaggio e della biodiversità vegetale della provincia di
Palermo. Quad Bot Amb Appl 8(1998):3–164 [text] + map 1:50,000 (Tables 1-10)
9. Raimondo FM (2004) Landscape and floristic features of Sicily. Bocconea 17:22–64
176 F. M. Raimondo

10. Raimondo FM (2000b) Cartografie (Tables 1–10) scale 1:50000 (cover). In: Carta del
paesaggio e della biodiversità vegetale della provincia di Palermo. Provincia Regionale di
Palermo/Università degli Studi di Palermo
11. Raimondo FM, Schicchi R, Surano N (2004) Carta del paesaggio e della biodiversità vegetale
del Parco delle Madonie (Sicilia). Naturalista Sicil, s. IV 28(1):71–137 [text]+map 1:50,000
12. Raimondo FM, Gianguzzi L, Surano N (1994) La carta del paesaggio vegetale della provincia
di Palermo (base I.G.M. scale 1:50,000): esemplificazione di elaborazione tramite G.I.S. Boll
Ass It Cartografia 89:71–76
13. Taffetani F, Santolini R (1997) Un metodo per la valutazione della biodiversità su base
fitosociologia e faunistica applicata allo studio di un’area collinare del Montefeltro (Provincia
di Pesaro-Urbino, Italia centrale). Fitosociologia 32:245–271
Cognitive Mapping of Forest Fragments

Frédéric Roulier, Emilie Gallet-Moron, and Guillaume Decocq

Abstract The analysis of spatial cognition (i.e., people’s ability to locate sites in
the geographic space) can help land planning and forest management by predicting
the behavior and reactions of people towards specific landscapes. To achieve this
goal, it is of utmost importance to design robust, replicable methods; this is the
purpose of this study. Here we develop a method which compares locations of an
objective set of points (e.g. extracted from topographic maps) with locations of a
homolog subjective set of points (e.g. collected by means of interviews and sketch
maps) using a bi-dimensional regression analysis. After explaining the method-
ological flow, we demonstrate the applicability of the method to the assessment of
spatial cognition of forest fragments in a 25 km2 agricultural landscape of northern
France.

Keywords Bidimensional regression analysis Cognitive mapping
Forest



fragmentation Objective landmarks Subjective landmarks

1 Introduction

The Sustainable Forest Management (SFM) paradigm was adopted at the

Conference of the Earth in 1992 [16]. It recommends combining material values
(economic and environmental) with non-material values (social, cultural, aesthetic,
spiritual) [1, 13, 18]. Therefore, it now makes sense to evaluate the mental repre-
sentations that people have about forests. We focused in this research on the
development of a method for mapping of spatial cognition, i.e. the ability to locate
places in geographic space. Analysis of this knowledge can be useful for the
management of the forest fragments and predicting the behaviour and reactions of
people.

F. Roulier
E. Gallet-Moron
G. Decocq (&)

Jules Verne University of Picardie, EDYSAN, UMR 7058 CNRS-UPJV, Amiens, France
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 177
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_9
178 F. Roulier et al.

We propose to assess these cognitive representations of space to show differ-

ences between objective spaces (measured according to specific rules) and sub-
jective spaces (knowledge of individuals about space). Basically, our method is
based on a bidimensional regression analysis [14, 15] that compares locations of an
objective set of points (collected from a topographic map) with locations of a
homologous subjective set of points (collected by means of a sketch map). The
deformations and statistics produced by the analysis teach us about the cognitive
representations of this specific space.
However, the bidimensional regression was applied most often to analyze
intra-urban cognition in Strasbourg [4–6]; Swansea [9]; Edmonton [8]; Beijing [12],
and Tokyo [17]. How to exploit this powerful method best when mental repre-
sentations of space are related to rural areas? To illustrate the possibilities of our
method, we will map the spatial cognition of forest patches located in an area of
25 km2 in Thiérache (northern part of the department of Aisne, France).

2 Methodological Issues

The general question is: how location of forest fragments (more generally, of
landscape features in rural areas) is transformed in our minds? Three issues arise
when one wants to answer this question using bidimensional regression. The first
issue is related to the designation of forest fragments. Comparing the objective
location of an object with its subjective location assumes this object is identified by
the individual [10] but also that it can be referred to by a name. Unlike the urban
environment where places have names and are clearly identifiable (library X, statue
Y, square Z …), forest fragments, especially the smallest, have no particular des-
ignation. Even if there is no doubt that forest fragments may register in the cog-
nitive representation of space, it is difficult to designate them with a name during
data collection. The second issue relates to the geometric nature of forest patches.
The parameters of the bidimensional regression are calculated originally for a set of
homologous points. However, the subjects of research do not necessarily corre-
spond to point features. How, in this case, assigning the information on spatial
cognition of polygonal forest fragments for comparison with other data, including
naturalistic data? The third issue concerns the data resource on small space
extents. Compared to a similar surface in urban areas, the number of people likely
to be questioned is necessarily limited and the number of points for bidimensional
regression is much lower.
Cognitive Mapping of Forest Fragments 179

3 Methodological Solution

Our proposal seeks to resolve the three issues identified above (Fig. 1). First, we
collect cognitive data using a survey. The issue of landmark scarcity in rural areas
can be solved by using a pre-survey to identify the remarkable places better (vil-
lages, buildings…) or by an extension of the study area. We chose the first option
because the window was predetermined in the research project. 1. We collect the
image (cognitive) points using a sketch map. 2. The designation of forest fragments
will be conducted indirectly during the data collection. We get the code for iden-
tified fragments by superimposing and intersecting forest areas drawn by each
individual on a reference map (without forest fragments) with maps of forest
fragments identified by a numerical code. 3. Then, we adjust the set of cognitive
points to the set of homologous reference points using bidimensional regression that
produces statistics about the cognitive models. 4. To [affect??] the results of
adjustment to polygonal forest fragments, we will generalize the cognitive models
by interpolation across the study area (using a multiquadric interpolation of a grid).
This produce a space deformation. 5. The interpolation also generates statistics over
the whole surface, which allows us to calculate the inaccuracy surface. 6. Then we
are able to calculate statistical areas corresponding to the averages of the values
contained in each forest patch. Statistics are mapped only for identified patches. We
can use for mapping both interpolated and non-interpolated patches.

Fig. 1 General methodology

180 F. Roulier et al.

4 Application

A test survey was conducted over an area of 25 km2 north of the Aisne1. The original
forest there was fragmented over time by human activities. This area is crossed from
north to south by the National Road 2. This context provides a framework to test our
method. In order to test it in rural areas, we chose to explore a barrier effect produced
by the National Road 2. The barrier effect corresponds to an overestimation of the
distances between places separated by a geographical barrier [7, 11]. For example, it
has been established that such distances were overestimated for places situated on
either side of a river [2] or former iron curtain [3]. The data are intended to be
compared and aggregated. So, only individuals who have drawn the same places
have been retained for the analysis, which explains that the sample has been
(drastically) reduced to a homogeneous set of 9 respondents and 11 landmarks.
For the full sample we note a relative lack of knowledge of the sector, revealed
by a rather low goodness-of-fit (r2= 0.69). The overestimation of distances suggests
the reality of the desired effect. This overestimation is not uniform in intensity, nor
in orientation. These variations lead to inconsistencies expressed through expan-
sions and compressions of space (Fig. 2c). The displacement vectors (Fig. 2a), the
vector field (Fig. 2b) and the deformation of the surface (Fig. 2d) show that sig-
nificant distortions are mostly located west of the National Road 2 (dilatation) and
to a lesser extent, east (compression). All analyses were done using morphoses
software (http://morphoses.eu) [19].
We chose to map the inaccuracy of responses through the error of location. The
error corresponds to the importance of the difference between objective and sub-
jective positions, regardless of the orientation error. By superimposing the error
surface over forest fragments, it is possible to identify those that are poorly located
(Fig. 3b). Some fragments are very large. The average values for a large fragment
could hide very different and even opposite local situations, as shown by the vector
field around the fragment BT10 that combines short and long magnitudes of vectors
(Fig. 3c). We have therefore chosen to transfer location errors to a grid made of
square cells 40 meters on a side (Fig. 3a). This method preserves the possibility of
comparing the data on knowledge with naturalistic data that can be easily associated
with each cell. Finally, the low density of landmarks (fortunately well distributed
here) creates areas where knowledge cannot be taken into account in the cognitive
model. All these calculations were made using the plugin Morphoses for
MapWindow, developed at EDYSAN research unit (http://morphoses.eu).

1
smallFOREST project, «Biodiversity and ecosystem services of small forest fragments in
European landscapes» - EraNet-BiodivERsA 2011-40.
Cognitive Mapping of Forest Fragments 181

Fig. 2 Inconsistencies in the cognitive space

182 F. Roulier et al.

Fig. 3 Error in the knowledge of location

References

1. Angelstam P, Persson R, Schlaepfer R (2004) The sustainable forest management vision and
biodiversity–barriers and bridges for implementation in actual landscapes. Ecol Bull 51:29–49
2. Canter D (1977) The psychology of place. Architectural Press, London
3. Carbon CC, Leder H (2005) The Wall inside the brain: overestimation of distances crossing
the former iron curtain. Psychon Bull Rev 12:746–750
4. Cauvin C (1984a) Espaces cognitifs et transformations cartographiques. Les conditions de la
comparaison des espaces cognitifs: de la carte aux configurations. Exemples de l´espace
urbain strasbourgeois, Strasbourg, Université de Strasbourg, thèse de doctorat, micro-fiches à
Lille n° 85 09 2134
5. Cauvin C (1984b) Une méthode générale de comparaison cartographique: la régression
bidimensionnelle. Travaux et Recherches. Fascicule 4. ERA 214 (CNRS). Thèse de doctorat
d’Etat, Fascicule E, Texte polycopié, Strasbourg
6. Cauvin C, Chernai Z, Daniilidis K (1998) Usagers et représentations cognitives de la ville:
exemples à Strasbourg. In: Reymond H, Cauvin C, Kleinschmager R (eds) dir L’Espace
géographique des villes: pour une synergie multistrates, Paris, Anthropos, coll. « Villes » ,
pp 301–348
7. Cohen R, Baldwin L, Sherman RC (1978) Cognitive maps of naturalistic setting. Child Dev
49:1216–1218
8. Friedman A, Kohler B (2003) Bidimensional regression: assessing the configural similarity
and accuracy of cognitive maps and other two-dimensional data sets. Psychol Methods 8
(4):468–491
9. Kitchin R (1996) Methodological convergence in cognitive mapping research: investigating
configurational knowledge. J Environ Psychol 16:163–185
10. Lynch K (1998) L’image de la cité. Dunod, Paris
11. Newcombe N, Liben L (1982) Barrier effects in the cognitive maps of children and adults.
J Exp Child Psychol 34:46–58
Cognitive Mapping of Forest Fragments 183

12. Shen S, Xue L, Wu L, Liu Y (2007) Cognitive distortions on MDS configuration an sketch
mapping: a case study in Beijing city. In: Li M, Wang J (eds) Geoinformatics 2007:
cartographic theory and models, Proceeding of SPIE, vol. 6751, p 67511A
13. Tabbush P (2004) Public money for public good? public participation in forest planning.
Forestry 77:145–156
14. Tobler WR (1965) Computation of the corresponding of geographical patterns. Papers Reg
Sci Assoc 15:131–139
15. Tobler WR (1994) Bidimensional regression. Geogr. Anal. 26(3):187–212
16. United Nations (1992) Earth Summit 1992: The United Nations Conference on Environment
and Development, Rio de Janeiro 1992, London, The Regency Press Corporation
17. Wakabayashi Y (1994) Spatial analysis of cognitive maps. Geogr Rep Tokyo Metrop Univ
29:57–102
18. Wyder J (2001) Multifunctionality in the Alps: challenges and the potential for conflict over
development. Mt Res Dev 21:327–330
19. Roulier F (2018) Une adaptation de la régression bidimensionnelle aux polygones: un
exemple en cognition spatiale. Cybergeo: European Journal of Geography. https://doi.org/10.
4000/cybergeo.29082
Phytosociology
California Ultramafic Vegetation:
A Phytosociological Update

Daniel Sánchez-Mata

Abstract The vegetation of ultramafics soils throughout the California state

landscapes are summarized. Most important features related with the biogeo-
graphical distribution, structure, dynamics, floristic composition and related data of
the main vegetation types are given and considered. Two appendices (taxonomic
and syntaxonomic) complete this contribution; taxonomic appendix includes some
nomenclatural proposals for the California flora.

Keywords California vegetation
Ultramafic vegetation
California flora

California serpentine vegetation

1 Introduction

Ultramafic rocks are a broad group that includes (besides serpentinite) peridotite,
ophiolite, gabbro, dunite, pyroxenite and hornblendite, among others. Frequent
minerals and elements in these rocks are olivine, chrome, nickel, antimony, cobalt,
mercury, and gold [8, 10, 11]. All these ultramafic rock types cause similar stresses
to plants, and all are present in the state of California (USA). Serpentinite is
relatively easy to determine in the field because of its shiny or silky luster, slightly
soapy feel, conchoidal fractures and greenish surface.
California ultramafic outcrops are typically associated with deep geological
processes such as fault lines, which is why they occur in strips trending southeast to
northwest, parallel to other neighboring fault lines (Fig. 1). Nearly all these ultra-
mafic materials were metamorphosed while being emplaced in, or rising up
through, the continental crust.
Soils developed on ultramafic areas (ultramafic soils) are usually shallow and
skeletal. The steepness of the slopes and sparseness of the vegetation allow for

D. Sánchez-Mata (&)
Department of Pharmacology, Pharmacognosy and Botany (Botany Unit),
Faculty of Pharmacy, Complutense University, 28040 Madrid, Spain
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 187
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_10
188 D. Sánchez-Mata

Fig. 1 Biogeographical map of the Californian Region (I) with the distribution of the main
ultramafic (serpentine) outcrops (black areas), adapted from Sánchez- Mata [35]. The geological
information was obtained from Kruckeberg [18] and the biogeographical limits and unit names are
adapted from Rivas- Martínez et al. [25] (II. Great Basin Region; III. Rocky Mountain Region)
North California Province: I.1.a. Klamath-Ciscascadian Sector; I.1.b. Sierra Nevada-Transcascadian
Sector; I.1.c. North Californian Coastal Sector; I.1.d. California Great Valley Sector. South
California Province: I.2.a. South Californian Ranges Sector; I.2.b. South California Coast &
Channel Islands Sector

continual erosion. The most frequent soils in California’s ultramafic territories are
Lithic Argixerolls, Lithic Haploxerolls and Pachic Argixerolls [1], which all have
low concentrations of calcium, an essential nutrient for plants and whose low
California Ultramafic Vegetation: A Phytosociological Update 189

availability reduces plants’ ability to grow [16, 17, 20, 21, 37–39, 42]. High
magnesium levels and the presence of several heavy metals add further stresses.
Metals are found in large quantities in the tissues of certain plants growing on
ultramafics [7]. Plants on ultramafics have heightened drought resistance [38] and
tolerance to heavy metals [3].
In summary, ultramafic soils impose the following stresses on plants: calcium
and magnesium imbalance, magnesium toxicity, low availability of molybdenum,
toxic levels of heavy metals, occasional high alkalinity, low concentrations of such
essential nutrients as nitrogen and phosphorus, and low soil water storage
capacity [19].
Nomenclature: Taxonomic nomenclature follows the proposals of the second
edition of The Jepson Manual: Vascular Plants of California [5] except those
included in the taxonomic appendix at the end of this contribution. The syntaxo-
nomical appendix includes all the syntaxa known and published from the California
ultramafic (serpentine) areas, and mentions their authority and year of publication;
the vegetation syntaxonomy follows the current issue of the International Code of
Phytosociological Nomenclature [41].

2 Background

The state of California has > 3200 km2 of ultramafic outcrops. These peculiar geo-
logical materials extend west-east from the North Coast Ranges to the west-facing
slopes of the Sierra Nevada, and north-south from the Klamath-Siskiyou Mountains
to the South Coast Ranges (Fig. 1).
A brief compilation of California’s ultramafic vegetation and its diversity was
published as a special chapter in the third edition of Barbour’s classic book
Terrestrial Vegetation of California [6, 32]. All the ultramafic vegetation types can
be considered as xero-edaphic or hygro-edaphic climax. Sargent cypress wood-
lands, pine-oak woodlands, serpentine chaparrals, pine woodlands and forests,
conifer forests, and riparian willow formations are the main vegetation types rec-
ognized throughout the California ultramafics [31, 34, 35].
Continentality is the most important environmental factor driving the community
diversity of ultramafic vegetation in California, as revealed by our own research and
field studies. In hyperoceanic areas (ultra-hyperoceanic, eu-hyperoceanic and
sub-hyperoceanic) with meso-mediterranean thermotypes (Ic < 11) along the Pacific
coast (from Mendocino to San Luis Obispo Counties) and in the San Francisco Bay
Area the potential natural vegetation is conifer forests structured by the endemic
Sargent cypress (Hesperocyparis sargentii). Permanent and seasonal fog, high air
humidity, and the maritime oceanic realm strongly support the existence of this
endemic California vegetation type, and disturbed stands of this peculiar forest
vegetation allow the development of ultramafic chaparral as seral vegetation types
on sunny slopes and in clearings. These climatic features are evidenced (in terms of
190 D. Sánchez-Mata

floristic composition) by the absence of Pinus sabiniana in the most oceanic seral
ultramafic chaparral association, Ceanotho jepsonii-Quercetum duratae, which is
replaced by another conifer tree, Hesperocyparis sargentii, forming potential forest
communities. However, differences in continentality are less marked toward the
interior of the continent (along similar altitudinal ranges), as occurs throughout the
inner Coast Ranges and the Sierra Nevada foothills (Great Valley territories, west to
east sides). The differentiation of ultramafic chaparral communities can thus be
explained mostly by biogeography and altitude.
Pine-oak woodlands on ultramafics at low elevations are related to formations
dominated by blue oak (Quercus douglasii) and grey pine (Pinus sabiniana), which
are widely distributed throughout the foothills that surround the Great Valley; all
these oak-dominated vegetation types are included in the phytosociological class
Heteromelo arbutifoliae-Quercetea agrifoliae (Fig. 2).

Fig. 2 Above, left: Sargent cypress woodland (Hesperocyparis sargentii) at Mount Tamalpais
State Park, 750 m asl; right: mainland ultramafic chaparral (Ceanotho albiflori-Quercetum
duratae) at Napa County, 300 m asl. Below, left: fruited branch of leather oak (Quercus durata),
Napa County, 250 m asl; right: grey pine-blue oak woodland (Pino sabinianae-Quercetum
douglasii) at Lake County, 350 m asl
California Ultramafic Vegetation: A Phytosociological Update 191

Chaparral is a complex of many shrub-dominated vegetation types that cover

8.5% of California [6, 15]. California chaparral can be climatic climax or seral
vegetation; that is, it can develop as climatic climax vegetation on sites that cannot
support large forested communities mainly owing to climatic or edaphic factors.
California ultramafic (serpentine) chaparrals support the potential vegetation in
oceanic ultramafic areas (semihyperoceanic, euoceanic and semicontinental terri-
tories with Ic = 11–21) following the proposals by Rivas-Martínez et al. [27] with
(thermo)-meso-mediterranean thermotypes and arid to lower humid ombrotypes
(Fig. 2). These chaparrals show an interesting spectrum of plant functional traits as
it has been recently published [13].
Throughout the Californian ultramafic areas with a supra-mediterranean,
oro-mediterranean and oro-submediterranean (oro-temperate sub-mediterranean)
thermotype (mountain and high-mountain areas and Klamath-Siskiyou territories),
the potential ultramafic (serpentine) chaparrals are replaced by conifer forests such
as Pacific ponderosa pine forests (Pinus ponderosa subsp. pacifica) in mountain
areas; Jeffrey pine forests (Pinus jeffreyi) at higher altitudes; northern foxtail pine
woodlands (Pinus balfouriana) and Shasta red fir forests (Abies shastensis) in
high-mountain landscapes in the Shasta-Klamath-Siskiyou territories; and southern
foxtail pine woodlands (Pinus austrina) and Critchfield red fir forests (Abies
critchfieldii) in the high mountains of southern Sierra Nevada. All these conifer
vegetation units are framed within the phytosociological class Calocedro
decurrentis-Pinetea jeffreyi (Fig. 3).
California ultramafic grasslands comprise both very diverse perennial and annual
formations; the plant communities structured by annuals (therophytes) were first
described and published by Rodríguez-Rojo et al. [28, 29] with the proposal of the
phytosociological class Vulpio microstachyos-Hesperolinetea micranthi. Barrens
are also devoid of woody vegetation, and occur on steep sites where the parent rock
is serpentinized, as also occurs in rocky outcrops. These habitats support diverse
plant communities that are rich in endemic taxa. In general, ultramafic areas,
especially barrens and rocky outcrops, are refugia for endemics and native taxa as
they serve as islands of native flora and are barely modified by exotic taxa that are
invasive elsewhere, but intolerant of ultramafic soil [2, 12, 23, 30].
192 D. Sánchez-Mata

Fig. 3 Above, left: Group of mature cones, Shasta red fir (Abies shastensis) in Mount Eddy
forests at 1800 m asl; above, right: cone of Siskiyou foxtail-pine (Pinus balfouriana). Below:
Mount Eddy woodlands on ultramafic soils (2300 m asl) are structured by noble fir (Abies
procera), mountain hemlock (Tsuga mertensiana), Shasta red fir (Abies shastensis), and foxtail
pine (Pinus balfouriana)
California Ultramafic Vegetation: A Phytosociological Update 193

3 A Summary of Ultramafic Vegetation Units

3.1 North Coast Ranges

Ultramafics are a frequent geological feature at low elevations in the North Coast
Ranges [22]. The most extensive vegetation types are Sargent cypress woodlands
(Hesperocyparis sargentii) toward the coast; pine-oak woodland (Pino
sabinianae-Quercetum douglasii) and chaparrals (Quercion duratae) in the more
continental areas, such as from Sonoma County to Mendocino and Humboldt
Counties (Fig. 2); and riparian scrub (Salicion breweri). All are xero-edaphic cli-
max vegetation types, except riparian scrub which is hygro-edaphic.

3.1.1 Sargent Cypress Woodland

Sargent cypress woodlands (Hesperocyparis sargentii) cover oceanic, maritime

ultramafic sites (e.g. San Francisco Bay region and some coastal territories of
Humboldt, Mendocino, Sonoma, Marin, Santa Clara, San Mateo, Monterey, San
Benito, San Luis Obispo and Santa Barbara Counties, Fig. 2). These woodlands
extend as a riparian formation close to Brewer willow scrub into more protected
inland areas such as canyons, inland valleys, at the head of streams, and where there
is a frequent occurrence of summer fog (Calycantho occidentalis-Hesperocyparetum
sargentii, Salicion breweri); disturbed sites allow the development of seral scrub
formations structured by Frangula tomentella and Rhododendron occidentale
(Frangulo tomentellae-Rhododendretum occidentale, Salicion breweri) [35].
The floristic composition of Sargent cypress woodlands growing on ultramafics
is very consistent from stand to stand. Hesperocyparis sargentii is usually the
overwhelming or sole dominant tree in the overstory, but Pinus sabiniana may
co-occur in northern and central California and P. coulteri in central and southern
California. In the most mesic or oceanic situations, scattered elements from mixed
evergreen forest are also present (e.g., Arbutus menziesii, Notholithocarpus den-
siflorus, Pseudotsuga menziesii, Quercus agrifolia, Q. chysolepis and Umbellularia
californica).
Leather oak (Quercus durata) is the most common shrub forming seral com-
munities in disturbed sites; Adenostoma fasciculatum, Arctostaphylos spp.,
and Pickeringia montana, among others, are also frequent shrubs. Adjacent
herbaceous communities include the perennial species Calamagrostis ophitidis,
Elymus multisetus, Festuca californica, Galium andrewsii, Melica californica, M.
torreyana, and Polygala californica, plus many ephemerals. Adjacent rocky places
are dominated by Allium falcifolium, Asclepias solanoana, Calochortus raichei
(The Cedars, Sonoma County), Cymopterus terebinthinus var. californica, Galium
californicum, Myriopteris gracillima, Streptanthus morrisonii s.l., and S. tortuosus
var. suffrutescens.
194 D. Sánchez-Mata

3.1.2 Riparian Scrub

Brewer willow formations (Salix breweri) grow on ultramafic streambeds, creeks

and canyons in lowlands throughout the North and South Coast Ranges (Salicion
breweri). They are vicariants of Del Norte willow groves (Salix delnortensis) on
Klamath-Siskiyou ultramafics (Salicion delnortensis).
Associated shrubs include Baccharis glutinosa, Calycanthus occidentalis, Salix
breweri, Salix delnortensis, and Salix lasiolepis. Springs and peaty communities
contain such rare and threatened plants as Carex mendocinensis, Cypripedium
californicum, Delphinium uliginosum, Epipactis gigantea, Gentiana setigera,
Helenium bigelovii, Parnassia palustris and Triteleia peduncularis.
Another riparian scrub type is dominated by Frangula tomentella and
Rhododendron occidentale, and related with the Sargent cypress woodlands as
hygro-edaphic climax formations in southern areas of the North Coast Ranges and
central Great Valley territories, as mentioned (Salicion breweri) [35].

3.2 Great Valley and Sierra Nevada

The most characteristic vegetation developed on ultramafic areas within the Great
Valley is pine-oak woodlands on deep soils developed from scarcely weathered
ultramafic parent rocks throughout the foothills of the inner Coast Ranges, diverse
chaparral types in the more continental areas spreading into the foothills on soils
derived from strongly serpentinized parent rocks, and conifer woodlands at higher
elevations that also grow on strongly serpentinized soil.

3.2.1 Pine-Oak and Pine Woodlands

Pine-oak woodlands are open formations dominated by blue oak (Quercus dou-
glasii) and grey pine (Pinus sabiniana) (Pino sabinianae-Quercetum douglasii).
They occur from the northwest of Glenn County to the Klamath region, east into the
montane zones of the Sierra Nevada, and in south to mid-elevations in Monterey,
San Benito, San Luis Obispo, Fresno, and Kern Counties. These woodlands can be
considered as climatic climax vegetation in lower-altitude ultramafic territories
(Fig. 2).
Pacific Ponderosa pine woodlands (Pinus ponderosa subsp. pacifica) occur
mainly on low-elevation ultramafics in the northwestern part of the Great Valley
and along some west-facing slopes of the Sierra Nevada foothills. They grow on
strongly serpentinized soil in the vicinity of Pacific ponderosa pine-California black
oak woodland on deep, non-ultramafic soil (Querco kelloggii-Pinetum pacificae)
[33]. These woodlands usually have three canopy layers. Pacific ponderosa pine is
frequently associated with Calocedrus decurrens, Pinus attenuata, P. lambertiana,
P. ponderosa subsp. pacifica x P. jeffreyi [4, 9], P. sabiniana, and Quercus
California Ultramafic Vegetation: A Phytosociological Update 195

chysolepis. The shrub layer is dominated by Arctostaphylos viscida and Ceanothus

cuneatus, associated with Arctostaphylos patula, A. viscida x A. patula, Ceanothus
lemmonii, Eriodictyon californicum, Hesperocyparis macnabiana (local),
Heteromeles arbutifolia, Pickeringia montana and Quercus breweri (northwest
border of Sacramento Valley). The herb layer is a mix of sparse perennials (Allium
campanulatum, Ceanothus prostratus, Elymus elymoides, Eriophyllum lanatum s.l.,
Monardella villosa s.l., Penstemon laetus s.l., Polygala cornuta and Salvia
sonomensis) and many annual grasses and forbs.
Jeffrey pine woodlands are also present on ultramafics, but growing at higher
elevations, especially in the Sierra Nevada (Querco vaccinifoliae-Pinetum jeffreyi).
At mid-elevations, they are present in Monterey, San Benito, San Luis Obispo,
Fresno and Kern Counties, and also in the Sierra Nevada. Hybrid populations of
Pinus ponderosa subsp. pacifica, P. jeffreyi and P. coulteri (central and south Coast
Ranges) can be found in some localities [4, 9]. Associated trees include Calocedrus
decurrens, Pinus attenuata, P. coulteri, and P. lambertiana.

3.2.2 Chaparral

Ultramafic chaparral types within the Great Valley are diverse, but all are domi-
nated and structured by the leather oak (Quercus durata) which is also a California
endemic (Quercion duratae). Scattered trees of grey pine (Pinus sabiniana) and
Coulter pine (Pinus coulteri) can be found in central and southern areas. Quercus
durata s.str., a Californian endemic distributed northward of Santa Barbara County,
is the only American species of Quercus that has a restricted distribution on
ultramafics (Fig. 2).
Four ultramafic chaparral types were recently recognized and published for
California by Sánchez-Mata and Rodríguez-Rojo [34], covering all the
xero-edaphic climatic communities developed on oceanic sites (see syntaxonomical
appendix), some ultramafic chaparrals in hyperoceanic areas (Ic < 11) can also be
considered as seral vegetation within the Sargent cypress forest vegetation series, as
occurs throughout the California Pacific coastal territories.
These chaparrals are low open formations growing on strongly weathered
(serpentinized) parent rocks that are widely distributed from Santa Barbara County
northward through the Coast Ranges and Sierra Nevada foothills. The shrubs
exhibit xeromorphic traits: they are dwarfed and compact, they have low produc-
tivity, and the evergreen leaves are mainly malacophyllous, hairy and with reduced
size [13]. Frequent dominants include Adenostoma fasciculatum, Arctostaphylos
spp., Ceanothus cuneatus, C. jepsonii subsp. albiflorus, Frangula californica s.l.,
Fremontodendron californicum, F. decumbens, Garrya congdonii, Heteromeles
arbutifolia, Quercus durata and Rhamnus crocea. Scattered trees, sometimes
growing as tall shrubs, include Pinus sabiniana, Pseudotsuga menziesii, Quercus
agrifolia, Q. breweri, Q. chrysolepis and Umbellularia californica. There is con-
siderable open intershrub space occupied by perennial bunch grasses, subshrubs,
196 D. Sánchez-Mata

and annual herbs. Arctostaphylos spp. and Ceanothus spp. are the most diverse
genera in California and contain many narrowly endemic taxa [14], they play an
important role in differentiating community diversity.
Throughout California ultramafic areas with a supra-mediterranean,
oro-mediterranean and oro-submediterranean (oro-temperate sub-mediterranean)
thermotype (mountain and high mountain areas, and Klamath-Siskiyou territories),
the potential ultramafic (serpentine) chaparrals are replaced by conifer woodlands
and forests, as mentioned.

3.3 Klamath-Siskiyou Mountain Region

The ultramafic vegetation in this unique area is very diverse and more closely
related to that of Sierra Nevada than to that of the North Coast. Typically, several
main ultramafic vegetation types are recognized: Jeffrey pine woodlands (Pinus
jeffreyi); Shasta red fir forests (Abies shastensis) with Brewer spruce (Picea
breweriana), noble fir (Abies procera) and mountain hemlock (Tsuga mertensiana);
subalpine woodlands; and whitebark-pine woodlands on rocky summits (Pinus
albicaulis). Jeffrey pine and whitebark-pine woodlands can be considered
xero-edaphic climax vegetation, while both fir forest formations are climatic climax
vegetation. In addition to these main forest types it is worth noting some unusual
vegetation types, often very rich in endemics and restricted to the ultramafic
Klamath-Siskiyou high-mountain areas, frequently with a local character.

3.3.1 Jeffrey Pine Woodland

The floristic composition of higher-elevation Jeffrey pine stands depends on the

type of ultramafic substrate, whether serpentinite (B1) or diorite/granodiorite (B2).
Associated trees include Abies lowiana, Calocedrus decurrens, Notholithocarpus
echinoides, Pinus contorta subsp. murrayana, P. jeffreyi x P. ponderosa
subsp. pacifica, P. lambertiana, P. monticola, P. sabiniana (local), Quercus
chrysolepis and Q. kelloggii, and scattered individuals of Pseudotsuga menziesii
and Umbellularia californica. There are numerous shrub species: Amelanchier
alnifolia s.l. (B1, B2), Arctostaphylos nevadensis (B2), A. patula (B2), A. patula x
A. nevadensis (A. xbarbouri, B2), A. viscida (B1), and Ceanothus cordulatus.
These open woodlands occur in the most xeric ultramafic sites (mainly growing
on peridotite, gabbro and serpentinite-derived soils) from low to higher elevations
(900–2000 m asl). Pinus jeffreyi and Calocedrus decurrens are the only common
trees present in these open woodlands; Abies lowiana, Hesperocyparis bakeri,
Pinus attenuata, P. monticola, P. lambertiana, P. ponderosa subsp. pacifica
P. ponderosa subsp. pacifica x P. jeffreyi, and Pseudotsuga menziesii co-occur as
scattered individuals. Open spaces are covered by shrubs and small trees such as
Arctostaphylos klamathensis, A. nevadensis, A. viscida s.l., A. patula, Ceanothus
California Ultramafic Vegetation: A Phytosociological Update 197

cuneatus, Frangula californica s.l., Garrya buxifolia, G. fremontii,

Notholithocarpus echinoides, Quercus breweri, Q. vacciniifolia, Vaccinium parv-
ifolium, and shrub forms of Umbellularia californica. Perennial grasses (Elymus
elymoides s.l., E. glaucus, Festuca californica, F. idahoensis, Melica californica,
M. geyeri and Stipa lemmonii) and forbs are often present. These woodlands are
framed in the association Aspidoto densae-Pinetum jeffreyi [24].

3.3.2 Fir forests

Fir forests with occasional Brewer spruce (Picea breweriana) occur in several
localities in northwest California and southern Oregon on ultramafics at elevations
of 1200–1800 m asl. The geographic area of Brewer spruce includes Curry,
Josephine, and Jackson Counties in Oregon and Del Norte, Siskiyou, Humboldt,
Trinity and Shasta Counties in California [40]. These forests are mainly structured
by white fir (Abies lowiana) at mid-elevations and by Shasta red fir (Abies shas-
tensis), noble fir (Abies procera) and mountain hemlock (Tsuga mertensiana) at
higher elevations. Pinus contorta subsp. murrayana, P. jeffreyi, P. lambertiana,
P.monticola, and Pseudotsuga menziesii are common associates (Fig. 3). The
understory is frequently dominated by shrubs such as Arctostaphylos canescens s.l.,
A. nevadensis, A. patula, Ceanothus pumilus, Quercus sadleriana, Q. vacciniifolia,
Vaccinium membranaceum, V. parvifolium and Xerophyllum tenax. Herbs include
Achlys triphylla, Asarum hartwegii, Corallorhiza maculata, Goodyera oblongifo-
lia, Linnaea americana, and Prosartes hookeri s.l., among others.
Noble fir and mountain hemlock forests grow at elevations of > 1800 m asl,
sometimes with Shasta red fir (Abies shastensis). These highland forests should be
framed in the phytosociological class Tsugetea mertensiano-heterophyllae [25, 26].
In a few localities, Abies lasiocarpa is an associate in this forest type [36] (Fig. 3).

3.3.3 Subalpine Woodland

In some high exposed rocky places that do not accumulate snow-packs there is a
subalpine woodland structured by Pinus albicaulis. The understory contains
Cercocarpus ledifolius s.l., Quercus vacciniifolia and Purshia tridentata, and can
be considered to be a xero-edaphic climax. Somewhat more protected mesic sites
support an open foxtail pine woodland (Pinus balfouriana); this type of vegetation
is more restricted to ultramafics than whitebark pine woodland [36]. Associated
trees include Abies shastensis, Pinus albicaulis, P. contorta subsp. murrayana and
P. monticola, and common understory herbaceous perennials in openings are
Aconogonon davisiae, Aspidotis densa, Carex rossii, Eriogonum alpinum and
Lewisia leana.
198 D. Sánchez-Mata

The subalpine woodland that dominates the most mesic subalpine sites, e.g. at
the bottom of late-melting snowpacks, is structured by Tsuga mertensiana in the
overstory and Phyllodoce empetriformis in the understory. This woodland type can
be considered as a permanent climatic vegetation type.

3.3.4 Port Orford Cedar Forest and Adjacent Wetlands

This unusual ultramafic forest type is dominated by Port Orford cedar

(Chamaecyparis lawsoniana) and grows in ravines, along streams and on protected
mesic slopes where there is a strong maritime influence [44]. Associated tree
species in this hygro-edaphic forest include Acer macrophyllum, Alnus rubra, Pinus
monticola, Pseudotsuga menziesii and Taxus brevifolia. Common sub-canopy trees
and shrubs are Cornus nuttallii, Quercus sadleriana, Rhododendron macrophyllum,
R. occidentale, Ribes bracteosum, Rosa californica and Physocarpus capitatus. The
perennial herbs and ferns Adiantum aleuticum, Darmera peltata, Polystichum
munitum, Streptopus amplexifolius var. americanus, and Xerophyllum tenax are
also frequent.
Permanent water runoffs on ultramafics produce fen meadows structured by
Cobra lily (Darlingtonia californica). Calliscirpus criniger, Caltha leptosepala,
Carex echinata, Castilleja miniata subsp. elata, Cypripedium californicum,
Epipactis gigantea, Gentiana setigera, Hastingsia alba, Lilium bolanderi, L.
pardalinum subsp. vollmeri, Narthecium californicum, Platanthera sparsiflora,
Raillardella pringlei, Rudbeckia glaucescens, Pseudotrillium rivale, Triantha
occidentalis, and Viola primulifolia subsp. occidentalis are common perennials.
Less acidic springs support megaforbs such as Aconitum columbianum, Angelica
arguta, Aquilegia eximia, Heracleum maximum, Lilium occidentale, Pteridum
aquilinum subsp. pubescens, Platanthera dilatata var. leucostachys, Senecio tri-
angularis, Valeriana sitchensis, Veratrum californicum, V. insolitum, and V. viride.

3.4 Southern California

The ultramafic vegetation of southern California basically consists of chaparral in

the more continental sites, pine woodlands at mid- and high elevations, Sargent
cypress woodlands on uplands that experience some maritime influence, and
pine-oak woodlands.
Ultramafic chaparral occurs at elevations of <500 m asl. It is structured by
Quercus durata, associated with scattered Pinus sabiniana and Umbellularia cal-
ifornica. Shrubs include Adenostoma fasciculatum, Arctostaphylos obispoensis, A.
pechoensis, Artemisia californica, Ceanothus cuneatus var. fascicularis, C. papil-
losus, C. spinosus, Dendromecon rigida, Diplacus aurantiacus, Eriodictyon
tomentosum, Hesperoyucca whipplei, Quercus berberidifolia, Q. palmeri, Salvia
California Ultramafic Vegetation: A Phytosociological Update 199

leucophylla, S. mellifera, and Trichostema lanatum. Herbaceous perennials and

annuals also have a floristic flavor of southern California.
Sargent cypress woodlands occur on slopes at <600 m asl in the Southern Coast
Ranges exposed to summer fog. Frequent (but scattered) associated trees include
Arbutus menziesii, Lithocarpus densiflorus, Pinus coulteri, P. sabiniana, Quercus
chrysolepis and Umbellularia californica. The understory is dominated by leather
oak (Quercus durata) and consists of the same chaparral shrubs mentioned above.
Jeffrey pine woodlands occupy xeric sites at mid- and high (>900 m asl) ele-
vations that experience strong continentality. Associated trees include Calocedrus
decurrens, Pinus coulteri and hybrid populations of Jeffrey pine and Coulter pine.
These hybrids are particularly common on ultramafics in the South Coast Ranges
[4, 9, 43]. Leather oak (Quercus durata) is a common and notable shrub in these
pine woodlands.
Pine-oak woodlands can grow very locally on ultramafics throughout the South
Coast Ranges. They are dominated by Pinus sabiniana and Quercus douglasii, but
may be joined by Pinus coulteri, Quercus john-tuckeri, and Juniperus californica
as the main tree and shrub species.

4 Appendices

4.1 Taxonomic Appendix

– Abies critchfieldii (Lanner) Rivas-Martínez & Sánchez-Mata

– Abies lowiana (Gordon) A. Murray bis
– Abies shastensis (Lemmon) Lemmon
– Arctostaphylos xbarbouri nothosp. nov. [Arctostaphylos nevadensis A. Gray x
Arctostaphylos patula Greene]. Erect subshrub 0.5–1 m, with intermediate
characters between the parental species: twig and nascent inflorescence axis with
both, short-glandular and nonglandular hairs; leaf erect, ovate to oblancelate,
commonly elliptic, bright green, shiny, puberulent, glabrous in age, base
wedge-shaped, tip abruptly soft pointed, margin entire, flat. Flowers whitish o
weakly pink. Naturally grows between the parent species.
Holotypus and isotypus preserved in MAF Herbaria (MAF 158,485, holotypus;
MAF 158,486, isotypus): Flora of California, UC Davis Herbaria, Nevada
County: Nevada City-Lake Spaulding. Tahoe National Forest. White fir forests
near Casci Ranch Road (Castanopsio sempervirentis-Abietetum lowianae
Rivas-Martínez & Sánchez-Mata 1997), 1220 m, July 31, 1997, D.
Sánchez-Mata & P. Rodríguez-Rojo.
Species devoted with much affection to Prof. Michael G. Barbour (Emeritus
Professor at University of California, Davis), an erudite expert on California and
Pacific Northwest vegetation who passed away on January 7th, 2021.
200 D. Sánchez-Mata

– Ceanothus jepsonii subsp. albiflorus (J.T. Howell) Sánchez-Mata &

Rodríguez-Rojo
– Linnaea americana J. Forbes
– Notholithocarpus echinoides (R.Br.ter) comb. nov. Bas.: Quercus echinoides
R.Br. ter in Ann Mag Nat Hist ser. 4, 7: 251. 1871
– Pinus austrina (R.J. Mastrog. & J.D. Mastrog.) Rivas-Martínez &
Sánchez-Mata
– Pinus ponderosa subsp. pacifica (J.R. Haller & N.J. Vivrette)
Rivas-Martínez & Sánchez-Mata
– Pteridium aquilinum subsp. pubescens (Underw.) Piper & Beattie
– Quercus breweri Engelm.

4.2 Syntaxonomic Appendix

Only the phytosociological units related to ultramafic vegetation types are

considered.
I. HETEROMELO ARBUTIFOLIAE-QUERCETEA AGRIFOLIAE Rivas-
Martínez 1997
• Quercetalia agrifolio-wislizenii Rivas-Martínez 1997
– Quercion douglasio-wislizenii Rivas-Martínez 1997
1. Pino sabinianae-Quercetum douglasii Rivas-Martínez 1997
• Adenostomo fasciculati-Rhamnetalia croceae Rivas- Martínez 1997
– Quercion duratae Sánchez-Mata, Barbour & Rodríguez-Rojo in
Rivas-Martínez 1997
2. Ceanotho jepsonii-Quercetum duratae Sánchez-Mata & Rodríguez-
Rojo 2016
3. Ceanotho albiflori-Quercetum duratae Sánchez-Mata, Barbour &
Rodríguez-Rojo [in Rivas-Martínez] 1997
4. Arctostaphylo glaucae-Quercetum duratae Sánchez-Mata &
Rodríguez-Rojo 2016
5. Arctostaphylo viscidae-Quercetum duratae Sánchez-Mata &
Rodríguez-Rojo 2016
6. Hesperoyucco whipplei-Quercetum duratae Sánchez-Mata &
Rodríguez-Rojo 2016
II. CALOCEDRO DECURRENTIS-PINETEA JEFFREYI Rivas-Martínez &
Sánchez-Mata [in Rivas-Martínez] 1997
• Arctostaphylo patulae-Pinetalia jeffreyi Rivas-Martínez & Sánchez-Mata
[in Rivas-Martínez] 1997
California Ultramafic Vegetation: A Phytosociological Update 201

– Querco vaccinifoliae-Pinion jeffreyi Rivas-Martínez & Sánchez-Mata

[in Rivas- Martínez] 1997
1. Querco vaccinifoliae-Pinetum jeffreyi Rivas-Martínez & Sánchez-
Mata [in Rivas- Martínez] 1997
2. Aspidoto densae-Pinetum jeffreyi Rivas-Martínez & Sánchez-Mata
[in Rivas- Martínez] 1997
III. SALICETEA LASIANDRO-EXIGUAE Peinado & al. 2011
• Salicetalia delnortensis-breweri Peinado & al. 2011
– Salicion delnortensis Sánchez-Mata & Barbour [in Sánchez-Mata]
2018
1. Salicetum delnortensis Sánchez-Mata & Barbour [in Sánchez-
Mata] 2018
– Salicion breweri Sánchez-Mata & Barbour 2007
2. Salicetum breweri Sánchez-Mata & Barbour 2007
3. Solidago guiradonis-Salicetum breweri Sánchez-Mata & Barbour
[in Sánchez-Mata] 2018
4. Calycantho occidentalis-Hesperocyparetum sargentii Sánchez-
Mata & Barbour [in Sánchez-Mata] 2018
5. Frangulo tomentellae-Rhododendretum occidentalis Sánchez-
Mata & Barbour [in Sánchez-Mata] 2018
IV. VULPIO MICROSTACHYOS-HESPEROLINETEA MICRANTHI
Rodríguez-Rojo, Sánchez-Mata, Rivas-Martínez & Barbour 2001
• Eriogono luteoli-Hesperolinetalia micranthi Rodríguez-Rojo, Sánchez-
Mata, Rivas-Martínez & Barbour 2001
– Hesperevaco sparsiflorae-Hemizonion congestae Rodríguez-Rojo,
Sánchez-Mata, Rivas-Martínez & Barbour 2001
1. Hesperevacetum sparsiflorae Rodríguez-Rojo & Sánchez-Mata [in
Rodríguez-Rojo, Sánchez-Mata, Gavilán, Rivas-Martínez &
Barbour] 2001
– Hesperolinion clevelendii Rodríguez-Rojo, Sánchez-Mata, Rivas-
Martínez & Barbour 2001
2. Hesperolinetum clevelandii Rodríguez-Rojo & Sánchez-Mata [in
Rodríguez-Rojo, Sánchez-Mata, Gavilán, Rivas-Martínez &
Barbour] 2001
– Hesperolino micranthi-Navarretion filicaulis Rodríguez-Rojo,
Sánchez-Mata, Rivas-Martínez & Barbour 2001
202 D. Sánchez-Mata

3. Hesperolino micranthi-Navarretietum filicaulis Rodríguez-Rojo &

Sánchez-Mata [in Rodríguez-Rojo, Sánchez-Mata, Gavilán, Rivas-
Martínez & Barbour] 2001
– Streptanthion polygaloidis Rodríguez-Rojo, Sánchez-Mata, Rivas-
Martínez & Barbour 2001
4. Streptanthetum polygaloidis Rodríguez-Rojo & Sánchez-Mata [in
Rodríguez-Rojo, Sánchez-Mata, Gavilán, Rivas-Martínez &
Barbour] 2001

References

1. Alexander EB, Coleman RG, Keeler-Wolf T, Harrison S (2007) Serpentine geoecology of

Western North America: geology, soils, and vegetation. Oxford University Press, New York,
p 512
2. Anderson RC, Fralish JS, Baskin JM (eds) (2007) Savannas, barrens, and rock outcrop plant
communities of North America. Cambridge University Press, New York, p 484
3. Antonovics J, Bradshaw AD, Turner RG (1971) Heavy metal tolerance in plants. Adv Ecol
Res 7:1–85
4. Armstrong WP (1977) Natural plant hybrids in San Diego County. Env South 477:14–16
5. Baldwin BG, Goldman DH, Keil DJ, Petterson R, Rosetti TJ, Wilken DH (2012) The Jepson
manual. Vascular plants of California, 2nd edn. California University Press, Berkeley, p 1568
6. Barbour MG, Keeler-Wolf T, Schoenherr AA (2007) Terrestrial vegetation of California, 3rd
edn. University of California Press, Berkeley, p 730
7. Bargagli R (1998) Trace elements in terrestrial plants: an ecophysiological approach to
biomonitoring and biorecovery. Springer-Verlag and RG Landes Co., Berlin, p 324
8. Bates RL, Jackson JA (eds) (1984) Dictionary of geological terms (rocks, minerals and
gemstones), 3rd edn. American Geological Institute. Anchor Books, Doubleday, New York,
p 576
9. Callaham RZ (1956) Needle oils of three pine species and species hybrids. Forest Sci 2
(2):101–105
10. Coleman RG (1977) Ophiolites. Springer, Berlin, p 229
11. Coleman RG, Jove C (1992) geological origin of serpentinites. In: Baker AJM et al (eds) The
Vegetation of Ultramafic (Serpentine) Soils. Intercept, Andover, pp 1–17, p 509
12. Harrison S, Safford HD, Grace JB, Viers JH, Davies KF (2006) Regional and local species
richness in an insular environment: serpentine plants in California. Ecol Mon 76:41–56
13. Hidalgo-Triana N, Pérez Latorre AV (2018) Plant functional traits and groups in a California
serpentine chaparral. Ecol Res 33:525–535
14. Keeley JE, Davis FW (2007) Chaparral. In: Barbour MF et al (eds) Terrestrial vegetation of
California, 3rd edn. University of California Press, pp 339–366, p 730
15. Keeley JE, Soderstrom TJ (1986) Postfire recovery of chaparral along an elevational gradient
in Southern California. Southw Nat 31:177–184
16. Kruckeberg AR (1951) Intraspecific variability in the response of certain native plant species
to serpentine soil. Am J Bot 38:408–419
17. Kruckeberg AR (1954) Plant species in relation with serpentine soils. In: Whittaker RH
(ed) The ecology of serpentine soils: a symposium. Ecology, vol 35, pp 267–274
California Ultramafic Vegetation: A Phytosociological Update 203

18. Kruckeberg AR (1984) California serpentines: flora, vegetation, geology soils, and
management problems, vol 78. University of California Publications in Botany. University
of California Press, Berkeley, p 180
19. Kruckeberg AR (2002) Geology and Plant Life. The Effects of Landforms and Rock Types on
Plants. University of Washington Press, p 362
20. Mason HL (1946) The edaphic factor in narrow endemism. I. The nature of environmental
influences. Madroño 8:209–226
21. Mason HL (1946) The edaphic factor in narrow endemism. II. The geographic occu rrence of
plants of highly restricted patterns of distribution. Madroño 8:241–257
22. McCarten N (1988) Serpentine plant communities of San Francisco Bay Region. In: Elias TS
(ed) Conservation and management of rare and endangered plants. Proceedings of a California
conference on the conservation and management of rare and endangered plants (1986).
California Native Plant Society, Sacramento, p 630
23. Rajakaruna N (2018) Lessons on evolution from the study of edaphic specialization. Bot Rev
84:39–78
24. Rivas-Martínez S, Sánchez-Mata D (1997) Calocedro decurrentis-Pinetea jeffreyi. In: Rivas-
Martínez S (ed) Syntaxonomical synopsis of the potential natural plant communities of North
America, I. Itinera Geobot, vol 10, pp 54–90
25. Rivas-Martínez S, Sánchez-Mata D, Costa M (1999) North American boreal and western
temperate forest vegetation (Syntaxonomical synopsis of the potential natural plant
communities of North America, II). Itinera Geobot 12:5–316
26. Rivas-Martínez S, Sánchez-Mata D, Costa M (1999) North American new phytosociological
classes. Itinera Geobot 13:349–352
27. Rivas-Martínez S, Penas A, del Río S, Díaz González TE, Rivas Sáenz S (2017)
Bioclimatology of the Iberian penisnula and the Balearic Islands. In: Loidi L (ed) The
vegetation of the Iberian Peninsula, vol 1. Plant and Vegetation 12. Springer, pp 29–80, p 676
28. Rodríguez-Rojo MP, Sánchez-Mata D, Gavilán RG, Rivas-Martínez S, Barbour MG (2001)
Typology and ecology of Californian serpentine annual grasslands. J Veg Sci 12:687–698
29. Rodríguez-Rojo MP, Sánchez-Mata D, Rivas-Martínez S, Barbour MG (2001)
Syntaxonomical approach for classification of the Californian serpentine annual grasslands.
Lazaroa 22:83–94
30. Safford HD, Viers JH, Harrison SP (2005) Serpentine endemism in the California flora: a
database of serpentine affinities. Madroño 52(4):222–257
31. Sánchez-Mata D, Rodríguez-Rojo MP, Barbour MG (2004) California ultramafic vegetation:
biodiversity and phytosociological survey. In: Boyd RS et al (ed) Ultramafics rocks: their
soils, vegetation and fauna (Rocas ultramáficas: sus suelos, vegetación y fauna), Proceedings
of the fourth international conference on serpentine ecology, St Albans (UK), pp 177–181,
p 347
32. Sánchez-Mata D (2007) Ultramafic vegetation. Chapter 3: California soils and examples of
ultramafic vegetation (A. T. O’Geen, R. A. Dahlgren, and D. Sánchez-Mata). In: Barbour MG
et al (eds) Terrestrial vegetation of California, 3rd edn. University of California Press, pp 93–
(71)–106, p 730
33. Sánchez-Mata D (2012) Overview of the forest vegetation of the Sierra Nevada Range
(California, USA). Berich RT-Gesells 24:141–153
34. Sánchez-Mata D, Rodríguez-Rojo MP (2016) Mediterranean ultramafic (serpentine) cha-
parrals of California (USA): a geobotanical overview. In: Box EO (ed) Vegetation structure
and function at multiple spatial, temporal and conceptual scales. Geobotany Studies, Springer,
pp 285–312, p 578
35. Sánchez-Mata D (2018) Studies on California ultramafic (serpentine) vegetation: riparian
woody formations. Berich RT-Gesells 30:175–190
36. Sawyer JO, Thornburgh DA, Bowman WF (1970) Extension of the range of Abies lasiocarpa
into California. Madroño 20:413–415
37. Walker RB (1948) Molybdenum deficiency in serpentine barren soils. Science 108:473–475
204 D. Sánchez-Mata

38. Walker RB (1954) Factors affecting plant growth on serpentine soils. In: Whittaker RH
(ed) The ecology of serpentine soils: a symposium. Ecology, vol 35, pp 258–266
39. Walker RB, Walker HM, Ashworth PR (1955) Calcium-magnesium nutrition with special
reference to serpentine soils. Plant Phys 30:214–221
40. Waring RH, Emmingham WH, Running SW (1975) Environmental limits of an endemic
spruce, Picea breweriana. Canad J Bot 53:1599–1613
41. Theurillat J-P, Willner W, Fernández-González F, Bultmnan H, Carni A, Gigante D,
Mucina L, Weber H (2021) International Code of Phytosociological Nomenclature, 4th edn.
App Veg Sci 2021:24:e12491
42. Whittaker RH (1954) The vegetational response to serpentine soils. In: Whittaker RH
(ed) The ecology of serpentine soils: a symposium. Ecology 35:275–288
43. Zobel B (1952) Jeffrey pine in the south Coast Ranges of California. Madroño 11:283–284
44. Zobel DG, Hawk GM (1980) The environment of Chamaecyparis lawsoniana. Am Mid Nat
103:280–297
Plant Associations of Petasition
officinalis Alliance in the East
Carpathians (Călimani and Gurghiu
Mountains Romania)

Silvia Oroian, Mihaela Sămărghiţan, and Erzsébet Domokos

Abstract This paper presents the results of phytosociological research of tall-herbs

vegetation from the East Carpathians Rank Călimani and Gurghiu Mountains.
A total of five plant associations were analysed with the classification in the
Petasition officinalis Sillinger 1933. The study area refers to plant communities
dominated by Petasites sp. which occur along the river banks and on the fringe of
forests in the montane belt. The paper is based on field researches during 2000–
2014. A data set of 70 relevés were analysed in term of flora composition and
geomorphological features of the habitat. For analysis of the relationship between
floristic composition and habitat variables the average non-weighted Ellenberg
indicator values (for light, moisture, temperature, soil reaction, soil nutrient content)
and the Shannon-Wiener indices of the relevés were used.

Keywords Phytosociology
Montane vegetation
Tall-herb plant communities

1 Introduction

The paper presents the results of the researches carried out in the mesophilous
communities belonging to the Petasition officinalis alliance, which develop along
the mountain valleys and the forest fringes of the Călimani-Gurghiu Mountains
(East Carpathians). The paper aims to emphasize the interconditioning and mutual

S. Oroian
E. Domokos
Department of Fundamental Pharmaceutical Sciences, Discipline of Pharmaceutical Botany,
University of Medicine, Pharmacy, Science and Technology of Tîrgu Mureş, 38 Gh.
Marinescu Street, 540139 Tîrgu Mureş, Romania
M. Sămărghiţan (&)
Department of Natural Sciences, Mureş County Museum, 24 Horea Street,
540036 Tîrgu Mureş, Romania
E. Domokos
Department of Horticulture,Faculty of Târgu Mureş, Sapientia Hungarian University
of Transylvania, 540485 Târgu-Mureş, Romania

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 205
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_11
206 S. Oroian et al.

influences between the plant communities and the complex environmental factors.
Features like geographical position, altitude, soil and rock types, as well as an
interspecific relationship determine a high structural diversity of studied plant
communities.
The study area stretches on the montane beech forest belt of the Călimani-
Gurghiu Mountains, with an altitude range of 500 to 1,600 m. The mountain belt
terrain is well represented in all mountains in the Romanian Carpathians. The
researched areas overlap with protected areas of the Upper Mures Gorge Natural
Park and ROSCI0019-Călimani-Gurghiu Natural Park Natura 2000 site. The entire
level is covered by beech and spruce forests that are in a good conservation status.
Riparian communities identified in the studied territory were assigned to the
Petasition officinalis alliance.

2 Materials and Methods

The field investigations carried out during the period 2000–2014 aimed to establish
the qualitative structure of phytocoenoses dominated by the Petasites sp. identified
in the area. A total of 70 surveys were analyzed which were assigned to 5 plant
associations. The analysis for each plant association consisted of: the specific
composition and belonging to the species of the higher coenotaxa.
For the study of vegetation the research was based on the methodology of the
Zurich-Montpellier Phytosociological School (Braun-Blanquet method) adapted to
the particularities of our country [2, 8]. Due to high number of relevés the obtained
results are presented in form of synoptic tables.
The species complex was analyzed according to species requirements to eco-
logical factors. The affinity of each species for different environmental conditions
was characterized using the Ellenberg indicator values for vascular plants (light,
temperature, humidity, soil reaction and nitrogen). The ecological indicators, as
well as the name of the species were based on Sârbu et al. [31]. The national and
international red lists were used [1, 3, 10, 11, 24] for the identification and clas-
sification of the species of plants at different degrees of vulnerability. The phyto-
coenoses were reported to Natura 2000 habitat type according to Gafta and
Mountford [13] and Doniţă et al. [12].

3 Results and Discussions

The general appearance of the vegetation of the Călimani and Gurghiu Mountains
on the territory of Mureş County is a reflection of a sum of factors such as past and
present physico-geographic conditions, the origin of the various floristic elements
and the impact of the anthropogenic factors. These anthropogenic factors induced
profound changes in the vegetations’ composition through extensive forestry,
Plant Associations of Petasition officinalis Alliance … 207

hydropower systems, and the change of land use by crops. In this examination we
took into account some tall-herbs (“megaphorbs”) with a zonal characteristic,
installed after the deforestation of forests and hedges [6]. The vegetation can reach
up to 50–150 cm high. Low altitude communities are often severely degraded and
invaded by anthropophilic weeds, sometimes even of neophyte species.
In this analysis, we considered the megaphorbs of the Petasition officinale alli-
ance. The studied plant associations spread over the mountain valleys or the fringe
of the mesophilous forests, forming borders along the river sides. Coenoses grow in
shady or semi-shady places, on more or less damp soils, that can include thick
gravel, but the soil is rich in humus. Research carried out refers to five plant
associations from the Petasition officinalis alliance in the habitat: 6430—
Hydrophilous tall-herb fringe communities of plains in the montane to alpine levels.
A total of 70 phytosociological surveys were performed in the studied area,
which were assigned to five plant associations that were reported in the Petasition
officinalis alliance, Petasito-Chaerophylletalia order.
The variability of the environmental conditions as well as the floristic compo-
sition of these tall-weeds gave rise to various opinions regarding their phytosoci-
ological framing.
Morariu in 1967 described the order Petasito-Chaerophylletalia and included it
in Epilobietea angustifolii Tx. et Prsg. 1950 class. This class includes the herba-
ceous vegetation of montane forest clearings, deforested sites. In the
Petasito-Chaerophylletalia order can meet species from montane forest clearings but
also species characteristic of mountane streams. Because of that, later the order was
classified in montane tall-weeds of Betulo-Adenostyletea Br.-Bl. et Tx. 1943,
Adenostyletea Knapp 1943 class (syn. Mulgedio-Aconitetea Hadać et Klika in
Klika et Hadać 1944) [28]. Some authors include Petasition officinalis alliance into
the Galio-Urticetea class based on the presence of many nitrophilic and synan-
thropic species [4, 7, 21, 26, 28]. Given the natural character of habitats, the
Petasition association is usually ranked also in the subalpine high-herbs vegetation
class Mulgedio-Aconitetea [16, 17, 19, 22].
Considering the floristic composition of the conenoses and the environmental
factors that determine the studied conenoses to spread over the altitudinal gradient,
we considered this last classification in our study.
Cls. Mulgedio-Aconitetea Hadać et Klika in Klika et Hadać 1944
Ord. Petasito-Chaerophylletalia Morariu 1967
All. Petasition officinalis Sillinger 1933
Ass. Telekio-Petasitetum hybridi (Morariu 1967) Resmeriţă et Raţiu 1974
Ass. Telekio-Filipenduletum Coldea 1996
Ass. Telekio speciosae-Aruncetum dioici Oroian 1998
Ass. Petasitetum kablikiani Szafer et al. 1926
Ass. Arunco-Petasitetum albi Br.-Bl. et Sutter 1977

1. Telekio-Petasitetum hybridi (Morariu 1967) Resm. et Raţiu 1974 (syn.

Petasiteto - Telekietum speciosae Morariu 1967)
208 S. Oroian et al.

This endemic association of the Romanian Carpathians (Table 1, column 1, Fig. 1)

is frequent along the streams, near the springs or on the edge of mesophilous
forests. This association has been detailed in several mountains of the Eastern
Carpathians: Maramureş [27], Baraolt [9], Rodnei [5], Calimani, Gurghiu [25],
Hășmaș [23], Gurghiu [30].
The phytocoenoses grow lushly on alluvial soil as well as on colluvial accu-
mulations, with favorable humidity and moderate or abundant quantities of nitrates.
They are very common in the study area. Due the characteristic species: Petasites
hybridus and Telekia speciosa, these phytocoenoses have a regional peculiarity in
the Romanian Carpathians, unlike those described in Central Europe. The coenoses
have considerable coverage, due to the broad leaves of the characteristic species that
shade the ground. Because of that other plant species develop only on the peripheral
sides of the phytocoenosis. The coenoses form more or less narrow strips, along river
banks and forest fringes. In their floral composition there are often found numerous
forest elements stemming from the woods in the close vicinity. The phytocoenoses
of the association were identifyed at altitudes between 660 and 1215 m. For this
association, 39 phytosociological surveys were performed. These phytocoenoses are
characterized by a great floral wealth, with 244 species being identified. The plant
communities are layered, achieving good coverage of up to 100%. The floral
composition includes the edifying species: Telekia speciosa and Petasites hybridus.
The diagnostic and dominant species identified are Telekia speciosa, Petasites
hybridus, Angelica sylvestris, Cirsium oleraceum, Geranium palustre, Scirpus syl-
vaticus, Filipendula ulmaria, Chaerophyllum hirsutum etc.
Regarding their belonging to different coenotic groups, there is a high share of
the species of the superior syntaxa, the Petasition officinalis alliance (Aruncus
dioicus, Cirsium oleraceum, Chrysosplenium alternifolium, Cirsium erisithales)
and Petasito-Cherophylletalia aliance (Angelica sylvestris, Chaerophyllum hirsu-
tum, Aegopodium podagraria, Impatiens noli-tangere, Rumex obtusifolius etc.).
The endemic or subendemic species with a remarkable phytogeographical interest
are: Aconitum moldavicum, Dianthus barbatus ssp. compactus and Leucanthemum
waldsteinii. There are also Carpathian-Balkan species (Campanula abietina,
Petasites kablikianus, and Pulmonaria rubra). Adenostyletalia order species like
Achillea distans, Cicerbita alpina, Deschampsia cespitosa, Gentiana asclepiadea,
Senecio ovatus, Cirsium waldsteinii, Doronicum austriacum are also present,
especially in higher elevations. This is due to the proximity of these coenoses with
tall-weeds of the mountain area.
2. Telekio-Filipenduletum Coldea 1996
Hygrophilic coenoses ediphied by Filipendula ulmaria and Telekia speciosa are
spread in the study area in intermontane valleys on humus-rich alluvial soils at
altitudes between 651 and 1063 m. For this association, 8 phytosociological sur-
veys were performed. Phytocoenoses consist of 132 layered species, achieving a
good coverage of about 80–100%. Diagnostic species are Myosoton aquaticum and
Scirpus sylvaticus. Among the characteristic species, we mention Telekia speciosa,
Petasites hybridus, and Filipendula ulmaria (Table 1, Column 2, and Fig. 2). The
structure of the association is well represented by the species characteristic of the
Plant Associations of Petasition officinalis Alliance … 209

Petasition officinalis alliance and the Petasitio-Chaerophylletalia order. In terms of

flora, these phytocoenoses in Romania differ from those described in the Czech
Republic [14] by the small number of species characteristic of the Calthion alliance,
but those of the Petasition officinalis alliance are much better represented similar to
those from Slovakia [18].
The phytocoenoses are distributed in more or less narrow strips along the forest
fringe or water courses. Due to the proximity of the forest, a floral composition with
noticeable forest elements is easily distinguishable.
Considering the coenotic groups one can notice a high weight of the species
belonging to the higher syntaxa, the alliance Petasition officinalis, Petasito-
Chaerophylletalia order (Chaerophyllum hirsutum, Stellaria nemorum, Mentha
longifolia, Myosotis scorpioides, Rumex alpinus, Carduus personata, Cirsium
erisithales, Valeriana sambucifolia) and the Mulgedio-Aconitetea class (Hypericum
maculatum, Rosa pendulina, Silene dioica, Veratrum album). Additionally,
Adenostyletalia species like Deschampsia cespitosa, Senecio ovatus, Cirsium
waldsteinii, Doronicum austriacum are also present. We note that the drought and
low precipitation in recent years have led to the reduction of the phytocoenoses
edifyed by Telekia speciosa with Filipendula ulmaria.

Fig. 1 Telekio-Petasitetum
hybridi (foto Mihaela
Sămărghiţan)

Fig. 2 Telekio-Filipenduletum (foto Mihaela Sămărghiţan)

210 S. Oroian et al.

3. Telekio speciosae-Aruncetum dioici Oroian 1998

The plant groups with Aruncus dioicus and Telekia speciosa develop in semi-shady,
inclined (up to 40°) slopes with northern exposure at altitudes ranging from 500 to
926 m in Sovata (Ilieşi), Bistra Mureşului on the Bistra Valley and Lunca Bradului
on the Ilişoara Mare valley. The association was also described from the Mureş
Gorge in the Borzia and Răstoliţa-Costeasa Valley [25]. The diagnostic species for
association are Gentiana asclepiadea, Spiraea chamaedryfolia, Aruncus dioicus,
Campanula trachelium, Pulmonaria rubra. The characteristic species of the asso-
ciation are Aruncus dioicus a circumpolar-boreal species and Telekia speciosa, a
Carpathian-Balkan-Caucasian species, which achieves considerable coverage due to
the broad leaves. In addition, Spiraea chamaedrifolia with constancy III is char-
acteristic (Table 1, column 3, Figs. 3, 4). 10 surveys were performed, and 138 taxa
were identified.
In the floristical composition of this association, we can observe a high partic-
ipation of species belonging to the higher syntaxa which subordinates the associ-
ation namely Petasition officinalis alliance (Spiraea chamaedrifolia, Petasites
hybridus, Carduus personata, Cirsium erisithales, Chaerophyllum hirsutum, etc.),
Petasito-Chaerophyletalia order (Anthriscus sylvestris, Aegopodium podagraria,
Galium aparine, Impatiens noli-tangere, Rumex obtusifolius) and Mulgedio-
Aconitetea class (Hypericum maculatum, Rosa pendulina, Silene dioica etc.).

Fig. 3 Telekio
speciosae-Aruncetum dioici at
Sovata (Ilieşi) (foto Silvia
Oroian)

Fig. 4 Telekio
speciosae-Aruncetum dioici at
Răstoliţa (foto Silvia Oroian)
Plant Associations of Petasition officinalis Alliance … 211

4. Petasitetum kablikiani Szafer et al. 1926

Southeast Carpathian high-weed communities of Petasites kablikianus form
enclaves on the sides of the mountain brooks on the montane beech and spruce
forest level. In Romania, from the Eastern Carpathians, this association has also
been described in Rodna Mountains, Bistriţa Aurie, Tazlăul Basin, Siriu Mountain,
Tiştiţa Gorges. They are spread at altitudes between 900 and 1280 m. Climatically,
phytocoenoses develop on temperatures between 5.8 and 4.5 °C with precipitation
of 900–1200 mm/year. They prefer folisoils with gravel and pebbles [12].
In the study area 5 phytosociological surveys were performed, identifying 65
taxons, at altitudes ranging from 900 to 1024 m (Table 1, column 4, Figs. 5, 6). The
characteristic and edifying species Petasites kablikianus achieve coverage of 77–
95%. The diagnostic species for association are Valeriana sambucifolia and
Petasites kablikianus. Other species with high constancy found in the floristic
composition are Petasites hybridus, Telekia speciosa, Geranium robertianum,
Aegopodium podagraria, Eupatorium cannabinum, Glechoma hederacea, Galium
aparine, Urtica dioica, Geranium phaeum, Rumex obtusifolius, Rumex alpinus.
The floristic composition of the association is completed by numerous species
belonging to higher syntaxa: Petasition officinalis alliance (Petasites hybridus,
Chaerophyllum hirsutum, Mentha longifolia, Myosotis scorpioides, Valeriana
sambucifolia, Carduus personata) and Petasito-Chaerophylletalia order
(Aegopodium podagraria, Galium aparine, Glechoma hederacea, Impatiens
noli-tangere, Rumex obtusifolius etc.) Species from Adenostyletalia order like
Achillea distans, Senecio ovatus are also present.

Fig. 5 Petasitetum kablikiani

at Lunca Bradului, Ilva Mare
(photo Silvia Oroian)

Fig. 6 Petasitetum kablikiani

(photo Silvia Oroian)
212 S. Oroian et al.

5. Arunco-Petasitetum albi Br.-Bl. et Sutter 1977

These coenoses meet on the montane level, from 698 to 1336 m altitude, occupying
larger areas in Răstoliţa: Borzia, on Răstoliţa Valey, Costeasa Valley, Iod Valley, at
Lăpușna on Secuieu Stream, Ibănești on Șirodul Mare Stream. A number of 99 taxa
were registered in 10 surveys. These tall-herbs are made up of species characteristic
of the mountain valleys: Petasites albus, Aruncus dioicus (Table 1, column 5,
Figs. 7, 8). Beside characteristic species of association, species belonging to
Petasition officinalis alliance (Telekia speciosa, Carduus personata, Cirsium eri-
sithales Chrysosplenium alternifolium Epilobium montanum Leucanthemum
waldsteinii, Mentha longifolia) as well as Petasito-Chaerophylletalia order
(Aegopodium podagraria, Impatiens noli-tangere, Lamium maculatum) record high
frequencies. We also mention the presence of many transgresive meso-hygrophilic
species from the Molinio-Arrhenatheretea class (incl. Molinietalia) grasslands, such
as Prunella vulgaris, Crepis biennis, Lysimachia nummularia, or from the vege-
tation of the forest areas with which the phytocoenoses often bind, from Fagetalia
order (Oxalis acetosella, Acer pseudoplatanus, Aconitum moldavicum ssp. mol-
davicum, Ajuga reptans, Dentaria glandulosa, Luzula luzuloides, Salvia glutinosa,
Stachys sylvatica etc.)

Fig. 7 Coenoses with

Petasites albus (photo Silvia
Oroian)

Fig. 8 Arunco-Petasitetum
albi (photo Silvia Oroian)
Plant Associations of Petasition officinalis Alliance … 213

Comparative analysis shows that all five associations grow on intermontane

valleys, on alluvial soils rich in humus, at altitudes between 500 and 1336 m, only
association Petasitetum kablikiani was identified at higher altitudes, between 905
and 1025 m, near forests.
The 5 associations identified are characteristic of the habitat: 6430 - Hydrophilous
tall-herb fringe communities of plains and in the montane to alpine levels [13]. It is a
habitat composed of tall herbaceous grass communities, diverse in terms of species
composition. The most representative communities of tall herbs (not to be confused
with the weed communities that are directly related to human activities) are those of
different species of butterbur, Petasites hybridus, Petasites albus, Petasites kab-
likianus, and also Telekia speciosa, Filipendula ulmaria, Cicerbita alpina, Digitalis
grandiflora, Calamagrostis arundinacea, Aruncus dioicus etc.
Considering the characteristic species to different coenotic groups, in all 5
associations there is a high share of the species belonging to the coenotaxa that
subordinate the associations Petasition officinalis alliance and Petasito-
Chaerophylletalia order (Aruncus dioicus, Cirsium oleraceum, Impatiens
noli-tangere, Leucanthemum waldsteinii, Poa trivialis, Petasites hybridus,
Petasites  reichingeri, Cirsium erisithales, Chaerophyllum hirsutum, Stellaria
nemorum, Mentha longifolia, Myosotis scorpioides, Rumex alpinus etc.), as well as
Mulgedio-Aconitetea class (Aconitum napellus ssp. tauricum, Hypericum macula-
tum, Rosa pendulina, Silene dioica) etc.
Characteristic species of the order Adenostyletalia and class Epilobietea
(Achillea distans, Cicerbita alpina, Deschampsia cespitosa, Gentiana asclepiadea,
Senecio ovatus, Cirsium waldsteinii, Doronicum austriacum) occur in the associ-
ations probably due the proximity of mountainous tall-weeds.
In the composition of studied phytocoenoses we mention species such as
Aconitum moldavicum, Dianthus barbatus ssp. compactus and Leucanthemum
waldsteinii. They are endemic or subendemic species and have a significant phy-
togeographical interest. Carpathian-Balkan species are also present; out of these, we
mention Campanula abietina, Petasites kablikianus, Pulmonaria rubra.

4 Data Analysis

Plant communities were ordered with Principal coordinate analysis (PCoA) based
on Jaccard (presence/absence) and Bray-Curtis (abundance-dominance) indices.
The main environmental gradients of the associations were analyzed with CCA
(Canonical Correspondence Analysis). For this analysis the average non-weighted
Ellenberg indicator values (for light, moisture, temperature, soil reaction, soil
nutrient content) and the Shannon-Wiener indices of the relevés were used. The
comparisons of the ecological indicator values and diversity indices were made
with the Kruskal-Wallis test. The significance of the differences was checked with
the Mann-Whitney U post hoc test. All statistics were performed with the Past
Program (v. 2.17 [15]).
214 S. Oroian et al.

5 Results and Discussion

Plant communities could not be differentiated with the PCoA based on the presence/
absence of the species (Fig. 9) (PC1: 9.70%, PC2: 6.30%). The ordering of data on
the basis of the Bray–Curtis index presented the grouping in two well differentiated
associations: Petasitetum kablikiani and Arunco-Petasitetum albi (PC1: 32.18%,
PC2: 14.27%). The other three associations sharing many plant taxa were not
differentiated (Fig. 10).
The ordering of plant communities (Eigenvalues for CCA1 and CCA2: 0.346
and 0.159, respectively) on basis of ecological indicator values and the
Shannon-Wiener indices presented the grouping in 5 different associations
(Fig. 11). No correlations were found between the six analyzed gradients. The first
CCA axis explained 51.03% variance of the species data, and the second axis
23.49%. The first axis was negatively correlated with light indicator values (cor-
relation coefficient: −0.7633), while the second axis was positively correlated with
moisture (correlation coefficient: 0.5239). Thus, the main environmental gradients
in ordering of the studied plant communities were the light and the moisture.
Communities that require intense light are located in the left side of the chart
(Telekio-Filipenduletum), while those developed on shady places with northern
exposure, on the right side (Telekio speciosae-Aruncetum dioici). On the upper part
of the ordination chart are the relevés of Petasitetum kablikiani edified on wet al-
luvial soils, and some of the relevés of mesophilous, meso-hygrofilous and
hygrofilous plant communities (Arunco-Petasitetum albi, Telekio-Petasitetum
hybridi, Telekio-Filipenduletum).

Fig. 9 PCoA of the studied plant communities based on the Jaccard index. Associations are
marked as follows: Petasitetum kablikiani; Arunco-Petasitetum albi; Telekio
speciosae-Aruncetum dioici; Telekio-Filipenduletum; Telekio-Petasitetum hybridi
Plant Associations of Petasition officinalis Alliance … 215

Fig. 10 PCoA of the studied plant communities based on the Bray–Curtis index. Associations are
marked as follows: Petasitetum kablikiani; Arunco-Petasitetum albi; Telekio
speciosae-Aruncetum dioici; Telekio-Filipenduletum; Telekio-Petasitetum hybridi

2.5

Moisture 2.0

1.5

1.0 Nutrients (N)

Axis 2

Reaction
Light 0.5

-4.0 -3.2 -2.4 -1.6 -0.8 0.8 1.6 2.4

Temperature
-0.5

-1.0

-1.5
Shannon-Wiener index
-2.0
Axis 1

Fig. 11 Canonical correspondence analysis (CCA) of the studied plant communities. The average
non-weighted Ellenberg indicator values and the Shannon-Wiener indexes of the relevés were
plotted on the CCA ordination chart. Associations are marked as follows: Petasitetum
kablikiani; Arunco-Petasitetum albi; Telekio speciosae-Aruncetum dioici;
Telekio-Filipenduletum; Telekio-Petasitetum hybridi
216 S. Oroian et al.

Fig. 12 Ellenberg indicator values and species richness in the studied phytocoenoses. Different
letters signify p < 0.05 (Pk Petasitetum kablikiani; A-P Arunco-Petasitetum albi; T-A Telekio
speciosae-Aruncetum dioici; T-F Telekio-Filipenduletum; T-P Telekio-Petasitetum hybridi)
Plant Associations of Petasition officinalis Alliance … 217

Fig. 13 Dianthus barbatus

ssp. compactus (photo Silvia
Oroian)

Fig. 14 Angelica
archangelica (photo Mihaela
Sămărghiţan)

There were significant differences between the studied plant communities

regarding the Ellenberg indicator values and the species number of the relevés
(100 m2) (Fig. 12). The highest values (significantly different from the other
studied communities) for light intensity were obtained in case of the association
Telekio-Filipenduletum and Telekio-Petasitetum hybridi, both encountered at higher
altitudes. In case of moisture, high values were found in the association Petasitetum
kablikiani (significantly higher than the associations with Telekia speciosa)
encountered in mountain valleys, at the level of beech forests. The association
Arunco-Petasitetum albi, specific for mountain valleys situated at lower altitudes,
presented also high values for moisture. The association Telekio-Filipenduletum
218 S. Oroian et al.

had the lowest indicator values for temperature and also for soil reaction. The
demands for nutrients (N) were significantly lower for the association Telekio-
Filipenduletum and Telekio-Petasitetum hybridi compared to the association
Petasitetum kablikiani. The association Telekio-Petasitetum hybridi had signifi-
cantly higher species number/relevés than the other communities, but outside the
association Telekio-Filipenduletum.
Analyzing the conservation status of habitat, we conclude that almost all the
described associations are in very good conservation status and presenting several
endemic and protected species (Figs. 13 and 14).
Although the conservative value of the 5 plant associations, which at their origin
have coenoses that spread on the edge of forests and along watercourses is low, they
have an ecotonal character that provides shelter for many animal species, shelter a
very wide range of invertebrates but also is a habitat representing a feeding place
for many small and large mammal species, therefore careful protection is a
necessity. With their rich biomass they often complement the habitat with alluvial
forests (91E0*) and its role as an ecological corridor.
In the phytocoenoses of 4 plant associations: Telekio-Petasitetum hybridi,
Telekio-Filipenduletum, Telekio speciosae-Aruncetum dioici and Arunco-
Petasitetum albi, were noted some invasive species such as: Erigeron annuus,
Galinsoga parviflora and Rudbeckia laciniata but they don’t have a significant impact
on phytocoenoses’ short term evolution.

Table 1 Plant associations of Petasition officinalis

T-P T-F T-A PK A-P

Association number 1 2 3 4 5
Relevés number 38 8 10 5 8
650- 500- 905- 525-
Altitude (m.s.m) 660-1215
1063 926 1025 1336
Number of species 244 132 138 65 99
Coverage (%) 87-100% 80-100 90-100 82-94 82-100

Dominant species (Characteristic

species)
1. Telekia speciosa V V V III V
2. Filipendula ulmaria V V I IV -
3. Cirsium oleraceum III II II II -
4. Carduus personata III II I IV I
Plant Associations of Petasition officinalis Alliance … 219

Table 1 (continued)

Diagnostic species
5. Myosoton aquaticum II III - - -
6. Scirpus sylvaticus II IV - II -
7. Gentiana asclepiadea I - III I -
8. Spiraea chamaedryfolia I - III - -
9. Aruncus dioicus I II V - V
10. Campanula trachelium - - II - -
11. Pulmonaria rubra I - III - II
12. Valeriana sambucifolia I II I IV -
13. Petasites kablikianus I - - V -
14. Scopolia carniolica - - - - III
15. Petasites albus I I I - V
16. Cardamine pratensis I - - - -
Petasition officinalis et Petasito-
Chaerophylletalia - - - - -
17. Aegopodium podagraria III IV I II III
18. Angelica sylvestris I II II - -
19. Anthriscus sylvestris I III - - -
20. Chaerophyllum aromaticum I - II - -
21. Chaerophyllum hirsutum IV II I IV IV
22. Chrysosplenium alternifolium I - - - IV
23. Cirsium erisithales I II I - II
24. Galium aparine II IV III II -
25. Geranium phaeum I - - II -
26. Geum rivale - - - - I
27. Glechoma hederacea II - III II -
28. Impatiens noli-tangere IV II III IV IV
29. Lamium maculatum I - - - II
30. Leucanthemum waldsteinii II II I II II
31. Mentha longifolia IV IV II IV I
32. Myosotis scorpioides IV IV I II II
33. Petasites hybridus V III II IV II
34. Poa trivialis I II I - I
35. Rumex alpinus I IV I III I
36. Rumex obtusifolius III II - II -
37. Stellaria nemorum III II I IV IV
Adenostyletalia
38. Achillea distans II II II II I
39. Angelica archangelica I - - - -
40. Cicerbita alpina I II - - -
41. Cirsium waldsteinii IV III II - I
220 S. Oroian et al.

Table 1 (continued)

42. Deschampsia cespitosa III IV I II III

43. Doronicum austriacum I II - - II
44. Senecio ovatus II II I II -
45. Silene vulgaris I - - - -
Mulgedio-Aconitetea
46. Aconitum napellus ssp. tauricum - - - II -
47. Hypericum maculatum III II I II III
48. Rosa pendulina - I I - -
49. Silene dioica I II I - I
50. Veratrum album ssp. lobelianum I I - - -
Geo urbani-Alliarion officinalis
(syn. Galio-Alliarion)
51. Alliaria petiolata I - - - II
52. Campanula rapunculoides I - I - -
53. Carduus crispus - - I - -
54. Chelidonium majus I - - - -
55. Clematis vitalba - - I - -
56. Geranium robertianum III II III IV IV
57. Mycelis muralis II II II - II
58. Rubus caesius II I II I -
Senecionion fluviatilis - - - - -
59. Heracleum spondylium II I II - -
60. Lamium album I - I - II
Epilobietea angustifolii
(incl.Epilobion angustifolii) - - - - -
61. Athyrium filix-femina IV IV IV III IV
62. Calamagrostis arundinacea II III III - -
63. Carex hirta I - - - -
64. Chamerion angustifolium II II I II I
65. Digitalis grandiflora I - II - I
66. Dryopteris filix-mas I - I - III
67. Eupatorium cannabinum III II I I II
68. Fragaria vesca II II I - I
69. Galeopsis speciosa II II - - -
70. Galeopsis tetrahit III II II II II
71. Galium odoratum - - - - III
72. Juncus tenuis I - - - -
73. Populus tremula I - - - -
74. Rubus idaeus III IV IV IV III
75. Scrophularia nodosa I - I I II
76. Senecio nemorensis III II - III I
77. Solanum dulcamara I - I - -
Plant Associations of Petasition officinalis Alliance … 221

Table 1 (continued)

Sambuco-Salicion - - - - -
78. Salix caprea III III III I II
79. Sambucus nigra I - I - -
80. Sambucus racemosa - I - II II
81. Urtica dioica V III IV V III
Arction lappae - - - - -
82. Arctium lappa I - - - II
83. Tanacetum vulgare I - - - -
Molinietalia (incl.Molinion) - - - - -
84. Agrostis stolonifera I II - II -
85. Carex distans I - - - -
86. Cirsium palustre I I - II -
87. Equisetum palustre II II II - -
88. Euphrasia rostkoviana I - - - -
89. Galium uliginosum II II I - I
90. Juncus effusus III IV - - -
91. Lathyrus pratensis I - I - -
92. Lychnis flos-cuculi I II - - I
93. Lycopus europaeus I - - II -
94. Lythrum salicaria III II I I I
95. Mentha × verticillata I - - - -
96. Potentilla erecta II I - - -
97. Rumex crispus I - - - -
98. Stachys officinalis I - - - -
99. Stellaria graminea III - I - -
100. Symphytum officinale I - I - I
101. Trifolium hybridum III - I II -
Filipendulion - - - - -
102. Valeriana officinalis III III II - -
103. Veronica chamaedrys I I - - -
104. Vicia sepium I - I - -
Molinio-Arrhenatheretea - - - - -
105. Achillea millefolium II II - - -
106. Agrostis capillaris II II I - -
107. Alchemilla xanthochlora I II I - -
108. Campanula glomerata I I I - -
109. Campanula abietina I I I II II
110. Campanula patula ssp. patula II I I - -
111. Centaurea phrygia. ssp. phrygia I I - - -
112. Cerastium holosteoides II IV I II -
113. Crepis biennis I II - - I
222 S. Oroian et al.

Table 1 (continued)

114. Cynosurus cristatus I - I - -

115. Dactylis glomerata IV III I IV I
116. Elymus repens I - - - -
117. Equisetum arvense I - II - III
118. Holcus lanatus I II - - -
119. Lapsana communis I - I - -
120. Leucanthemum vulgare II - III I I
121. Lotus corniculatus II II I - -
122. Lysimachia nummularia IV IV III III II
123. Lysimachia vulgaris II II I II -
124. Medicago lupulina I II - - I
125. Ranunculus repens IV IV I III IV
126. Plantago lanceolata I II I - -
127. Plantago major III II I - -
128. Poa palustris I - - - -
129. Poa pratensis II II I II -
130. Potentilla reptans I - - - -
131. Prunella vulgaris IV III II III IV
132. Taraxacum officinale - - - - II
133. Trifolium repens III III II - II
134. Vicia cracca I - - - -
Alno-Ulmion - - - - -
135. Brachypodium sylvaticum II - I - -
136. Circaea lutetiana I II - III -
137. Elymus caninus II II II I II
138. Epilobium montanum III II II IV II
139. Equisetum hyemale I - - - -
140. Equisetum pratense - - II - -
141. Equisetum sylvaticum II II I - -
142. Equisetum telmateia I - - - -
143. Festuca gigantea I II II - -
144. Listera ovata - - I - -
145. Matteuccia struthiopteris II - - I III
146. Primula elatior - - I - -
147. Rumex sanguineus I I - - -
148. Salix alba I - - - -
149. Thalictrum aquilegiifolium I I - - -
Fagetalia - - - - -
150. Acer pseudoplatanus I II - - -
151. Aconitum moldavicum I - - - -
152. Ajuga reptans I II II - -
Plant Associations of Petasition officinalis Alliance … 223

Table 1 (continued)

153. Anemone nemorosa - - - - I

154. Anemone ranunculoides - - - - I
155. Asarum europaeum - - II - II
156. Carex sylvatica I - - - -
157. Daphne mezereum I - - - -
158. Dentaria glandulosa - - - - II
159. Euphorbia amygdaloides I - - - II
160. Fagus sylvatica I - - - -
161. Fraxinus excelsior I - - - -
162. Galium schultesii - - II - -
163. Luzula luzuloides I I II I II
164. Luzula sylvatica II I - - -
165. Mercurialis perennis - - II - -
166. Oxalis acetosella I - II - III
167. Pulmonaria officinalis - - I - -
168. Salvia glutinosa II IV III III -
169. Stachys sylvatica IV IV II IV -
170. Veronica urticifolia - - II - II
171. Viola riviniana - - I - -
Variae Syntaxa - - - - -
172. Alnus incana I - - - -
173. Alopecurus pratensis I - - - -
174. Anthriscus cerefolium I - I - -
175. Arrhenatherum elatius I - - - -
176. Artemisia vulgaris I - - I -
177. Asplenium trichomanes - - I - -
178. Astragalus glycyphyllos I - I - I
179. Bidens tripartita I - - I -
180. Bromus commutatus I - - - -
181. Calamagrostis pseudophragmites I - - III I
182. Caltha palustris ssp. laeta I I - - II
183. Calystegia sepium II II I II -
184. Campanula bononiensis I - - - -
185. Cardamine impatiens I - - - II
186. Carex leporina I II - - -
187. Carex pallescens I - - - -
188. Carex remota I - - - II
189. Carex vulpina I II - - -
190. Catabrosa aquatica I - - - -
Centaurea biebersteinii
191.
subsp. biebersteinii - I - - I
224 S. Oroian et al.

Table 1 (continued)

192. Centaurea nigrescens I - - - -

193. Centaurea pseudophrygia I II - - -
194. Chenopodium album I - - - -
195. Cichorium intybus I - - - -
196. Cirsium arvense II II I - -
197. Cirsium canum I II - - -
198. Cirsium vulgare I - - - -
199. Clinopodium vulgare II - II - -
200. Convolvulus arvensis I - - - -
201. Cornus sanguinea - - I - -
202. Coronilla varia I - - - -
203. Corylus avellana I - - - II
204. Crataegus monogyna I - - - -
205. Crepis paludosa I - - - -
206. Cruciata glabra I I - - -
207. Cruciata laevipes I - - - -
208. Cystopteris fragilis - - I - -
209. Dianthus barbatus ssp. barbatus I - - - II
210. Dianthus barbatus ssp. compactus I - II - -
211. Dianthus carthusianorum - II - - -
212. Dipsacus pilosus I - - - -
213. Epilobium hirsutum I - - - -
214. Epilobium palustre I - I - -
215. Epilobium parviflorum I - I - -
216. Erigeron annuus II IV I - I
217. Euphorbia carniolica - I I - -
218. Euphorbia epithymoides II - - - I
219. Festuca pratensis - I - - -
220. Galinsoga parviflora I - - - -
221. Galium mollugo I I - - -
222. Galium verum I - I - -
223. Geranium palustre I - - - -
224. Geum urbanum I - I - -
225. Glechoma hirsuta I - II - I
226. Glyceria fluitans I I - - -
227. Glyceria maxima I - - - -
228. Glyceria notata I - - II -
229. Gnaphalium sylvaticum I II - - -
230. Gymnocarpium robertianum I - I - -
231. Hieracium aurantiacum - I - - -
232. Hieracium umbellatum I - - - I
Plant Associations of Petasition officinalis Alliance … 225

Table 1 (continued)

233. Homogyne alpina - - - - I

234. Hypericum montanum - - I - -
235. Hypericum perforatum I I I - II
236. Hypericum tetrapterum I - - - -
237. Lathyrus tuberosus I - - - -
238. Lonicera xylosteum - - I - -
239. Lychnis viscaria I - - - -
240. Lysimachia punctata I - - - -
241. Matricaria perforata I - - - -
242. Medicago minima I I - - -
243. Melilotus albus I - - - -
244. Mentha aquatica I - - - -
245. Mentha arvensis I I I - II
246. Origanum vulgare I - - - I
247. Petasites × reichingeri - - - I -
248. Phalaris arundinacea I - - - -
249. Phegopteris connectilis - - I - -
250. Phleum montanum I - - - -
251. Phleum phleoides I - - - -
252. Picea abies I II I - -
253. Plantago media I II I - III
254. Poa nemoralis I - I - I
255. Polygonum hydropiper I - - - -
256. Polygonum lapathifolium I - - - -
257. Polygonum mite I I - - -
258. Polypodium vulgare - - I - -
259. Rhinanthus rumelicus I - - - -
260. Rorippa pyrenaica I - - - -
261. Rorippa sylvestris I - I - -
262. Rubus hirtus I II - - -
263. Rudbeckia laciniata I - - - -
264. Salix cinerea I - - - -
265. Salix fragilis I - - - -
266. Salix purpurea I - - - -
267. Saponaria officinalis I I I - -
268. Scutellaria galericulata I - I - I
269. Sedum maximum - - I - -
270. Senecio erucifolius - - - - I
271. Senecio vulgaris I - - - -
272. Seseli libanotis - - - - I
273. Silene armeria - - - - I
226 S. Oroian et al.

Table 1 (continued)

274. Silene latifolia ssp. alba I - - - II

275. Silene nutans ssp. dubia - - I - I
276. Stellaria media I - - - -
277. Trifolium alpestre - II - - -
278. Trifolium campestre I - I - -
279. Trifolium medium I - - - -
280. Trifolium montanum - I - - -
281. Trifolium pratense I IV - - -
282. Tussilago farfara IV IV II - IV
283. Valeriana tripteris - - I - -
284. Veronica beccabunga I - - - -
285. Veronica officinalis II II I - II
286. Veronica teucrium I II - - -
287. Vicia sylvatica I I - - -
288. Viola arvensis I I - - -
289. Viola tricolor - II - - -

Date and location of relevés:

1. Telekio-Petasitetum hybridi (Morariu 1967) Resmeriţă et Raţiu 1974: R1 -
Ciobotani, Mermezeu (15.07.2014); R2 - Câmpul Cetății, valea Nirajul Mare
(01.07.2014); R3 - Lăpușna (05.07.2014); R4 - Lăpușna, pârâul Secuieu
(07.07.2014); R5 - Lăpușna, pârâul Negru (07.07.2014); R6 - Ibănești, Șirodul
Mare (10.07.2014); R7 - Câmpul Cetății, valea Nirajul Mare (10.07.2014); R8 -
Câmpul Cetății, valea Nirajul Mare (15.07.2014); R9 - Sovata, Ilieși
(18.07.2014); R10 - Răstolița, Valea Răstoliței (19.07.2014); R11 - Bistra
Mureșului, Valea Bistrei (23.07. 2014); R12 - Bistra Mureșului, Valea Bistrei
(23.07. 2014); R13 - Răstolița, Gălăoaia (25.07. 2014); R14 - Răstolița,
Gălăoaia (25.07. 2014); R15 - Răstolița, Gălăoaia (25.07. 2014); R16 -
Răstolița, Valea Vișa (28.07. 2014); R17 - Răstolița, Valea Iodului (30.07.
2014); R18 - Răstolița, Valea Iodului (30.07. 2014); R19 - Sovata, valea Sebeș
(04.08. 2014); R20 - Sovata, valea Sovata (06.08. 2014); R21 - Sovata, valea
Sovata (06.08. 2014); R22 - Sovata, Târnava Mică (06.08.2014); R23 - Lunca
Bradului, Ilișoara Mare (11.08.2014); R24 - Lunca Bradului, Ilișoara Mică
(11.08. 2014); R25 - Lunca Bradului, Ilva Mare (11.08.2014); R26 - Stânceni,
Gudea Mare (12.08.2014); R27 - Stânceni, Gudea Mică (12.08.2014); R28 -
Răstolița, Valea Rusu (13.08.2014); R29 - Sălard, Hidegag (16.08.2014); R30 -
Sălard, Șolea (16.08.2014); R31 - Sălard, valea Sălard (16.08.2014); R32 -
Răstolița, Borzia (17.08.2014); R33 - Sălard, Țâba Mică (18.08.2014); R34 -
Sălard, Valea Belciu (18.08.2014); R35 - Sălard, Jirca (18.08.2014); R36 -
Neagra, pârâul Neagra (23.08.2014); R37 - Ibănești, Fâncel (26.08.2014); R38 -
Sălard, Pârâul Wagner (26.08.2014); R.39 - Ciobotani, Zebrac (16.09.2014).
Plant Associations of Petasition officinalis Alliance … 227

2. Telekio-Filipenduletum Coldea 1996: R1 - Câmpul Cetății, valea Nirajul Mare

(01.07.2014); R2 - Câmpul Cetății, valea Nirajul Mare (12.07.2014); R3 -
Răstolița, Valea Răstoliței (19.07.2014); R4,5 - Bistra Mureșului, valea Bistrei
(23.07.2014); R6- Sălard, valea Hidegag (16.08.2014); R7 - Sălard, valea Șolea
(16.08.2014); R8 - Stânceni, Ciobotani (16.09.2014).
3. Telekio speciosae-Aruncetum dioici Oroian 1998: 1-Răstolița, Peșcoasa Mare,
06.07.2014; R:2,3 - Sovata-pe valea Târnava Mică, Ilieși, 18.07.2014; R4:
Bistra Mureșului, valea Bistrei (23.07.2014); R5 - Lunca Bradului Ilișoara Mare
(11.08.2014); R6,7 - Răstolița,Borzia (07.2000); R 8,9,10 –Răstolița, Valea
Costeasa (07.2000).
4. Petasitetum kablikiani Szafer et al. 1926: R1 - Lunca Bradului, Ilișoara Mare
(11.08.2014); R2,3 - Lunca Bradului, Ilva Mare (11.08.2014); R4,5 - Stânceni,
Gudea Mare (12.08.2014).
5. Arunco-Petasitetum albi Br.-Bl. et Sutter 1977: R1 - Răstolița, Valea Răstoliței
(19.07.2014); R2 - Lăpușna, Pârâul Secuieu (07.07.2014); R3 –Ibănești, Șirodul
Mare (10.07.2014); 4-Răstolița “Costeasa” (2000.06.15); 5-Răstolița, Iod
(2000.06.17); 6-Răstolița “Costeasa” (2001.07.11); 7,8 -lunca Mureșului între
Răstolița și Borzia (2001.07.28).

6 Conclusions

A data set of 70 relevés were analysed. They were recorded in five associations
belonging to Petasition officinalis alliance.
If we compare the 5 plant associations, from the floral point of view, the coe-
noses resemble but are distinguished by the abundance of characteristic species.
It is worth mentioning the significant percentage of Alpine-Carpathian,
Carpathian and Carpathian-Balkan elements, which give a particular aspect to this
area, conferring phytocenoses a distinct regional peculiarity of that of the vicarians
in Central Europe, demonstrating the intensity of the process of differentiation and
preservation from the flora of tertiary and pleistocene origin and highlights the close
florogenetic connections between the flora of this Carpathian area, strictly geo-
morphologically delimited and that of the Balkan mountains. Among these we
mention: Petasites kablikianus, Leucanthemum waldsteinii, Achillea distans,
Aconitum napellus ssp. tauricum, Telekia speciosa, Scopolia carniolica, Cirsium
waldsteinii, Pulmonaria rubra, Silene nutans ssp. dubia.
We report the presence in the composition of plant associations of some
endangered species found in the Red Lists: Campanula abietina (in all associations)
and Angelica archangelica (association 1). A remarkable phytogeographical
interest has the endemic or subendemic species: Aconitum moldavicum (associa-
tion1), Dianthus barbatus ssp. compactus (associations 1,3), Cirsium waldsteinii
(associations 1, 2, 3, 5) Leucanthemum waldsteinii (in all 5 associations) and Silene
nutans ssp. dubia (associations 3, 5). There are also Carpathian-Balkan species
(Campanula abietina, Petasites kablikianus, Pulmonaria rubra).
228 S. Oroian et al.

References

1. Bilz M, Kell SP, Maxted N, Lansdown RV (2011) European red list of vascular plants.
Publications Office of the European Union, Luxembourg
2. Borza A, Boşcaiu N (1965) Introducere în studiul covorului vegetal, Ed. Acad. R.P.R.,
Bucureşti
3. Boşcaiu N, Coldea G, Horeanu C (1994) Lista roşie a plantelor vasculare dispărute,
periclitate, vulnerabile şi rare din flora României. Ocrot. Nat. Med. Înconj. 38(1):45–56
4. Chytrý M (ed) Vegetace České republiky. 2. Travinná a keříčková vegetace Academia, Praha
5. Coldea G (1990) Munţii Rodnei. Studiu geobotanic. Ed. Acad. Bucureşti, p 183
6. Coldea G (1991) Prodrome des associations vegetales des Carpates du sud-est (Carpates
Roumaines). Doc Phytosociol Camerino 13:317–359
7. Coldea G (2012) Les association végétales de Roumanie, T. 2 (Les associations
anthropogènes), Edit. Presa Universitară Clujeană, Cluj Napoca
8. Cristea V, Gafta D, Pedrotti F (2004) Fitosociologie. Edit. Presa Universitară Clujeană,
Cluj-Napoca
9. Danciu M (1973) Contribuţii la studiul asociaţiilor alianţei Filipendulo-Petasition din sudul
munţilor Baraolt. Bul. Univ. Braşov. Seria C. Matematică-Fizică-Chimie-Şti. Nat. Braşov
15:123–131
10. Dihoru G, Negrean G (2009) Cartea roşie a plantelor vasculare din România, Edit. Acad.
Române, Bucureşti, p 630
11. Dihoru GH, Dihoru A (1993–1994) Plante rare, periclitate şi endemice în flora României –
Lista Roşie, Acta Botanica Horti Bucurestiensis, Bucureşti, pp 173–179
12. Doniţă N, Popescu A, Paucă-Comănescu M, Mihăilescu S, Biriş I-A (2005) Habitate din
România, Edit. Tehnică Silvică Bucureşti
13. Gafta D, Mountford O (coord.) (2008) Manual de interpretare a habitatelor Natura 2000 din
România, Edit. Risoprint, Cluj-Napoca
14. Hájková P, Hájek M (2007) Calthion palustris. In: Chytrý M (ed) Vegetace České republiky.
1. Travinná a keříčková vegetace Academia, Praha, pp 238–280
15. Hammer Ø, Harper DAT, Ryan PD (2001) PAST: paleontological statistics software package
for education and data analysis. Palaeontologia Electronica 4(1):9
16. Jarolimek I, Kliment J, Valachovič M (2002) The syntaxonomicalrevision of the riparian plant
communities dominated byPetasites hybridus in Slovakia. Biologia 57:471–492
17. Kliment J, Jarolimek I (2002) Syntaxonomical revision of the Petasites kablikianus
communities (Petasition officinalis) in the West Carpathians. Biologia 57:101–118
18. Kliment J, Valachovic M (eds) (2007) Rastlinné spolocenstvá Slovenska. 4. Vysokohorská
vegetátia. Veda, Bratislava
19. Kliment J, Šibík J, Šibíková I, Jarolímek I, Dúbravcová Z, Uhlířová J (2010) High-altitude
vegetation of the Western Carpathians – a syntaxonomical review. Biologia 65/6:965–989,
Section Botany. https://doi.org/10.2478/s11756-010-0109-4
20. Morariu I (1967) Clasificarea vegetatţiei nitrofile din România.Contr. Bot. 1967: 233–246
21. Mucina L (1993) Galio-urticetea. In: Mucina L, Grabherr G, Ellmauer T (eds) Die
Pflanzengesellschaften Österreichs. Teil I, Gustav Fischer Verlag, Jena, DE, pp 203–251
22. Mucina L, Bültmann H, Dierßen K, Theurillat J‐P, Raus T, Čarni A, Šumberová K,
Willner W, Dengler J, Tichý L (2016) Vegetation of Europe: hierarchical floristic
classification system of vascular plant, bryophyte, lichen and algal communities. Appl Veg
Sci 19(Suppl. 1):3–264
23. Nechita N (2003) Flora şi vegetaţia cormofitelor din Masivul Mășmaș, Cheile Bicazului și
Lacul Roșu, Ed. “Constantin Matasăˮ Piatra Neamț
24. Oltean M, Negrean G, Popescu A, Roman N, Dihoru G, Sanda V, Mihăilescu S (1994) Listă
roşie a plantelor superioare din România, Studii, sinteze, documentaţii de ecologie, 1, Acad.
Română, Instit. de Biologie, Bucureşti
Plant Associations of Petasition officinalis Alliance … 229

25. Oroian S (1998) Flora şi Vegetaţia Defileului Mureşului între Topliţa şi Deda, Casa de Editură
Mureş
26. Pott R (1995) Die Pfl anzengesellschaften Deutschlands. 2nd edn. Ulmer, Stuttgart, p 622
27. Resmeriţă I, Raţiu O (1974) Vegetaţia higro-şi hidrofilă din Maramureş. Contrib. Bot.,
Cluj-Napoca, pp 115–129
28. Sanda V, Popescu A, Barabaş N (1997) Cenotaxonomia şi caracterizarea grupărilor vegetale
din România. St. Şi Com. Muz. Şt. Nat. Bacău, Biol. Veget., 14: 365 p.
29. Sanda V, Popescu A, Stancu DI (2001) Structura cenotică şi caracterizarea ecologică a
fitocenozelor din România, Ed. Conphis, Piteşti
30. Sămărghiţan M (2005) Flora şi vegetaţia Văii Gurghiului, Ed. University Press, Târgu-Mureş
31. Sârbu I, Ştefan N, Oprea A (2013) Plante vasculare din România, determinator ilustrat de
teren. Edit. VictorBVictor, Bucureşti
Arthrocnemum macrostachyum plant
communities in the Iberian Peninsula,
Balearic and Canary Islands
(Spain and Portugal)

Esteban Ramírez, Lourdes Rufo, Daniel Sánchez-Mata,

Irene Sánchez-Gavilán, and Vicenta de la Fuente

Abstract We present a synthesis of the halophilous fruticose vegetation structured

by Arthrocnemum macrostachyum on the Iberian Peninsula, Balearic Islands
and Canary Islands (Arthrocnemo macrostachyi-Suaedetalia braun-blanquetii,
Salicornietea fruticosae). The analysis and study of over 200 selected phytosoci-
ological relevés supports the conclusions reflected in the syntaxonomical checklist
of the units recognised in the study area. The floristic appendix also contains
taxonomic and floristic-chorological novelties.

Keywords Arthrocnemion macrostachyi Salicornietea fruticosae
Halophytic



vegetation Iberian Peninsula vegetation Saline soils habitats

1 Introduction

Plants adapted to live in saline environments, a group known as halophytic plants,

are found along coastlines and inland salt marshes around the world [43]. The main
ecological factors that determine the habitat type and limit the development of
certain species in these environments are the high concentration of salts present in
the soils, and the daily and seasonal variation in the degree of soil waterlogging due
to tidal action along the coast and the rainfall regime in inland environments [4]. To
respond to these peculiar, extreme environmental conditions, species in the family

E. Ramírez
I. Sánchez-Gavilán
V. de la Fuente

Department of Biology (Botany), Faculty of Sciences, Autonomous University of Madrid,
Cantoblanco, 28049 Madrid, Spain
L. Rufo
Department of Pharmacy, Faculty of Experimental Sciences, Francisco de Vitoria University,
Pozuelo de Alarcón, 28223 Madrid, Spain
D. Sánchez-Mata (&)
Department of Pharmacology, Pharmacognosy and Botany (Botany Unit), Faculty of
Pharmacy, Complutense University, 28040 Madrid, Spain
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 231
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_12
232 E. Ramírez et al.

Chenopodiaceae (subfamily Salicornioideae) are among the lineages with the lar-
gest proportion of genera that mainly form plant communities in saline environ-
ments. Of particular note are the communities formed by succulent therophytes
belonging to the phytosociological class Therosalicornietea (mainly species from
the genera Halopeplis, Microcnemum and Salicornia), and/or by succulent
chamaephytes [mainly species from the genera Sarcocornia (hereafter S.),
Halocnemum and Arthrocnemum (hereafter A.)] in the class Salicornietea fruti-
cosae [5, 13, 19, 23–26, 36, 40–42, 44, 49–52]. This last class is characterised by
permanent halophilous fruticose communities and includes vegetation units struc-
tured by Arthrocnemum macrostachyum (hereafter A. macrostachyum) in territories
of Spain and Portugal reviewed by [49, 50, 52].
The hierarchical classification and prior knowledge of these communities are
essential prerequisites for determining, managing and protecting their diversity.
These are also fundamental instruments for interpreting the landscape in refor-
estation programs and for recovery of vegetation cover [14, 33]. Halophilous fru-
ticose communities (Salicornietea fruticosae) and their characteristic biotopes are
included in Annex I of the Habitats Directive 92/43/EEC [17].
From knowledge of the taxonomy, ecology and habitats of the subfamily
Salicornioideae in the Mediterranean [23–26, 42], we undertake and assess the
diversity of the halophilous communities structured by A. macrostachyum and
contribute a proposal for their syntaxonomical classification, based on analysis of
phytosociological data by means of K-means clustering.

2 Nomenclature and References

The taxonomic nomenclature follows the Euro+Med database [16], exceptions are
included in the floristic appendix with their recognized authority. The main refer-
ences were the biogeographical units of Western Mediterranean Europe synthesized
by Rivas-Martínez et al. [53] and the syntaxonomy of the Salicornietea fruticosae
class proposed by Rufo et al. [26].

3 Material and Methods

3.1 Data Sources

The diversity of halophytic vegetation structured by A. macrostachyum on the

Iberian Peninsula and Canary and Balearic Islands was analysed using phytosoci-
ological relevés collected between 1958 and 2019 from the Iberian and
Macaronesian Vegetation Information System [21, 22], from other bibliographical
Arthrocnemum macrostachyum plant communities ... 233

sources, and from our own unpublished fieldnotes. Relevés compiled according to
the phytosociological methodology of the Zürich-Montpellier school [7] were
recovered from the following vegetation studies: [1–3, 8–12, 23, 28, 30, 32, 34, 35,
37–39, 44, 47, 48, 54, 56–58].
The total number of relevés was 293 initially. We screened the data manually to
exclude duplicate and heterogeneous relevés (those that included species indicating
a mixture of ecologically different plant communities) and those where the identity
and presence of some species were unclear. This reduced the number of relevés to
225. Taxa determined to supraspecific, sensu lato, and uncertain taxa were elimi-
nated. Synonyms were merged and taxa present in only one or two relevés were
eliminated. After this screening, the number of taxa to analyse was 89.

3.2 Data Analysis

The following protocol was used to assess the halophytic vegetation in the terri-
tories studied. The data matrix was stratified by biogeographic sectors [53], then
subjected to a heterogeneity-constrained random (HCR) re-sampling using the
Bray–Curtis dissimilarity index [31]. The number of relevés selected was propor-
tional to beta diversity within strata, based on the Bray-Curtis index [61], and the
number of relevés to be re-sampled was between 10 and 40 per stratum. The matrix
obtained includes 187 relevés and 89 species, and these data were analysed by the
K-means method using the Hellinger Distance as a dissimilarity measure, Ward's
hierarchical clustering method and a minimum of two and a maximum of seven
groups. Finally, a synthetic table was created with the groups generated and rear-
ranged based on the percentage of present and faithful species contributed by the
K-means analysis. A synthesis of the initial and screened relevés, and those
remaining for analysis, is given in Table 1.
Data editing and K-means analysis were carried out with the Quercus edition
module in the management package Vegana for ecological data [20] and re-sampled
with the JUICE 7.0 program [59]. The methodology was based on our previous
study focused on plant communities of Sarcocornia spp. [26], with some variations
adapted to the new data matrix.
234 E. Ramírez et al.

Table 1 References used in References Table Initial Filtered Analysed

the analysis, with the number
of tables consulted, number of [1] 12 5 5 5
initial relevés, number of [2] 2.5 5 5 4
relevés remaining after [3] 34 4 4 4
filtering, number of relevés [6] _ 12 11 11
remaining for analysis and the
[8] 11 8 4 4
global sum of each column
[9] 20 16 _ _
[10] 9 10 7 5
[11] 10 17 16 11
[12] 7 19 17 15
[18] 1 15 13 10
[23] 3 5 5 4
[28] 11 5 4 4
[30] 01 3 3 3
[32] 4 4 4 4
[34] 9 14 10 10
[35] 2 8 4 4
[37] 20 6 5 5
[38] 62 10 10 9
[39] 129 4 4 4
[44] 1 12 12 6
[47] 20 5 5 5
[48] 7 12 2 2
[54] 2 22 8 5
[56] 7 4 4 4
[57] 7 4 4 4
[58] 2 8 3 3
Unpublished _ 56 56 42
own data
R _ 293 225 187

4 Results and Discussion

The analysis corresponding to the separation in six groups has the highest silhouette
value (K = 6; average silhouette = 0.157) of all the different K-means analyses
carried out, and is the grouping that has been considered. Table 2 shows the
synthetic relevés corresponding to this grouping.
Group 1: Represents the shrubland community in which A. macrostachyum
coexists with S. hispanica; it contains relevés from the Spanish provinces of
Albacete, Alicante, Murcia, Almería, and Granada.
Arthrocnemum macrostachyum plant communities ... 235

Table 2 Synthetic table with the six groups generated by K-means analysis. Numbers represent
the presence percentage in all relevés in each group. Grey-shaded values indicate faithful taxa
Taxon name/K-means groups G1 G2 G3 G4 G5 G6
Nº of relevés 26 50 25 15 17 53
Nº of species 33 45 25 18 16 55
Arthrocnemum macrostachyum 100 100 100 100 100 100

Faithful species G1
Sarcocornia hispanica 96 3 8 . . .
Limonium supinum 30 . . . . .
Lygeum spartum 46 . 8 6 5 5
Suaeda vera 88 33 4 . 5 58
Puccinellia fasciculata 11 . . . . .
Puccinellia caespitosa 11 . . . . .
Limonium majus 11 . . . . .
Phragmites australis 23 3 . 6 . 1
Limonium cossonianum 42 1 8 . 70 3

Faithful species G2
Halimione portulacoides 7 100 40 . 5 3
Limonium algarvense . 19 . . . .
Sarcocornia pruinosa . 23 . . . 3
Sarcocornia alpini . 27 . . 5 9
Myriolimon ferulaceum 3 21 . . . 3
Limoniastrum monopetalum . 27 . . 11 9

Faithful species G3
Sarcocornia lagascae . . 100 . . .
Limonium bellidifolium . . 16 . . 1

Faithful species G4
Suaeda braun-blanquetii . . . 93 . .
Spergularia diandra . . . 60 . .
Frankenia pulverulenta . . 12 80 . 1
Hornungia procumbens . 1 8 66 . 1
Sphenopus divaricatus 3 9 . 73 . .
Parapholis incurva 7 . . 53 . 3
Limonium echioides . 1 . 33 . .
Rostraria cristata . . . 13 . .
Aizoon hispanicum . . . 13 . .

Faithful species G5
Frankenia corymbosa . 7 4 . 76 9
Phragmites communis . . . . 29 .
Limbarda crithmoides 19 27 16 . 82 .
Pancratium maritimum . . . . 17 .

Faithful species G6
Tetraena fontanesii . . . . . 20

Other species
Parapholis filiformis . 5 24 . . 5
Juncus maritimus 15 13 . . . 3
Frankenia thymifolia 15 . . 13 . .
Plantago coronopus . . 12 6 . 9
Atriplex glauca 15 . 4 . . 7
236 E. Ramírez et al.

Limonium delicatulum 19 3 . . . 3
Artemisia gallica 3 9 . 6 . 1
Limonium angustebracteatum 3 5 . . 11 .
Polypogon maritimus 3 1 8 6 . .
Cistanche phelypaea . 15 . . . 3
Spartina densiflora . 13 . . . 5
Juncus subulatus 3 3 4 . . 7
Limonium latebracteatum 11 . 4 . . 1
Suaeda maritima . 13 . . . 3
Salsola soda . . . 13 . 3
Limonium caesium 3 5 . . . 7
Limonium gibertii . . 8 . . 7
Cynomorium coccineum . . 12 . . 1
Limonium retusum . . 12 . . 1
Suaeda spicata 3 . 8 . . 1
Frankenia laevis . 7 . . . 5
Mesembryanthemum nodiflorum . 5 . . . 7
Plantago crassifolia 7 . . . 5 .
Frankenia capitata . . . . . 11
Limonium lanceolatum . 11 . . . .
Myriolimon diffusum . 11 . . . .

In addition to A. macrostachyum, the faithful species in this group are Limonium

cossonianum, L. majus, L. supinum, Lygeum spartum, Phragmites australis,
Puccinellia caespitosa, P. fasciculata and S. hispanica. Another species often
present in this group is the Murcian-Almerian endemic Limonium delicatulum.
The floristic composition and biogeographical distribution of the relevés inclu-
ded in this group lead us to consider its inclusion in the association Arthrocnemo
macrostachyi-Sarcocornietum hispanicae [24] (Albacete, Alicante, Murcia and
Almería) and in the endemic association of the Guadician-Bacensean sector
Limonio majoris-Sarcocornietum hispanicae (Granada) [24, 26, 30]. The biogeo-
graphical distribution of these communities covers inland saline areas throughout
the Central Iberian Mediterranean biogeographical province (Castilian subprovince,
southwest territories of the La Mancha sector), Murcia and Almería biogeographical
province (Murcia and Almería subprovince, Alicante and Murcia and Almería
sectors), and the Bética province (Bética subprovince, Hoyas de Guadix and Baza
sector).
These communities can withstand temporary flooding (winter, spring), summer
drought and a high concentration of salts in the soil due mainly to sulphates,
carbonates and bicarbonates.
The relevés dominated by A. macrostachyum, with the presence of S. hispanica [S.
fruticosa sensu auct. pl.] in these territories in southeast Iberia have traditionally been
ascribed to the association Frankenio corymbosae-Arthrocnemetum macrostachyi [1,
3, 35, 58]. Salazar et al. [56] include the relevés controlled by A. macrostachyum in
the inland salt pans of Granada (Guadix, Baza) in an association structured by what
they interpret as Sarcocornia fruticosa. The relevés we use to describe the phytoso-
ciological association Arthrocnemo macrostachyi-Sarcocornietum hispanicae for the
eastern Iberian Peninsula also reflect this last syntaxonomical interpretation [24, 25].
Arthrocnemum macrostachyum plant communities ... 237

Species with low presence in G1: Puccinellia stenophylla 3, Elytrigia curvifolia

3, Helianthemum polygonoides 7, Lycium intricatum 3, Aeluropus littoralis 7,
Sonchus tenerrimus 3; in G2: Limonium virgatum 1, Limonium daveaui 1, Elytrigia
elongata 3, Tripolium pannonicum 3, Puccinellia stenophylla 1, Atriplex halimus 3,
Polygonum equisetiforme 5, Triglochin barrelieri 7, Asparagus horridus 3,
Spergularia media 9, Sporolobus pungens 3, Sonchus tenerrimus 7, Hordeum
marinum 1; in G3: Lycium intricatum 4, Salicornia patula 8, Limonium delicatulum
subsp. formenterae 8; in G4: Hordeum marinum 6; in G5: Halocnemum cruciatum
5, Asparagus horridus 5, Sporolobus pungens 5; and in G6: Limonium virgatum 1,
Limonium daveaui 1, Trisetaria panicea 3, Anisantha madritensis 3, Lactuca
tenerrima 3, Spergularia heldreichii 3, Elytrigia juncea 3, Traganum moquinii 3,
Elytrigia curvifolia 1, Spergularia marina 5, Juncus bufonius 5, Suaeda vermicu-
lata 5, Atriplex halimus 3, Polygonum equisetiforme 2, Limonium tuberculatum 7,
Aeluropus litoralis 1, Sporolobus pungens 1, Hordeum marinum 3.
Group 2: this group includes relevés from the west coast of the Iberian Peninsula
(Portugal, Spain) in the biogeographical territories of the Coastal Lusitania-West
Andalusia province (Divisorio Portuguese, and Cádiz and Sado subprovinces;
Divisorio Portuguese, Ribatejo and Sado, Algarve and Monchique, Cádiz and
Huelva coastline, and Algeciras and Aljibe sectors). It contains relevés with 46 taxa;
the faithful species are A. macrostachyum, Halimione portulacoides, Limoniastrum
monopetalum, Limonium algarvense, Myriolimon ferulaceum, S. alpini and S.
pruinosa. Other species with a relevant presence are Cistanche phelypaea,
Limbarda crithmoides, Limonium lanceolatum, Myriolimon diffusum, Spartina
densiflora and Suaeda maritima.
This set of relevés corresponds closely to the association Inulo crithmoidis-
Arthrocnemetum macrostachyi. The territory where these halophilous communities
develop has an oceanic Mediterranean pluvi-seasonal bioclimate characterised by
high precipitation, absence of frost in coastal areas, and high oceanicity. These
thrive in the higher parts of Atlantic coastal marshes and in the brinier and more
emerged zones of the wetlands, occupying the driest parts of salt marshes [10, 11,
34, 37, 57].
A few relevés from the east of the peninsula have joined the western block.
These share the presence of coastal species with wide distributions (Limoniastrum
monopetalum and Halimione portulacoides), which, due to the low biomass of the
other species, may have influenced the result. It should also be noted that Halimione
portulacoides may predominate exceptionally over A. macrostachyum in these
communities in the southwestern Iberian Peninsula [46]. Similarly, a relevé from
Lanzarote (Canary Islands) with a high cover of Halimione portulacoides appears
in this group.
Group 3: comprises 25 relevés from the eastern Iberian Peninsula and the Balearic
Islands that can be ascribed clearly to the association Sphenopo
divaricati-Arthrocnemetum glauci. The faithful species are Limonium bellidifolium
and S. lagascae. This last species has the highest presence percentage and clearly
238 E. Ramírez et al.

defines the community dominated by the succulent halophyte A. macrostachyum.

Other species with significant presence are Cynomorium coccineum, Halimione
portulacoides, Limonium retusum and Parapholis filiformis. The community is
characteristic of eastern coastal areas of the Iberian Peninsula and the Balearic
archipelago, and is included in the Valencia-Provençe and Balearic biogeographical
province. The association occupies low-lying zones with high salt concentrations
[8, 12, 32, 39, 47, 48].
Group 4: is formed by relevés from the Spanish provinces of Toledo, Madrid and
Zaragoza, and groups in a single block the two associations described for the central
Iberian Peninsula and the Ebro Valley. The most important faithful species are
Aizoon hispanicum, Frankenia pulverulenta, F. thymifolia, Hornungia procum-
bens, Limonium echioides, Parapholis incurva, Rostraria cristata, Salsola soda,
Spergularia diandra, Sphenopus divaricatus and Suaeda braun-blanquetii. The
result of the numerical analysis supports the observations made in the field on these
ecosystems: A. macrostachyum does not coexist with S. carinata in the salt flats of
Toledo and Madrid, and there are no supported references from the province of
Zaragoza; although Puccinellia fasciculata is not present in the community dom-
inated by A. macrostachyum, Suaeda braun-blanquetii contributes biomass and
singularity to the association when considered as a whole.
When Castroviejo and Cirujano [9] describe the association that they propose for
the center of the peninsula (Puccinellio fasciculatae-Arthrocnemetum macro-
stachyi), they explain that it is very close to the association described for the Ebro
Valley. They argue, however, that the differential species S. carinata (sub
Sarcocornia perennis ssp. alpini) and Puccinellia fasciculata allow one syntaxon to
be separated from the other. The table complied by the authors includes 16 relevés
from the same locality (Aranjuez), in seven of which they indicate the joint pres-
ence of A. macrostachyum and S. carinata. In our field campaigns in these and
neighbouring salt marshes we never observed these species growing together. For
this reason it was decided to exclude these relevés from the numerical analysis.
Taking into account the floristic similarities between these two communities, we
propose their assimilation in the alliance Suaedion braun-blanquetii under the name
Suaedo braun-blanquetii-Arthrocnemetum macrostachyi, which takes precedence
due to its age [60].
The halophilous community as a whole belongs to the Central Iberian
Mediterranean biogeographical province distributed throughout the territories of
Lower Aragón and Upper Ebro, and the Castilian subprovinces. It thrives in
endorrheic lakes with high salt concentrations in which the soil is waterlogged in
winter and spring, and becomes dry and intensely saline in summer [6, 9].
Arthrocnemum macrostachyum plant communities ... 239

Fig. 1 Geographical location of the relevés grouped by the different plant communities
recognized

Group 5: is defined by the majority presence of Frankenia corymbosa, Limbarda

crithmoides, Limonium cossonianum, Pancratium maritimum and Phragmites
communis, with relevés from the Spanish provinces of Alicante, Almería
and Murcia (Murcia and Almería biogeographical province). This block groups
the most clearly defined relevés in the association Frankenio corymbosae-
Arthrocnemetum macrostachyi. This is a strictly coastal community characteristic
of the Murcian-Almerian territory and grows on soils with a compact
texture, occasionally flooded by saltwater and with frequent saline efflorescences
[1, 2, 38, 44].
Group 6: combines a total of 53 relevés from different points throughout Spanish
and Portuguese territories which have been grouped without a completely clear
pattern of blocks. The presence percentage of the different taxa in this group is
lower than the other groups. Most of the relevés screened from the Canary Islands,
however, have been included in this block, in which Tetraena fontanesii is the
faithful species. The Canary association Zygophyllo fontanesii-Arthrocnemetum
macrostachyi is differentiated clearly, floristically and biogeographically, from the
240 E. Ramírez et al.

communities on the Iberian Peninsula and Balearic Islands, and must therefore be
maintained in its entirety. Species such as Frankenia capitata, Limonium tuber-
culatum and Traganum moquinii are only found in these islands within the
framework studied.
A common feature of the relevés in the group is a high presence of Suaeda vera,
a halo-nitrophilous taxon that occurs in wetlands and salt marshes over almost all of
the Iberian Peninsula and islands. Some of these relevés may represent alterations in
community structure. On Lanzarote and Fuerteventura the abundance of this species
is interpreted as the result of a dynamic caused by contamination and/or erosion
[18, 54].
The association Zygophyllo fontanesii-Arthrocnemetum macrostachyi has a
coastal distribution throughout the Canary biogeographical province (East Canary
subprovince, Lanzarote and Fuerteventura islands). The type community occupies
upper levels of salt marshes and is subjected to long periods of drought when the
tides fluctuate only slightly. Saline efflorescences emerge in winter in the dried
substrate and lead to the development of communities of Frankenia capitata
and Mesembryanthemum nodiflorum.
The geographical distributions of the associations treated in this study are shown
on Fig. 1. On the other hand, Fig. 2 shows details of A. macrostachyum and the
different communities recognized in this study.

5 Syntaxonomy

SALICORNIETEA FRUTICOSAE Br.-Bl. & Tüxen ex A. & O. Bolòs 1950.

* Arthrocnemo macrostachyi-Suaedetalia braun-blanquetii Rufo, Fuente &
Sánchez-Mata 2016 [in Phytocoenol 46(4):391]
• Arthrocnemion glauci (Rivas-Martínez in Rivas-Martínez & al. 1980)
Rivas-Martínez & Costa 1984 [in Doc. Phytosociol 8:18]
1. Inulo crithmoidis-Arthrocnemetum macrostachyi Fontes ex Géhu &
Géhu-Franck 1977 [in Acta Bot. Malacitana 3:149]
2. Frankenio corymbosae-Arthrocnemetum macrostachyi Rivas-Martínez, Alcaraz,
Belmonte, Cantó & Sánchez-Mata 1984 [in Doc. Phytosociol 8:336]
3. Sphenopo divaricati-Arthrocnemetum glauci Br.-Bl. 1933 [Prodr Group Veg
1:21]
4. Arthrocnemo macrostachyi-Sarcocornietum hispanicae Fuente, Rufo, Teijeiro
& Sánchez-Mata 2013 [in Lazaroa 34:270]
5. Limonio majoris-Sarcocornietum hispanicae Lendínez, Marchal & Salazar
2012 corr. Rufo, Fuente & Sánchez-Mata 2016 [in Phytocoenol 46(4):392]
6. Zygophyllo fontanesii-Arthrocnemetum macrostachyi Fernández Galván &
Santos 1984 [in Lazaroa 5:146]
Arthrocnemum macrostachyum plant communities ... 241

Fig. 2 A. Detail of Arthrocnemum macrostachyum in flowering state. B. Succulent stems in the

fruiting stage. C. Wetlands dominated by A. macrostachyum with a high presence of Halimione
portulacoides (Portugal, Tavira). D. Appearance of the association Inulo crithmoidis-
Arthrocnemetum macrostachyi (Spain, Huelva, La Rábida). E. A. macrostachyum in an inland
salt marsh (Suaedo braun-blanquetii-Arthrocnemetum macrostachyi); in the background Suaeda
braun-blanquetii (Spain, Madrid, Colmenar de Oreja). F. Appearance of the association
Arthrocnemo macrostachyi-Sarcocornietum hispanicae; in the foreground with more modest
height is A. macrostachyum; behind and with higher branches is S. hispanica (Spain, Albacete,
Cordovilla)
242 E. Ramírez et al.

• Suaedion braun-blanquetii Br.- Bl. & O. Bolòs 1958 corr. Rivas-Martínez,

Báscones, T.E. Díaz, Fernández-González & Loidi 1991 [in Itinera Geobot
5:407]
1. Suaedo braun-blanquetii-Arthrocnemetum macrostachyi (Br.-Bl. & O. Bolòs
1958) Rivas-Martínez, Báscones, T.E. Díaz, Férnandez-Gónzalez & Loidi 1991
[in Itinera Geobot 5: 407, incl. Puccinellio fasciculatae- Arthrocnemetum
macrostachyi Castroviejo & Cirujano 1980 in Anales Jard Bot Madrid 37
(1):144]

6 Floristic Appendix and New Herbarium Records

• Sarcocornia alpini (Lag.) Rivas Mart.

• Sarcocornia carinata (Fuente, Rufo & Sánchez-Mata) Fuente, Rufo &
Sánchez-Mata
• Sarcocornia lagascae Fuente, Rufo & Sánchez-Mata
• Sarcocornia pruinosa Fuente, Rufo & Sánchez-Mata
• A new nomenclatural combination is obliged if Sarcocornia is taxonomically
treated as synonym of Salicornia sensu Piirainen et al. [41]: Salicornia carinata
(Fuente, Rufo & Sánchez-Mata) comb. nov. Bas.: Sarcocornia alpini
subsp. carinata Fuente, Rufo & Sánchez Mata in Pl. Biosyst 147(1): 170. 2013.
• Suaeda braun-blanquetii (Pedrol & Castrov.) Rivas Mart., Cantó & Sánchez
Mata
• New herbarium records for Sarcocornia hispanica (SPAIN):
ESP. Albacete: Hellín, Cordovilla, Saladares de Cordovilla, entre Cordovilla y
Fuente García, 06.02.2019, comunidades fruticosas halófilas (Arthrocnemo-
Sarcocornietum hispanicae). D. Sánchez-Mata, V. de la Fuente & al. (MAF
177418).
ESP. Albacete: Hellín, Agramón, Saladares de Agramón, 400 m, 06.02.2019,
comunidades fruticosas halófilas (Arthrocnemo-Sarcocornietum hispanicae).
D. Sánchez-Mata, V. de la Fuente & al. (MAF 177419, 177420).
ESP. Almería: Roquetas de Mar. Paraje Natural Punta Entinas-Sabinar, cerca
del Charcón del Flamenco, 10.08.2017, comunidades halófilas fruticosas
(Arthrocnemo-Sarcocornietum hispanicae). D. Sánchez-Mata & I. Sánchez-
Gavilán (MAF 177277, 177278).
ESP. Almería: Vera, Saladar de los Canos, entre Garrucha y Villaricos, 24–
08.2016, comunidades halófilas fruticosas (Arthrocnemo-Sarcocornietum his-
panicae). D. Sánchez-Mata & I. Sánchez-Gavilán (MAF 177279).
Arthrocnemum macrostachyum plant communities ... 243

References

1. Alcaraz FJ (1984) Flora y vegetación del N.E. de Murcia. Secretariado de Publicaciones de la

Universidad de Murcia, Murcia, p 406
2. Alcaraz FJ, Díaz TE, Rivas-Martínez S, Sánchez-Gómez P (1989) Datos sobre la vegetación
del sureste de España: provincia biogeográfica Murciano-Almeriense. Itinera Geobot 2:5–133
3. Alonso MA (1996) Flora y vegetación del valle de Villena (Alicante). Publicaciones del
Instituto Alicantino de Cultura Juan Gil-Albert, Alicante, p 263
4. Álvarez J (1997) Relaciones suelo-vegetación en saladares del Sureste de España. Ph
Dissertation. Universidad de Murcia, Murcia, p 456
5. Biondi E, Casavecchia S, Estrelles E, Soriano P (2013) Halocnemum M. Bieb. vegetation in
the Mediterranean Basin. Pl Biosyst 147:536–547
6. Braun-Blanquet J, de Bolòs O (1958) Les groupements végétaux du bassin moyen de l’Ebre
et leur dynamisme. Anales Estac Exp. Aula Dei 5:1–266
7. Braun-Blanquet J (1979) Fitosociología: Bases para el estudio de las comunidades vegetales.
Blume Ediciones, Madrid, p 820
8. Cantó P, Laorga S, Belmonte D (1986) Vegetación y catálogo florístico del Peñón de Ifach
(Penyal d´Ifac) (Alicante, España). Opusc Bot Pharm Complutensis 3:3–86
9. Castroviejo S, Cirujano S (1980) Sarcocornietea en la Mancha (España). Anales Jard Bot
Madrid 37(1):143–154
10. Costa JC, Lousã M (1989) Communautes psammophiles et halophiles du “Ria du Alvor”.
Coll Phytosociol XVIII:120–135
11. Costa JC, Lousã M, Espírito-Santo MD (1996) A vegetaçao do parque natural da Ria Formosa
(Algarve, Portugal). Stud Bot Univ Salamanca 15:69–157
12. Curcó A (1996) La vegetación del Delta de l´Ebro (II): les comunitats halòfiles i
halo-nitròfiles (Classes Puccinellio-Salicornietea i Cakiletea maritimae). Fol Bot Misc
10:113–139
13. Del Arco Aguilar MJ, Izco J, Rodríguez Delgado O (2017) The name Salicornietum perennis
(sensu lato) and global nomenclatural issues. Lazaroa 38(2):203–206
14. Díaz-González TE (2004) Pasado, presente y futuro de la Fitosociología española. Lazaroa
25:3–13
15. DOCE (1992) Directiva 92/43/CEE del Consejo, de Conservación de Hábitats Naturales y de
la Fauna y Flora Silvestres
16. Euro+Med (2006) Euro+Med PlantBase. The information resource for Euro-Mediterranean
plant diversity. https://ww2.bgbm.org/EuroPlusMed/. Accessed Apr 2021
17. European Commission (2013) Interpretation Manual of European Union Habitats – EUR 28.
DG Environment, Nature ENV B.3, Strasbourg
18. Fernández M, Santos A (1984) La vegetación del litoral de Canarias. I. Arthrocnemetea.
Lazaroa 5:143–155
19. Flowers TJ, Galal HK, Bromham L (2010) Evolution of halophytes: multiple origins of salt
tolerance in land plants. Funct Pl Biol 37:604–612
20. Font X (2005) La gestión de la biodiversidad mediante bases de datos en línea y el
programario b-vegana. Rec Rurais, Serie Cursos 2:65–72
21. Font X, Rodríguez-Rojo MP, Acedo C, Biurrun I, Fernández-González F, Lence C, Loidi J,
Ninot JM (2010) SIVIM: an online database of Iberian and Macaronesian vegetation. Waldök
Landschafts Natur 9:15–22
22. Font X, Pérez-García N, Biurrun I, Fernández-González F, Lence C (2012) The Iberian and
Macaronesian vegetation information system (SIVIM, www.sivim.info), five years of online
vegetation’s data publishing. Plant Sociol 49:89–95
23. de la Fuente V, Rufo L, Sánchez-Mata D (2011) Sarcocornia hispanica (Chenopodiaceae), a
new species from the Iberian Peninsula. Lazaroa 32:9–13
244 E. Ramírez et al.

24. de la Fuente V, Rufo L, Teijeiro A, Sánchez-Mata D (2013) Arthrocnemo

macrostachyi-Sarcocornietum hispanicae, a new halophytic plant community from eastern
Iberian Península. Lazaroa 34:267–272
25. de la Fuente V, Oggerin M, Rufo L, Ortúñez E, Sánchez-Mata D, Amils R (2013) A
micromorphological and phylogenetic study of Sarcocornia A.J. Scott (Chenopodiaceae) on
the Iberian Peninsula. Plant Biosyst 147:158–173
26. de la Fuente V, Rufo L, Rodríguez N, Sánchez-Mata D, Amils R (2016) A study of
Sarcocornia A.J. Scott (Chenopodiaceae) from Western Mediterranean Europe. Plant Biosyst
150(2):343–356
27. Géhu J-M, Géhu-Franck J (1977) Quelques données dur les Arthrocnemetea fruticosi
ibériques sud-occidentaux. Acta Bot Malacitana 3:145–157
28. Giménez E, Navarro J, Oña JA, Gómez F (2003) Paraje Natural Punta Entinas-Sabinar
(Almería): Flora, vegetación y ornitofauna. Servicio de Publicaciones, Universidad de
Almería, Almería, p 171
29. Kadereit G, Mucina L, Freitag H (2006) Phylogeny of Salicornioideae (Chenopodiaceae):
diversification, biogeography and evolutionary trends in leaf and flower morphology. Taxon
55:617–642
30. Lendínez ML, Marchal FM, Salazar C (2012) Una nueva asociación de vegetación halófila en
el sureste de la Península Ibérica (España): Limonio majoris-Sarcocornietum fruticosae.
Lagascalia 32:229–236
31. Lengyel A, Chytrý M, Tichý L (2011) Heterogeneity constrained random resampling of
phytosociological databases. J Veg Sci 22:175–183
32. Llorens L (1986) La vegetación de los saladares de la isla de Formentera (Baleares). Anales
Jard Bot Madrid 42(2):469–479
33. Loidi J (2008) La fitosociología como proveedora de herramientas de gestión. Lazaroa 29:7–
17
34. Luis-Dias TM (2009) Flora e vegetação dos salgados de Coina, Corrios e Alcochete. Ms
Dissertation. Universidade Técnica de Lisboa, p 79
35. Molina JA, Pertíñez C, Cruz MT (2001) Dos sobre la relación suelo-vegetación en los
saladares de Cordovilla (Albacete, España). Sabuco 1:217–232
36. Mucina L., Bültmann H, Dierssen K, Theurillat J-P, Raus T, Čarni A, Šumberová K,
Willner W, Dengler J, Gavilán García R, Chytrý M, Hájek M, Di Pietro R, Iakushenko D,
Pallas J, Daniëls FJA, Bergmeier E, Santos Guerra A, Ermakov N, Valachovič M, Schaminée
JHJ, Lysenko T, Didukh YP, Pignatii S, Rodwell JS, Capelo J, Weber HE, Solomeshch A,
Dimopoulos P, Aguiar C, Hennekens SM, Tichý L (2016) Vegetation of Europe: hierarchical
floristic classification system of vascular plant, bryophyte, lichen and algal communities. App
Veg Sci (supp 1): 3–264
37. Neto C, Arsénio P, Costa JC (2009) Flora e Vegetaçã do sudoeste de Portugal continental.
Quercetea 9:43–144
38. Peinado M, Alcaraz F, Martínez-Parras JM (1992) Vegetation of Southeastern Spain. Flora et
Vegetatio Mundi, Stuttgart, p 487
39. Pérez Badia MR (1997) Flora vascular y vegetación de la comarca de la Marina Alta
(Alicante). Colección Técnica 16, Instituto de Cultura Juan Gil Albert, Alicante, p 566
40. Pigott CD, Ratcliffe DA, Malloch AJC, Birks HJB, Proctor MCF, Shimwell DW, Huntley JP,
Radford E, Wigginton MJ, Wilkins P (2000) Annual Salicornia salt-marsh community:
Salicornietum europaeae Warming 1906. In: JS Rodwell (ed) british plant communities, vol 5,
pp 45–46. Cambridge, p 526
41. Piirainen M, Liebisch O, Kadereit G (2017) Phylogeny, biogeography, systematics and
taxonomy of Salicornioideae (Amaranthaceae/Chenopodiaceae) – a cosmopolitan, highly
specialized hygrohalophyte lineage dating back to the Oligocene. Taxon 66(1):109–132
42. Ramírez E, Rufo L, Sánchez-Mata D, de la Fuente V (2019) Arthrocaulon meridionalis, a new
species of mediterranean flora. Med Bot 40(1):33–41
Arthrocnemum macrostachyum plant communities ... 245

43. Redondo-Gómez S, Mateos-Naranjo E, Figueroa ME, Davy AJ (2010) Salt stimulation of

growth and photosynthesis in an extreme halophyte, Arthrocnemum macrostachyum. Plant
Biol 12:79–87
44. Rivas-Martínez S (1984) Vegetatio hispaniae Notula VI. Stud Bot Univ Salamanca 3:7–16
45. Rivas-Martínez S, Alcaraz F, Belmonte D, Cantó P, Sánchez-Mata D (1984) Contribución al
conocimiento de la vegetación de los saladares del SE de la Península Ibérica (Arthrocnemion
glauci). Doc Phytosociol 8:335–342
46. Rivas-Martínez S, Lousã M, Díaz TE, Fernández-González F, Costa JC (1990) La vegetación
del sur de Portugal (Sado, Alentejo y Algarve). Itinera Geobot 3:5–126
47. Rivas-Martínez S, Costa M, Loidi J (1992) La vegetación de las islas de Ibiza y Formentera
(Islas Baleares, España). Itinera Geobot 6:99–236
48. Rivas-Martínez S, Costa M, Soriano P, Pérez Badia MR, Llorens L, Roselló JA (1992) Datos
sobre el paisaje vegetal de Mallorca e Ibiza (Islas Baleares, España). Itinera Geobot 6:5–98
49. Rivas-Martínez S, Fernández-González F, Loidi J, Fernandes M, Penas A (2001)
Sintaxonomical checklist of vascular plant communities of Spain and Portugal to association
level. Itinera Geobot 14:5–41
50. Rivas-Martínez S, Díaz-González TE, Fernández-González F, Izco LJ, Fernandes M, Penas A
(2002) Vascular plant communities of Spain and Portugal: addenda to the syntaxonomical
checklist of 2001. Itinera Geobot 15(1):5–432
51. Rivas-Martínez S et al (2007) Mapa de series, geoseries y geopermaseries de vegetación de
España. Parte I. Itinera Geobot 17:5–436
52. Rivas-Martínez et al (2011) Mapa de series, geoseries y geopermaseries de vegetación de
España: [Memoria del mapa de vegetación potencial de España]. Itinera Geobot 18(2):425–
800
53. Rivas-Martínez S, Penas A, Díaz-González TE, Cantó P, Del Río S, Costa JC, Herrero L,
Molero J (2017) Biogeographic units of the Iberian Peninsula and Balearic Islands to district
level: a concise sinopsis. In: Loidi J (ed). The vegetation of Iberian Peninsula, plant and
vegetation, vol 12, pp 131–188. Springer, Cham
54. Rodríguez O, García A, Reyes AJ (2000) Estudio fitosociológico de la vegetación actual de
Fuerteventura (Islas Canarias). Vieraea 28:61–98
55. Rufo L, de la Fuente V, Sánchez-Mata D (2016) Sarcocornia plant communities of the Iberian
Peninsula and the Balearic Islands. Phytocoenol 46(4):388–396
56. Salazar C, Torres JA, Marchal MF, Cano E (2002) La vegetación edafohigrófila del distrito
Guadiciano-Bastetano (Granada-Jaén, España). Lazaroa 23:45–64
57. Tamajón R, Muñoz J (2001) La vegetación de las marismas y lagunas de la hoja cartográfica
de Lebrija (Suroeste de España). Studia Bot Univ Salamanca 20:93–114
58. Valdés A, González JL, Molina R (1993) Flora y vegetación de los saladares de Cordovilla y
Agramon (SE de Albacete). Publicaciones del Instituto de Estudios Albacetenses, Albacete,
p 158
59. Tichý L (2002) JUICE, software for vegetation classification. J Veg Sci 13:451–453
60. Weber HE, Moravec J, Theurillat J-P (2000) International code of phytosociological
nomenclature. J Veg Sci 11:739–768
61. Wiser SK, De Cáceres M (2013) Updating vegetation classifications: an example with New
Zealand’s woody vegetation. J Veg Sci 24:80–93
Geosynphytosociological Typology
of French Atlantic Coastal Rocky-Cliff
Vegetation

Charlotte Demartini and Frédéric Bioret

Abstract Coastal cliff vegetation is spatially organized into parallel belts according
to ecological gradients caused by salt spray, wind exposure, relief energy, substrate
nature and depth. Along the French Atlantic coasts, a recent research programme set
up in 2011 is dedicated to a geosynphytosociological study of coastal vegetation.
The present work focuses on permanent vegetation of the French Atlantic coastal
rocky cliffs, corresponding to permaseries. Permaseries occuring within homoge-
neous geomorphological units are listed within geosynrelevés. The analysis of a set
of geosynrelevés from northern France up to the Pyrenees-Atlantic allows us to
characterise phytocoenotical diversity and to propose a preliminary typology of
coastal cliff vegetation. The geopermaseries assemblies follow Dupias’s and Rey’s
phytoecological regions. These biogeographical units are more precisely charac-
terized by substrate and climate. This analysis completes Géhu’s works about
sigma-association description of French Atlantic cliffs. Eleven geopermaseries of
rocky coast are defined: Sandstone or marly-cliff geopermaseries of Boulonnais;
Marly and chalky cliff geopermaseries of Boulonnais; Chalky-cliff geopermaseries
of Pays de Caux; Crystalline-cliff geopermaseries of the Gulf Normand-Breton;
Crystalline-cliff geopermaseries of the western Massif Armoricain, under a hyper
Atlantic climate; Crystalline-cliff geopermaseries of the southern Massif
Armoricain, under a thermo-Atlantic climate; Crystalline-cliff geopermaseries of
islands of the southern Massif Armoricain, under a thermo-Atlantic climate;
Cliff-top fossil dune geopermaseries on sandy decalcified substrates; Calcareous cliff
geopermaseries of the right-bank of the Gironde estuary, under a thermo-Atlantic
climate; Chalky and sandstone cliff geopermaseries of the south-Aquitan Basin; and
marly-chalky or sandstone cliff geopermaseries of the Corniche Basque.

Keywords Rocky cliff vegetation
French Atlantic coast
Geopermaseries

Geosynphytosociology

C. Demartini
F. Bioret (&)

EA 7462 Géoarchitecture, University of Western Brittany, Brest, France
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 247
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_13
248 C. Demartini and F. Bioret

1 Introduction

Financed by the French Ecology, Energy and Sustainable Development Ministry,

the national habitats and vegetation mapping program (CarHAB) was launched in
2011. Based on the dynamico-catenal phytosociological approach, its purpose is to
produce a map for French vegetation series and geoseries up to 2025, to answer the
needs of the European Flora-Fauna Habitat Directive of 1992 [42]. Analysis of
vegetation along the Channel-Atlantic coast requires a specific approach, due
mainly to vegetation zonation induced by specific ecological constraints. This work
started in 2012 and aims to create a typology and mapping methodology focused on
geopermaseries and geominoriseries of Channel-Atlantic coastal vegetation [14].
This paper focuses exclusively on permanent vegetation developing along rocky
coasts, on the supra-coastal zone. Several geopermaseries are identified by ana-
lysing géosynrelevés, each of them distinguished by regional climatic and geo-
morphological conditions.

2 Study Area

Our study focuses on the phytocoenotic diversity of the cliffs of the French
Channel-Atlantic coastline, from the Belgian to the Spanish borders. Most of the
sites inventoried are located between Blanc-Nez cape and the Corniche Basque
(Fig. 1).

3 Concepts and Method

3.1 Concepts: Series and Geoseries

Analysis of coastal vegetation landscapes is based on the synphytosociological and

geosynphytosociological methods. Synphytosociology, also known as serial phy-
tosociology, has vegetation series as its fundamental unit. It studies spatially sets of
plant communities that are linked dynamically, including substitute communities up
to potential vegetation [24, 37, 46], on ecologically homogeneous areas called
tesselas.
Geosynphytosociology (or catenal phytosociology) has geoseries as its funda-
mental unit. It studies homogeneous geomorphological units (e.g. crest, coastal
dune, rocky coast…) composed of vegetation series belonging to sets of adjacent
tesselas called a catena (Fig. 2).
In the coastal zone, ecological constraints (salt and wind exposure, dryness,
oligotrophic and superficial substrates…) block vegetation succession so that plant
Geosynphytosociological Typology of French Atlantic Coastal … 249

Fig. 1 Location of the study sites

communities form permanent vegetation types, called permaseries, that do not

correspond to the climatic potential.
For Rivas-Martínez: “permaseries or permasigmeta are perennial, stable com-
munities that populate microtesselas or microtessela complexes that are very similar
to each other, in particular areas such as: polar regions, crests of high mountains,
coastal zones, mobile dunes, cliffs, coastal reefs washed by seawater, etc. The
steady mature stage, or climax, corresponds to a perennial vascular community that
is generally poorly stratified (…)” [47, 48].
The new concept of mesoseries [44] has been replaced by the term ‘curtaseries’
or truncated series [43]. It concerns cases intermediate between permaseries and
series, where the vegetation is under less extreme ecological constraints, which
permit the first steps of succession but are strong enough to prevent reaching the
stage of natural potential vegetation, which is forest.
250 C. Demartini and F. Bioret

Fig. 2 Theoretical representation of a vegetation geoseries inside a catena composed of three

homogeneous tesselar units (series 1, 2 and 3) (after [4])

For grammatical reasons, Rivas-Martínez [49] proposes to name this concept

“minoriseries” instead of “curtaseries”. For this author, these terms are
synonymous.
Coastal tesselas grouping minoriseries are usually inserted, continuously or in a
dotted line [45], between the microtesselas of permaseries and the tesselas of
adjacent series (Fig. 3).

3.2 Method: Characterization of Coastal Geopermaseries

Associations of rocky coasts are identified in various phytosociological works

[3, 5–8, 16, 17, 19, 25–30, 32–36, 38–41]. The synsystem was developed from
different works: Géhu [27], Bardat et al. [1], Bioret and Géhu [6]; De Foucault and
Bioret [12] and De Foucault and Catteau [13].
In the field, the method consists on making systematic geopermasynrelevés. The
aim is to establish the list of permaseries (permanent plant communities) present in
a homogeneous geomorphological and biogeographical area. (In our study this
corresponds to a structurally homogeneous rocky coast) [24]. Coefficients of spatial
shape and abundance-dominance are added to each permaseries. The
abundance-dominance scale follows Braun-Blanquet [9–11]. The spatial shapes of
syntaxa in the landscape follow Béguin et al. [2], Géhu [18], Géhu &
Geosynphytosociological Typology of French Atlantic Coastal … 251

Fig. 3 Theoretical zonation for permaseries and minoriseries of Finistere rocky coasts

Rivas-Martínez [37], and Tüxen [50]: O spatial; o limited spatial;/linear; 0 spatial

linear, wide fringe; punctual; … mosaic.
Geopermasynrelevés have been made throughout a large part of the French
coastal territory, from the cliffs of the Gris-Nez cape in the Pas-de-Calais to the
Corniche Basque in the Atlantic Pyrenées. As an example, geopermasynrelevés
made on a part of the Finistère rocky coast are presented in Table 1.

4 Results

Analysis of geopermasynrelevés allows one to propose a zonation of French rocky-

coast vegetation. The geopermaseries assemblies follow the phyto-ecological regions
of Dupias and Rey [15]: chalky or marly cliffs of the Blanc-Nez and Gris-Nez capes
(Nord Pas-de-Calais) and of Etretat (Haute-Normandie) belong to the sedimentary
Parisian area,western Channel crystalline rocks of the southern part of Vendée are
linked to the Armorican massif; limestone or sand topped cliffs (Jard-sur-Mer,
Saint-Palais-sur-Mer), chalky and sandstone cliffs of Biarritz and marly-chalky or
sandstone cliffs of the Corniche Basque are typical of the Aquitaine area.
252 C. Demartini and F. Bioret

Table 1 Geopermasynrelevés realized on a part of the Finistère rocky coast (Brittany)

The distribution of coastal phytocoenoses, and by extension of coastal geoper-

maseries, depends on bioclimatic, geomorphologic and pedological factors. In 1984
Géhu and colleagues presented a map of French bioclimatic stages, proposing
creation of a thermo-Atlantic plain level and the subdivision of the coastal hill level
into two types: hyperAtlantic and Atlantic [31]. These great units have been divided
more precisely based on bioclimatic divisions, recent meteorological data (23
coastal stations studied from 1981 until 2010), and the chorology of plant associ-
ations. Eleven geopermaseries have been distinguished on the French
Channel-Atlantic coast (Table 2). This division completes the Géhu publications, in
which he described four sigmassociations on cliffs of Nord Pas-de-Calais,
Normandy and Brittany [20–22].
Geosynphytosociological Typology of French Atlantic Coastal … 253

Table 2 Summary table of géosynrelevés from French Atlantic cliffs

Cliffs geopermaseries
Cliffs geopermaseries of Massif Cliffs geopermaseries of
of sedimentary Parisian
armoricain Aquitain area
area
1 2 3 4 5 6 7 8 9 10 11
22 - 35 - 56 - 44 -
Department
62 62 76 50 22 - 29 85 56 85 17 64 64

Groix - Belle-île
Ploemeur to St-
Gilles-Croix-de-
Cap Blanc-Nez
Cap Gris-Nez /

Vaux-sur-Mer /
Cap d'Alprech

Pays de Caux

Perros-Guirec
Perros-Guirec

Jard-sur-Mer
Barfleur to

to Plomeur

St-Palais &

Corniche
(Etretat)

basque
Biarritz
Suzac
- Yeu
Vie
Location

Number of géosynrelevés 2 2 5 39 120 51 17 4 3 2 7

Number of syntaxa 12 4 5 37 48 44 37 14 12 4 3

Chasmo-halophytic vegetation
Spergulario rupicolae-Crithmetum maritimi (Roux & lahondère 1960) Géhu
2 . . V V V V 3 1 . .
1962 nom. inv.
Armerio maritimae-Asplenietum marini Géhu & Géhu-Franck 1984 . . . II III I I . . . .
Crithmo maritimi-Betetum maritimae Géhu 2008 . . . II I I + . 2 . .
Armerio maritimae-Inuletum crithmoidis Géhu 2006 . . . I I + I . . . .
Crithmo maritimi-Juncetum acuti Bioret 2008 . . . I r . + . . . .
Crithmo maritimi-Juncetum maritimi Géhu & Duquef 2009 . . . + r I I . . . .
Gpt à Carex extensa et Armeria maritima . . . + + I . . . . .
Umbilico rupestris-Asplenietum billoti de Foucault 1979 . . . II I r . . . . .
Crithmo maritimi-Limonietum normannici Lahondère & Bioret 1995 . . . r . . . . . . .
Umbilico rupestris-Asplenietum obovati Géhu & Biondi 1994 . . . . + . . . . . .
Armerio maritimae-Cochlearietum officinalis Géhu & Géhu-Franck 1984 . . . . IV . . . . . .
Soncho arvensis-Rumicetum rupestris Bioret & Géhu 2002 . . . . r . . . . . .
Apio graveolens-Rumicetum rupestris Bioret & Géhu 2002 . . . . . I . . . . .
Crithmo maritimi-Limonietum ovalifolii (Kuhnholtz-Lordat 1926) Lahondère,
. . . . . I + . 2 . .
Bioret & Botineau 1991
Crithmo maritimi-Limonietum occidentalis Pavillard 1928 . . . . . . . . . 2 .
Crithmo maritimi-Plantaginetum maritimae (Pavillard 1928) Guinea 1949 . . . . . . . . . . V
Gpt à Asplenium marinum . . I . . . . . . . .
Gpt à Asplenium marinum et Sonchus maritimus . . . . . . . . 1 . .
Vegetation of Riccardio pinguis-Eucladion verticillatii Bardat in Bardat et al.
1 2 . . . . . . . . .
2004 prov.
Halo-chomophytic vegetation
Spergulario rupicolae-Halimionetum portulacoidis Bioret & Géhu 2008 . . . + I IV IV 1 . . .
Spergulario rupicolae-Limonietum occidentalis Bioret & Géhu 2008 . . . I I II III . . . .
Armerio maritimae-Plantaginetum maritimae Géhu 2000 . . . + + I . . . . .
Spergulario rupicolae-Frankenietum laevis Bioret & Lahondère 2010 . . . . I III IV 1 . . .
Spergulario rupicolae-Limonietum dodartii Géhu, Franck & Scoppola 1984 . . . . r II I 1 . . .
Gpt à Phragmites australis (on the plane or at the foot of the cliff) . . . r + + I 1 . . .
Tuberario guttatae-Romuleetum columnae Provost 1978 . . . + . r + . . . .
Jasiono maritimae-Anthyllidetum maritimae Géhu & Duquef 2009 . . . + r . . . . . .
Anthemido nobilis-Agrostietum capillaris (Allorge 1922) de Foucault in
. . . . + . . . . . .
Royer et al. 2006
Ophioglosso lusitanici-Isoetetum histricis de Foucault 1988 . . . . r . . . . . .
Gpt à Juncus maritimus et Inula crithmoides . . . . . r I . . . .
Crithmo maritimi-Senecionetum cinerariae Géhu 2005 . . . . . + . . . . .
Junco maritimi-Caricetum extensae (Corillion 1953) Géhu 1976 . . . . . r . . . . .
Spergulario rupicolae-Puccinellietum maritimae Bioret & Gallet 2014 . . . . . + I . . . .
Brassicetum oleraceae Géhu 1962 . 2 III . . . . . . . .
Gpt à Brassica oleracea et Diplotaxis tenuifolia . 2 . . . . . . . . .
Samolo valerandi-Caricetum vikingensis Géhu 1982 2 . . . . . I 1 . . .
Agrostio stoloniferae-Caricetum vikingensis Géhu 1979 . . . . . . + . . . .
Soncho maritimi-Schoenetum nigricantis (Lahondère 1979) de Foucault
. . . . . . + . 1 . .
1984
Sagino maritimae-Spergularietum marinae Bioret 2006 . . . . . . . . 1 . .
Azonal vegetation along freshwater seeps
Scirpetum compacti (Van Langendonck 1931) Beeft. 1957 . . . . r + . . . . .
Pulicario dysentericae-Juncetum inflexi de Foucault in Royer et al. 2006 1 . . . . . . . . . .
Gpt à Epilobium hirsutum et Pulicaria dysenterica 2 . . . . . . . . . .
Aerohaline grass
Armerio maritimae-Festucetum pruinosae Géhu 2008 . . . V V V V . . . .
Carici arenarii-Festucetum pruinosae Bioret & Géhu 2008 . . . + I II + . . . .
Spergulario rupicolae-Silenetum maritimae Bioret & Géhu 2000 . . . I III + III . . . .
Armerio maritimae-Juncetum gerardii Bioret et al. 2014 . . . + r II II . . . .
Spergulario rupicolae-Armerietum maritimae Géhu 2008 . . . III IV I . . . . .
Festuco pruinosae-Elymetum atherici Géhu 2008 . . . I II I . . . . .
Gpt à Silene maritima et Festuca pruinosa . . . II I r . . . . .
Festuco pruinosi-Osmundetum regalis Géhu & Bioret 2000 . . . + + . . . . . .
Dactylo oceanicae-Daucetum gummiferi Géhu 2008 . . . II + . . . . . .
Sileno zetlandicae - Festucetum pruinosae Bioret & Géhu 2008 . . . r r . . . . . .
Armerio maritimae-Agrostietum maritimae Bioret & Géhu 2008 . . . . I II + . . . .
Festuco pruinosae-Ononidetum maritimae Bioret & Géhu 2008 . . . . + II III . . . .
Dauco gadeceaui-Festucetum pruinosae Bioret & Géhu 2008 . . . . . r . . . . .
Sileno zetlandicae-Festucetum pruinosae Bioret & Géhu 2008 . . . . + . . . . . .
Festuco huonii-Plantaginetum littoralis (J.-M & J. Géhu 1976) Bioret,
. . . . . . III . . . .
Bouzillé, de Foucault, Géhu & Godeau 1988
Dauco intermedii-Festucetum pruinosae (Liger 1956) Géhu & de Foucault
. . V . . . . . . . .
1978
Cirsio acaulis-Festucetum pruinosae Bioret & Géhu 2008 2 . . . . . . . . . .
Apio graveolentis-Tussilaginetum farfarae Géhu et al. 1983 ex Bioret &
2 . . . . . . . . . .
Géhu 2008
Tuberario guttatae-Corynephoretum canescentis Géhu 1964 . . . . . + + 4 . . .
Leucanthemo crassifolii-Festucetum pruinosae Géhu & Géhu-Franck 1980
. . . . . . . . . . V
corr. F. Pietro & Loidi 1984
Gpt à Carex distans et Festuca pruinosa . . . . . . . . . 2 .
Gpt à Festuca pruinosa et Spartina versicolor . . . . . . . . . 1 .
Crithmo maritimi-Elytrigietum acutae Bioret, Demartini, Géhu & Glemarec
. . . . . II . 3 1 . .
2016
254 C. Demartini and F. Bioret

Cliffs geopermaseries
Cliffs geopermaseries of Massif Cliffs geopermaseries of
of sedimentary Parisian
armoricain Aquitain area
area
1 2 3 4 5 6 7 8 9 10 11
22 - 35 - 56 - 44 -
Department
62 62 76 50 22 - 29 85 56 85 17 64 64

Groix - Belle-île
Ploemeur to St-
Gilles-Croix-de-
Cap Gris-Nez /

Cap Blanc-Nez

Vaux-sur-Mer /
Cap d'Alprech

Pays de Caux

Perros-Guirec

Perros-Guirec

Jard-sur-Mer

St-Palais &
to Plomeur
Barfleur to

Corniche
(Etretat)

basque
Biarritz
Suzac
- Yeu
Vie
Location

Number of géosynrelevés 2 2 5 39 120 51 17 4 3 2 7

Number of syntaxa 12 4 5 37 48 44 37 14 12 4 3

Flayed grass
Armerio maritimae-Plantaginetum coronopodis (Vanden Berghen 1979)
1 . . II III IV V . . . .
Géhu & Bioret 2008
Sagino maritimae-Catapodietum marini R. Tx. 1963 2 . I II II IV V 1 . 2 .
Dactylo oceanicae-Sedetum anglici J.-M. & J. Géhu & Caron 1978 corr.
. . . IV III II III . . . .
Géhu 2008
Catapodio marini-Trifolietum scabri Géhu & de Foucault 1978 nom. inv. 1 . . I r III II 1 . . .
Desmazerio marinae-Brometum ferronii Bioret 1994 1 . . II I II + 1 . . .
Soncho oleracei-Cochlearietum danicae Géhu & de Foucault 1978 . . . I r II II 1 . . .
Bromo ferronii-Trifolietum arvensis Géhu 2008 . . . I I r . 1 . . .
Catapodio marini-Parapholisetum incurvae Géhu & de Foucault 1978 1 . . . r II I 1 1 . .
Sedo anglici-Festucetum huonii Bioret 2014 . . . r I . + . . . .
Sedo anglici-Festucetum armoricanae Bioret et al. 2014 . . . + . . . . . . .
Sedo anglici-Scilletum vernae Bioret 1994 . . . . I . . . . . .
Bromo ferronii-Anthoxanthetum aristati Bioret, Bouzillé, de Foucault, Géhu
. . . . . . II . . . .
& Godeau 1988
Spergulario rupicolae-Salicornietum ramosissimae Bioret, Lahondère &
. . . . . . + . . . .
Khelifi 2001
Spergulario rupicolae-Plantaginetum coronopodis Bioret & Géhu 2013 . . . . . . + . . . .
Dactylo oceanicae-Limonietum dodartii (Lahondère 1979) Géhu, Franck &
. . . . . . . . 2 . .
Scoppola 1984
Dactylo hispanicae-Helichrysetum stoechadis Géhu, Franck & Scoppola
. . . . . . . . 1 . .
1984
Leucanthemo crassifolii-Helichrysetum staechadis (Allorge & Jovet 1941)
. . . . . . . . . . IV
Géhu & Géhu-Franck 1980
Gpt à Helichrysum stoechas et Helianthemum apenninum . . . . . . . . 1 . .
Pre-mantle
Orobancho hederae-Hederetum helicis Géhu (1961) 2006 . . . II II I . . . . .
Hedero helicis-Silenetum maritimae Géhu 2008 . . . I + I . . . . .
Hedero helicis-Ruscetum aculeati Géhu 2007 . . . r r . . . . . .
Nitrophilous grass
Atriplici hastatae-Betetum maritimae (Arènes 1933) Géhu 1968 . . II . r r . . . . .

1 – Sandstone-marly cliff geopermaseries of Boulonnais (Fig. 4)

This geopermaseries includes vegetation growing on marly cliffs, cut through by
Portlandian sandstone (Gris-Nez Cape, Alprech Cape). Meteorological data come
from the Calais station. The 10.8 °C mean annual temperature and 722.6 mm
average annual rainfall are characteristic of a subhumid Atlantic bioclimate. The
substrate is quite soft, with fresh water running slowly down from the top of the
cliffs, favorable to hygrophilous permaseries (Samolo valerandi-Caricetum vikin-
gensis Géhu et al. 1982, Apio graveolentis-Tussilaginetum farfarae Géhu et al.
1983 ex Bioret et Géhu 2008…).

2 – Marly and chalky-cliff geopermaseries of Boulonnais

This geopermaseries occurs on limestone cliffs of Cape Blanc-Nez (Pas-de-Calais).
The bioclimate is subhumid Atlantic. Only a few permaseries occur on the steep
slopes of Cape Blanc-Nez. The Cape Blanc-Nez cliffs are phytosociologically and
geomorphologically similar to those of Haute-Normandie, and involve in particular
the Dauco intermedii-Festucetum pruinosae (Liger 1956) Géhu and de Foucault
[32]. Maritime grassland is missing.

3 – Chalky-cliff geopermaseries of Pays de Caux (Fig. 5)

This geopermaseries occurs on the steep cliffs of Etretat. The 11.4 °C mean annual
temperature and 785 mm average annual rainfall (Cape Hève meteorological sta-
tion) are characteristic of a subhumid Atlantic bioclimate. A few permaseries grow
Geosynphytosociological Typology of French Atlantic Coastal … 255

Fig. 4 Theoritical zonation of the Boulonnais coastal cliffs permaseries

(1) Samolo valerandi-Caricetum vikingensis; (2) Apio graveolentis-Tussilaginetum farfarae;
(3) Spergulario rupicolae-Crithmetum maritimi; (4) Pulicario dysentericae-Juncetum inflexi;
(5) Ricardio pinguis-Eucladion verticillatii; (6) Cirsio acaulis-Festucetum pruinosae; (7) Contact:
agricultural crops.

in vertical side crevices (Brassicetum oleraceae [16]) or in less steep parts of the
cliff, where the substrate is deeper (Dauco intermedii-Festucetum pruinosae Liger
1956). Géhu and de Foucault [32].
The following three geopermaseries, growing on the ancient bedrock of the
Armorican massif, share the Armerio maritimae-Festucetum pruinosae Géhu [28]
maritime grassland. They are distinguished by the presence of differential associ-
ations, which indicate chorological or ecological variations.

4 – Crystalline-cliff geopermaseries of the Golfe Normand-Breton

This geopermaseries is characterized by a hyper-oceanic Atlantic subhumid bio-
climate. At the four meteorological stations the average annual temperature varies
between 11.9 °C and 12.2 °C, while the average annual rainfall varies between
615.5 mm (Granville) and 754.2 mm (Bréhat). The geopermaseries are differenti-
ated by the chasmo-halophytic vegetation of the Crithmo maritimi-Limonietum
normannici Lahondère & Bioret 1995 and the flayed grass of the Sedo
anglici-Festucetum armoricanae Bioret et al. 2014.
256 C. Demartini and F. Bioret

Fig. 5 Theoritical zonation of the Pays-de-Caux coastal cliffs permaseries

(1) Brassicetum oleraceae; (2) Dauco intermedii-Festucetum pruinosae; (3) Contact: agricultural
crops.
Geosynphytosociological Typology of French Atlantic Coastal … 257

Fig. 6 Theoritical zonation of the Western part of Massif armoricain coastal cliffs permaseries
(1) Armerio maritimae-Cochlearietum officinalis; (2) Spergulario rupicolae- Crithmetum mar-
itimi; (3) Desmazerio marinae-Brometum ferronii; (4) Armerio maritimae-Festucetum pruinosae;
(5) Dactylo oceanicae-Sedetum anglici; (6) Contact: Ulici maritimi-Ericetum cinereae.

5 – Crystalline-cliff geopermaseries of the western Massif Armoricain, under a

hyper-Atlantic climate (Fig. 6)
This geopermaseries, confined to extreme western France, is characterized by
hyper-Atlantic associations such as the Armerio maritimae-Cochlearietum anglicae
Géhu & Géhu-Franck 1984. Except for the stations at Brest and Penmarc’h, mean
annual temperatures range from 11.9 to 12.2 °C, and average annual rainfall is
between 819 and 1006 mm. Brest is slightly cooler (11.5 °C) and has more annual
precipitation (1210 mm). On the contrary, conditions are milder at Penmarc’h:
12.5 °C on average, and 664.4 mm of rainfall. The bioclimate is hyper-Atlantic
subhumid inferior or humid inferior.

6 – Crystalline-cliff geopermaseries of the southern Massif Armoricain, under

a thermo-Atlantic climate
More thermophilous than the previous, this geopermaseries prefers a mild climate,
less subject to storms. The temperatures are cooler at the northern limit of geop-
ermaseries (12 °C in Lorient) and higher at its southern limit (12.9 °C in Château
d’Olonne). Rainfall decreases progressively from north to south (950.9 mm in
Lorient to 761.7 mm in Château d’Olonne). The bioclimate is thermo-Atlantic
subhumid superior to humid inferior. The geopermaseries is defined mainly by the
chasmo-halophytic vegetation of the Apio graveolens-Rumicetum rupestris Bioret
& Géhu 2002 and the Crithmo maritimi-Limonietum ovalifolii (Kuhnholtz-Lordat
1926) Lahondère, Bioret & Botineau 1991.
258 C. Demartini and F. Bioret

Fig. 7 Theoritical zonation of the right bank of Gironde estuary cliffs permaseries
(1) Sagino maritimae-Spergularietum marinae; (2) Soncho maritimi-Schoenetum nigricantis;
(3) Crithmo maritimi-Limonietum ovalifolii; (4) Dactylo oceanicae-Limonietum dodartii;
(5) Astragalus monspessulanus and Helianthemum apenninum community; (6) Contact: Phillyreo
latifoliae-Quercetum ilicis.

7 – Crystalline-cliff geopermaseries of islands of the southern Massif

Armoricain, under a thermo-Atlantic climate
In general, islands are characterized by original vegetation, often linked to endemic
plant species and syntaxa, justifying the creation of new geopermaseries. The South
Armorican islands (Groix, Belle-Île, Yeu) are characterized by the Festuco
huonii-Plantaginetum littoralis (J.-M & J. Géhu 1976) Bioret, Bouzillé, de
Foucault, Géhu & Godeau 1988 and by the Bromo ferronii-Anthoxanthetum aristati
Bioret, Bouzillé, de Foucault, Géhu & Godeau 1988. Rainfall is lower than on the
mainland (from 701.4 to 769.8 mm per year). Temperatures are similar to those of
geopermaseries of the southern Armorican Massif (11.9 °C to 13 °C). The biocli-
mate is still thermo-Atlantic subhumid inferior or subhumid superior.

8 – Cliff-top dune geopermaseries on decalcified sand substrates

The distinction of this geopermaseries is more related to its ecology than to its
geographical distribution. It develops on low cliffs covered with a veneer of
decalcified sands. The characteristic association is the Tuberario guttatae-
Corynephoretum canescentis Géhu 1964. The bioclimate is thermo-Atlantic.

9 – Calcareous-cliff geopermaseries of the right-bank of the Gironde estuary,

under a thermo-Atlantic climate (Fig. 7)
Besides a chalky substrate, there is no other similarity between the geopermaseries
of the small limestone cliffs around Royan (Vaux-sur-Mer, Saint-Palais-sur-Mer)
and the Normando-Picardian geopermaseries. Located much farther south in
Geosynphytosociological Typology of French Atlantic Coastal … 259

Fig. 8 Theoritical zonation of the Basque Corniche coastal cliffs permaseries

(1) Crithmo maritimi-Plantaginetum maritimae; (2) Leucanthemo crassifolii-Helichrysetum stae-
chadis; (3) Leucanthemo crassifolii-Festucetum pruinosae; (4) Contact: Leucanthemo crassifolii-
Ericetum vagantis, Rubo ulmifolii-Tametum communis.
(A) Leucanthemo crassifoliae-Festuco pruinosae geopermasigmetum; (B) Contact: Rubo ulmifolii-
Tamo communis geominorisigmetum.

France, under a much milder climate, these vegetation types are more ther-
mophilous. Unfortunately, frequent tourist activity has led to the destruction of
most places where maritime grasslands can grow. The geopermaseries is charac-
terized by its chasmo-halophytic vegetation: the Sagino marinae-Spergularietum
marinae Bioret 2006 and the Soncho maritimi-Schoenetum nigricantis (Lahondère
1979) de Foucault 1984. The bioclimate is thermo-Atlantic.

10 - Chalky and sandstone-cliff geopermaseries around Biarritz, under a

thermo-Atlantic climate
After a gap of 230 km, where the landscape of Gironde and Landes is dominated by
coastal dunes, cliffs reappear south of the Adour (Pyrénées-Atlantiques). The mean
annual temperature is higher (14.3 °C in Biarritz), and average annual precipitation
is important (1450.9 mm in Biarritz). The rocky coast of Biarritz is composed of
calcareous sandstone and Nummulites calcareous sandstone. The geopermaseries is
very localized and characterized by Crithmo maritimi-Limonietum occidentalis
Pavillard 1928, strictly limited to this sector. The maritime grassland that grows on
260 C. Demartini and F. Bioret

the slopes is quite different from those found in other coastal geopermaseries. It has
not yet been characterized phytosociologically.

11 - Marly-chalky or sandstone cliff geopermaseries of the Corniche Basque,

under a thermo-Atlantic climate (Fig. 8)
At the extreme south of the Aquitaine Basin the rocky coasts of the Corniche
Basque stand out for their geologic and phytosociologic originality. These are high
cliffs, almost vertical, carved in the Cretaceous flysch. The chasmo-halophytic
vegetation and maritime grassland are synendemic to this coast: Crithmo
maritimi-Plantaginetum maritimae (Pavillard 1928) Guinea 1949, Leucanthemo
crassifolii-Festucetum pruinosae Géhu & Géhu-Franck 1980 corr. F. Pietro &
Loidi 1984, Leucanthemo crassifolii-Helichrysetum staechadis (Allorge & Jovet
1941) Géhu & Géhu-Franck 1980. As at Biarritz, temperature and precipitation are
much higher than in the north: 14.7 °C and 1483 mm annually at Socoa station.

Synsystem of Rocky Coastal cliffs of the French Channel-Atlantic Coastline

The 75 syntaxa are organized into the following synsystem:
AGROSTIETEA STOLONIFERAE Th. Müll & Gors 1969
Potentillo anserinae-Polygonetalia avicularis Tüxen 1947
Loto tenuis-Festucenalia arundinaceae palustris Julve ex de Foucault, Catteau &
Julve in de Foucault & Catteau 2012
Loto tenuis-Trifolion fragiferi (Westhoff, Van Leeuwen & Adriani 1962) de
Foucault 2008
Agrostio stoloniferae-Caricetum vikingensis Géhu 1979
Apio graveolens-Rumicetum rupestris Bioret & Géhu 2002
Samolo valerandi-Caricetum vikingensis Géhu 1982
Soncho arvensis-Rumicetum rupestris Bioret & Géhu 2002
Loto pedunculati-Cardaminenalia pratensis Julve ex de Foucault, Catteau & Julve
in de Foucault & Catteau 2012
Mentho longifoliae-Juncion inflexi Th. Müller & Görs ex de Foucault 2008
Pulicario dysentericae-Juncetum inflexi de Foucault in Royer et al.
2006

ARMERIO MARITIMAE-FESTUCETEA PRUINOSAE Bioret & Géhu 2008

Crithmo maritimi-Armerietalia maritimae Géhu 1964
Crithmo maritimi-Armerion maritimae Géhu 1968
Armerio maritimae-Asplenienion marini Géhu 2008
Armerio maritimae-Asplenietum marini Géhu & Géhu-Franck 1984
Crithmo maritimi-Limonienion binervosi Géhu & Géhu-Franck 1984
Armerio maritimae-Cochlearietum officinalis Géhu & Géhu-Franck
1984
Armerio maritimae-Inuletum crithmoidis Géhu 2006
Armerio maritimae-Plantaginetum coronopodis (Vanden Berghen
1979) Géhu & Bioret 2008
Armerio maritimae-Plantaginetum maritimae Géhu 2000
Crithmo maritimi-Betetum maritimae Géhu 2008
Crithmo maritimi-Juncetum acuti Bioret 2008
Crithmo maritimi-Juncetum maritimi Géhu & Duquef 2009
Crithmo maritimi-Limonietum normannici Lahondère & Bioret 1995
Crithmo maritimi-Limonietum occidentalis Pavillard 1928
Crithmo maritimi-Limonietum ovalifolii (Kuhnholtz-Lordat 1926)
Lahondère, Bioret & Botineau 1991
Geosynphytosociological Typology of French Atlantic Coastal … 261

Crithmo maritimi-Plantaginetum maritimae Guinea 1949

Crithmo maritimi-Senecionetum cinerariae Géhu 2005
Dactylo oceanicae-Limonietum dodartii (Lahondère 79) Géhu, Franck
& Scoppola 1984
Plantagini coronopodis-Spergularietum rupicolae Bioret & Géhu 2011
Spergulario rupicolae-Crithmetum maritimi (Roux & lahondère 1960)
Géhu 1962 nom. inv.
Spergulario rupicolae-Frankenietum laevis Bioret & Lahondère 2010
Spergulario rupicolae-Halimionetum portulacoidis Bioret & Géhu
2008
Spergulario rupicolae-Limonietum dodartii Géhu, Franck & Scoppola
1984
Spergulario rupicolae-Limonietum occidentalis Bioret & Géhu 2008
Spergulario rupicolae-Puccinellietum maritimae Bioret & Gallet 2014
Sileno maritimae-Festucenion pruinosae (Géhu & Géhu-Franck 1984) Bioret
& Géhu 2008
Apio graveolentis-Tussilaginetum farfarae Géhu et al. 1983 ex Bioret
& Géhu 2008
Armerio maritimae-Agrostietum maritimae Bioret & Géhu 2008
Armerio maritimae-Festucetum pruinosae Géhu 2008
Armerio maritimae-Juncetum gerardii Bioret, Demartini, Géhu &
Glémarec 2014
Carici arenarii-Festucetum pruinosae Bioret & Géhu 2008
Cirsio acaulis-Festucetum pruinosae Bioret & Géhu 2008
Dactylo oceanicae-Daucetum gummiferi Géhu 2008
Dauco gadeceaui-Festucetum pruinosae Bioret & Géhu 2008
Dauco intermedii-Festucetum pruinosae (Liger 1956) Géhu & de
Foucault 1978
Festuco huonii-Plantaginetum littoralis (J.-M & J. Géhu 1976) Bioret,
Bouzillé, de Foucault, Géhu & Godeau 1988
Festuco pruinosae-Elymetum atherici Géhu 2008
Festuco pruinosae-Ononidetum maritimae Bioret & Géhu 2008
Festuco pruinosae-Osmundetum regalis Géhu & Bioret 2000
Hedero helicis-Silenetum maritimae Géhu 2008
Jasiono maritimae-Anthyllidetum maritimae Géhu & Duquef 2009
Leucanthemo crassifolii-Festucetum pruinosae Géhu & Géhu-Franck
1980 corr. F. Pietro & Loidi 1984
Leucanthemo crassifolii-Helichrysetum staechadis (Allorge & Jovet
1941) Géhu & Géhu-Franck 1980
Orobancho hederae-Hederetum helicis Géhu (1961) 2006
Sileno zetlandicae - Festucetum pruinosae Bioret & Géhu 2008
Spergulario rupicolae-Armerietum maritimae Géhu 2008
Spergulario rupicolae-Silenetum maritimae Bioret & Géhu 2000
Gpt à Carex cf. repens et Festuca pruinosa
Gpt à Festuca pruinosa et Heracleum sphondylium subsp. trifoliolatum
Gpt à Festuca pruinosa et Spartina versicolor
Brassicion oleracei Rivas-Martínez, Fernández González & Loidi 1999
Brassicetum oleraceae Géhu 1962

ARRHENATHERETEA ELATIORIS Braun-Blanq.1949 nom. nud.

262 C. Demartini and F. Bioret

Plantaginetalia majoris Tüxen ex Von Rochow 1951

Lolio perennis-Plantaginion majoris Sissingh 1969
Anthemido nobilis-Agrostietum capillaris (Allorge 1922) de Foucault in
Royer et al. 2006

ASPLENIETEA TRICHOMANIS (Braun-Blanq. in Meier & Braun-Blanq. 1934) Oberd. 1977

Androsacetalia vandellii Braun-Blanquet in Meier & Braun-Blanquet 1934 nom. corr. in
Bardat et al. 2004
Asplenio billotii-Umbilicion rupestris de Foucaut 1988
Umbilico rupestris-Asplenietum billoti de Foucault 1979
Umbilico rupestris-Asplenietum obovati Géhu & Biondi 1994

ASTERETEA TRIPOLII Westhoff & Beeftink in Beeftink 1962

Glauco maritimae-Puccinellietalia maritimae Beeftink & Westhoff in Beeftink 1962
Limonio vulgaris-Plantaginenion maritimae Géhu & Géhu-Franck 1984 nom. nud.
Glauco maritimae-Juncion maritimi Géhu & Géhu-Franck ex Géhu in Bardat
et al. 2004
Junco maritimi-Caricetum extensae (Corillion 1953) Géhu 1976

CAKILETEA MARITIMAE Tüxen & Preising in Tüxen 1950

Cakiletalia integrifoliae Tüxen ex Oberdorfer 1950 corr. Rivas-Martínez, Costa & Loidi
1992
Atriplicion littoralis Nordhagen 1940
Atriplici hastatae-Betetum maritimae (Arènes 1933) Géhu 1968

CRATAEGO MONOGYNAE-PRUNETEA SPINOSAE Tüxen 1962

Prunetalia spinosae Tüxen 1952
Lonicerion periclymeni Géhu, de Foucault & Delelis 1983
Hedero helicis-Ruscetum aculeati Géhu 2007

HELIANTHEMETEA GUTTATI (Braun-Blanq. ex Rivas Goday 1958) Rivas Goday & Rivas
Mart. 1963 em. Boullet 2004
Helianthemetalia guttati Braun-Blanquet in Braun-Blanquet, Molinier & Wagner 1940
Thero-Airion Tüxen ex Oberdorfer 1957
Bromo ferronii-Anthoxanthetum aristati Bioret, Bouzillé, de Foucault, Géhu
& Godeau 1988
Tuberario guttatae-Romuleetum columnae Provost 1978

ISOETO DURIEUI-JUNCETEA BUFONII Braun-Blanq. & Tüxen ex V.West., Dijk & Paschier
1946
Isoetalia velatae (Braun-Blanquet 1936) de Foucault 1988
Ophioglosso lusitanici-Isoetion histricis (Braun-Blanquet 1936) de Foucault 1988
Ophioglosso lusitanici-Isoetetum histricis de Foucault 1988

KOELERIO GLAUCAE-CORYNEPHORETEA CANESCENTIS Klika in Klika & V.Novak

1941
Corynephoretalia canescentis Klika 1934
Miboro minimae-Corynephorion canescentis Loiseau & Felzines 2007
Tuberario guttatae-Corynephoretum canescentis Géhu 1964
Geosynphytosociological Typology of French Atlantic Coastal … 263

PHRAGMITI AUSTRALIS-MAGNOCARICETEA ELATAE Klika in Klika & V.Novák 1941

Scirpetalia compacti Hejny in Holub, Hejny, Moravec & Neuhausl 1967 corr. Rivas-
Martínez, J.C. Costa, Castroviejo & Valdes Berm. 1980
Scirpion compacti A.E.Dahl & Hadac 1941 corr. Rivas- Martínez, J.C. Costa,
Castroviejo & Valdes Berm. 1980
Scirpetum compacti (Van Langendonck 1931) Beeft. 1957

SAGINETEA MARITIMAE V.Westh., C.Leeuwen & Adriani 1962

Saginetalia maritimae Westhoff, C.Leeuwen & Adriani 1962
Saginion maritimae Westhoff, C.Leeuwen & Adriani 1962
Bromo ferronii-Trifolietum arvensis Géhu 2008
Catapodio marini-Parapholisetum incurvae Géhu & de Foucault 1978
Catapodio marini-Trifolietum scabri Géhu & de Foucault 1978 nom. inv.
Desmazerio marinae-Brometum ferronii Bioret 1994
Sagino maritimae-Catapodietum marini R. Tx. 1963
Soncho oleracei-Cochlearietum danicae Géhu & de Foucault 1978
Frankenietalia pulverulentae Rivas-Martínez ex Castroviejo & Porta 1976
Frankenion pulverulentae Rivas-Martínez ex Castroviejo & Porta 1976
Sagino maritimae-Spergularietum marinae Bioret 2006

SCHEUCHZERIO PALUSTRIS-CARICETEA FUSCAE Tüxen 1937

Caricetalia davallianae Braun-Blanquet 1949
Hydrocotylo vulgaris-Schoenion nigricantis de Foucault 2008
Caricenion pulchello-trinervis Julve ex de Foucault 2008
Soncho maritimi-Schoenetum nigricantis (Lahondère 1979) de
Foucault 1984

SEDO ALBI-SCLERANTHETEA BIENNIS Braun-Blanquet 1955

Sedo albi-Scleranthetalia biennis Braun-Blanquet 1955
Sedion anglici Braun-Blanquet in Braun-Blanquet & Tüxen 1952
Dactylo oceanicae-Sedetum anglici J.-M. & J. Géhu et Caron 1978 corr.
Géhu 2008
Sedo anglici-Scilletum vernae Bioret 1994
Festucenion huonii Bioret et al. 2014
Sedo anglici-Festucetum armoricanae Bioret et al. 2014
Sedo anglici-Festucetum huonii Bioret et al. 2014

THERO-SUAEDETEA SPLENDENTIS Rivas Mart. 1972 em. Rivas Mart., Fern.Gonz. & Loidi
1998
Thero-Salicornietalia dolichostachyae Tüxen ex Boullet & Géhu in Bardat et al. 2004
Salicornion europaeo-ramosissimae Géhu & Géhu-Franck ex Rivas-Martínez 1990
Spergulario rupicolae-Salicornietum ramosissimae Bioret, Lahondère &
Khelifi 2001

5 Conclusion

By highlighting 11 geopermaseries, the geoserial analysis revealed the diversity of

plant landscapes of French Channel-Atlantic rocky coastline.
264 C. Demartini and F. Bioret

By studying catenas at the regional scale, geosynphytosociology provides an

overview of the layout of coastal vegetation, at the national scale.
This study shows that geopermaseries of the supra-littoral zone are characterized
by particular geological, geomorphological, bioclimatic, and phytoecological
contexts.
For conservation, analysis of geopermasynrelevés can identify the most diverse
or the most vulnerable coastal areas and vegetation (incomplete geopermaseries,
presence of indicator associations of anthropogenic disturbance…) in order to
implement appropriate protection measures.

References

1. Bardat J, Bioret F, Botineau M, Boullet V, Delpech R, Géhu J-M, Haury J, Lacoste A,

Rameau J-C, Roux G, Royer J-M, Touffet J (2004) Prodrome des végétations de France.
Collection Patrimoines naturels, 61. Muséum national d’histoire naturelle, Paris, p 171
2. Béguin C, Géhu J-M, Hegg O (1979) La synphytosociologie: une approche nouvelle des
paysages végétaux. Doc Phytosociol N.S. IV:49–69
3. Bioret F (1989) Contribution à l’étude de la flore et de la végétation de quelques îles et
archipels ouest et sud armoricains. Ph. D. thesis, Nantes University, p 480
4. Biondi E, Calandra R, Gigante D, Pignatelli S, Rampiconi E, Venanzoni R (2002) Il
paesaggio vegetale della provincia di Terni. Arti grafiche Iezzi, Terni, p 104
5. Bioret F, Géhu J-M (2002) Révision phytosociologique des végétations à Rumex rupestris sur
les littoraux atlantiques européens. J Botanique de la Société Botanique de France 20:45–54
6. Bioret F, Géhu J-M (2008) Révision phytosociologique des végétations halophiles des
falaises littorales atlantiques françaises Fitosociologia 45(1):75–116
7. Bioret F, Lahondère C (2010) Les végétations à Frankenia laevis et à Limonium à nervures
parallèles des hauts schorres sablonneux et des falaises du littoral atlantique français.
Braun-Blanquetia 46:129–137
8. Bioret F, Lazare J-J, Géhu J-M (2011) Evaluation patrimoniale et vulnérabilité des
associations végétales du littoral atlantique français. J Botanique de la Société Botanique de
France 56:39–67
9. Braun-Blanquet J (1928) Pflanzensoziologie. Grundzüge der Vegetationkunde. Biol.
Studientbücher 7:1–330
10. Braun-Blanquet J (1932) Plant sociology - The study of plant communities (translation from
Pflanzensoziologie by Fuller & Conard), p 439
11. Braun-Blanquet J, Pavillard J (1922) Vocabulaire de sociologie végétale, vol 1, p 16.
Montpellier
12. De Foucault B, Bioret F (2010) Contribution au prodrome des végétations de France: les
Saginetea maritimae Westoff et al. 1962. J de Botanique de la Société Botanique de France
50:59–83
13. De Foucault B, Catteau E (2012) Contribution au prodrome des végétations de Botanique de
France 59:5–131
14. Demartini C, Bioret F, Lazare J-J (2017) Typologie et cartographie des géopermaséries et des
géocurtaséries des végétations littorales des côtes Manche-Atlantique françaises. Doc
Phytosociol, Série 3, 6:316–332
15. Dupias G, Rey P (1985) Document pour un zonage des régions phyto-écologiques. CNRS,
p 39
Geosynphytosociological Typology of French Atlantic Coastal … 265

16. Géhu J-M (1962) Quelques observations sur la falaise crétacée du Cap Blanc-Nez (P. de C.) et
étude de la végétation de la paroi abrupte: Brassicetum oleraceae ass. nov. Bulletin de la
Société Royale de Botanique de Belgique 95(1):109–129
17. Géhu J-M (1963) L’excursion dans le nord et l’ouest de la France de la Société internationale
de phytosociologie. Bulletin de la Société Botanique du Nord de la France 16(3):105–189
18. Géhu J-M (1974) Sur l’emploi de la méthode phytosociologique sigmatiste dans l’analyse, la
définition et la cartographie des paysages. Comptes Rendus de l’Académie des Sciences Paris
279(Série D):1167–1170
19. Géhu J-M (1975) Essai systématique et chorologique des principales associations du littoral
atlantique français. Los anales de la real academia de farmacia XLI(2):207–227
20. Géhu J-M (1977) Le concept de sigmassociation et son application à l’étude du paysage
végétal des falaises atlantiques françaises. Vegetatio 34:117–125
21. Géhu J-M (1979) La phytocoenologie comme outil de définition des secteurs côtiers à
préserver. Publications du CNEXO: Actes de Colloques, no 9, pp 154–164
22. Géhu J-M (1986) Des complexes de groupements végétaux à la phytosociologie paysagère
contemporaine. Informatore Botanico Italiano 18:53–83
23. Géhu J-M (1988) Schéma synsytématique des principales classes de végétations littorales
sédimentaires européennees avec références à d’autres territoires holarctiques. Annali di
Botanica LVI(1):5–52
24. Géhu J-M (1991) L’analyse synphytosociologique et géosynphytosociologique de l’espace.
Théorie et méthodologie. Colloques Phytosociologiques, Phytosociologie et Paysages,
Versailles, XVII:11–45
25. Géhu J-M (1994) Schéma synsystématique et typologique des milieux littoraux français
atlantiques et méditerranéens. Colloques Phytosociologiques, Typologie phytosociologique
des habitats, Bailleul, XXII:183–212
26. Géhu J-M (1995) Inventaire typologique des milieux terrestres littoraux menacés à des fins
conservatoires. Vol I-VI. Fondation de France, les côtes de France (Corse inclue), p 608
27. Géhu J-M (1998) Schéma synsystématique des principales classes de végétations littorales
sédimentaires européennes avec références à d’autres territoires holarctiques. Annali di
Botanica 61(1):5–52
28. Géhu J-M (2008) Étude des associations végétales des sentiers littoraux de Dinard et Saint-
Énogat (France, Ille-et-Vilaine) suivie d’un guide itinéraire. J Botanique de la Société
Botanique de France, 41:47–80
29. Géhu J-M (2009) Signification phytosociologique de Brassica oleracea ssp. oleracea sur les
falaises atlantiques européennes. Fitosociologia 46(1):3–10
30. Géhu J-M, Bioret F (2000) Les végétations à Osmunda regalis des falaises littorales
armoricaines. Colloques Phytosociologiques, Données de la Phytosociologie sigmatiste:
Structure, gestion, utilisation, Bailleul, XXVII:203–210
31. Géhu J-M, Géhu-Franck J, Bournique C (1984) Sur les étages bioclimatiques de la région
eurosibérienne française. Doc Phytosociol N.S., VIII:29–43
32. Géhu J-M, De Foucault B (1978) Phytosociologie de la pelouse aérohaline des falaises de
craie de Haute-Normandie (France). Doc Phytosociol N.S. 3:289–294
33. Géhu J-M, Duquef H (2009) Observations phytosociologiques complémentaires sur les
falaises de la Côte d’émeraude (France, Côtes-d’Armor et Ille-et-Vilaine). Bulletin de la
Société Botanique du Centre-Ouest, N.S. 40:231–280
34. Géhu J-M, Franck J, Bournique C (1982) La végétation du littoral Nord-Pas-de-Calais.
Synthèse phytocoenologique. Université de Paris V, p 361
35. Géhu J-M, Franck J, Scoppola A (1984) Observations sur la végétation aérohaline des falaises
maritimes du Centre-Ouest français. Doc Phytosociol N.S. VIII:147–164
36. Géhu J-M, Géhu J (coll. M. Bigot) (1980) Aperçu phytosociologique sur les falaises
d’Hendaye et de Saint-Jean-de-Luz (Pays basque). Doc Phytosociol N.S. 5:363–374
37. Géhu J-M, Rivas-Martínez S (1981) Notions fondamentales de phytosociologie. In:
Dierschke H (ed) Syntaxonomie. Berichte der Internationalen Symposien der
Internationalen Vereinigung für Vegetationkunde, pp 5–33
266 C. Demartini and F. Bioret

38. Lahondère C (1980) La végétation des falaises autour de Biarritz. Bulletin de la Société
Botanique du Centre-Ouest, N.S. 10:37–44
39. Lahondère C (1986) La végétation des falaises des côtes charentaises. Bulletin de la Société
Botanique du Centre-Ouest, N.S. 17:33–53
40. Lahondère C (1990) Les pelouses sèches littorales autour de Royan (Charente-Maritime).
Bulletin de la Société Botanique du Centre-Ouest, N.S. 21:29–39
41. Lahondère C (2005) Les formations sèches de la partie charentaise de l’estuaire de la Gironde,
de Saint-Palais-sur-Mer à Mortagne-sur-Gironde. Compte rendu de la session phytosoci-
ologique 2004. Bulletin de la Société Botanique du Centre-Ouest, N.S. 36:481–512
42. Lalanne A (2017) Le programme de cartographie nationale des végétations naturelles et
semi-naturelles de France: CarHAB, Doc Phytosociol Série 3, 6:236–244
43. Lazare J-J (2009) Phytosociologie dynamico-caténale et gestion de la biodiversité. Acta
Botanica Gallica 156(1):49–61
44. Lazare J-J, Lanniel K (2003) Une sous-association nouvelle de fourrés du Rubo
ulmifolii-Tametum communis du littoral basque. J Botanique de la Société Botanique de
France 21:33–35
45. Lazare J-J, Dartiguelongue S, Dicharry C, Lartigue A (2001) Érosion des communautés
végétales dunaires du littoral aquitain. Poster, XVIIIe Jornadas de fitosociología,
“Condiciones extremas y vegetación”, León, 19–22 IX 2001
46. Rivas-Martínez S (1976) Sinfitosociologia, una nueva metodologia para el estudio del paisaje
vegetal. Anales del instuto Botanico Cavanilles 33:179–188
47. Rivas-Martínez S (2005) Notions on dynamic-catenal phytosociology as a basis of landscape
science. Plant Biosystems 139(2):135–144
48. Rivas-Martínez S (2007) Mapa de series, geoseries y geopermaseries de vegetación de España
(Mémoria del mapa de vegetacion potential de España) parte I. Itinera Geobotanica, 17(new
series):1–435
49. Rivas-Martínez S (2011) Mapa de series, geoseries y geopermaseries de vegetación de España
(Mémoria del mapa de vegetacion potential de España) parte II. Itinera Geobotanica, 18
(2):425–800
50. Tüxen R (1973) Vorschlag zur Aufnahme von Gesellschaftskomplexen in potentiell
natürlichen Vegetationgebieten. Acta Botanica Academiae Scientiarum Hungaricae 19(1–
4):379–384
Symphytosociology, a Tool
for Landscape Monitoring: A Case
Study from the Swiss Alps

Jean-Paul Theurillat, Michele Di Musciano, Olivier Duckert,

and Claude Béguin

Abstract A comparison of vegetation using a symphytosociological approach was

carried out in the Aletsch region, a UNESCO World Heritage Centre located around
the Aletsch Glacier in the Valais Alps (Switzerland). For this, relevés performed in
1982 and 2014 in 61 operational geographic units (OGUs) located near the treeline,
with a mean elevation above 2000 m, were analysed. During the 32-year interval
between the two surveys, several environmental changes occurred, mainly a decrease
in traditional pasturing, a 1.5 °C increase in the mean annual temperature, and an
increase in deer population. Numerical analyses of the relevés showed that the
landscape remained globally stable. Nevertheless, in 56% of the OGUs, there was a
change in cover greater than 10% either for trees (> 2.5 m), dwarf shrubs (< 0.5 m) or
the herbaceous vegetation. In particular, Picea abies increased at lower elevation, and
Pinus cembra dominated the regeneration of young trees (0.5–2.5 m) outside the
forested area. Signs of thermophilisation were found in 74% of the OGUs. Deer were
present in all the OGUs, and in some places, they had a strong impact on the vegetation,
such as in the Aletschwald Nature Reserve. The resurvey confirms that symphy-
tosociological relevés are a promising tool for landscape monitoring.

J.-P. Theurillat (&)

Centre Alpien of Phytogeography, Foundation J.-M. Aubert, CP 71, 1938 Champex-Lac,
Switzerland
e-mail: [email protected]
J.-P. Theurillat
Department of Botany and Plant Biology, Section of Biology, University of Geneva,
CP 71, 1292 Chambésy, Switzerland
M. Di Musciano
Department of Life, Health and Environmental Sciences, University of L’Aquila,
67100 Coppito, Italy
e-mail: [email protected]
O. Duckert
Grenat Sarl, 1955 Chamoson, Switzerland
e-mail: [email protected]
C. Béguin
2067 Chaumont, Switzerland
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 267
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_14
268 J.-P. Theurillat et al.

Keywords Symphytosociology Vegetation Landscape Deer Climate







change Thermophilisation Monitoring Switzerland European Alps Valais

1 Introduction

Symphytosociology (also termed symphytocoenology) is a branch of phytosoci-

ology aiming at the analysis of vegetation specifically at landscape scale. It is based
on comparison of quantitative relevés of all the vegetation units occurring within a
given portion of a landscape or operational geographic unit (OGU). The particular
assembly of vegetation units forms a so-called vegetation complex, and abstract
types of such complexes are called sigmassociations,
P a term meaning “a sum of
associations” (from the Greek letter sigma, , meaning sum). Consequently, it is
possible to build a typology to describe the recurring vegetation complexes within a
landscape. According to the complexity of the landscape itself, and the size of the
OGU, different organisational levels of sigmassociations have been distinguished.
The theoretical ground and the methodological approach were developed in the
1970s, mainly by Tüxen [14], Béguin and Hegg [2], Rivas-Martínez [9] and Géhu
[6], see also Theurillat [12, 13] and Béguin et al. [3]. Several approaches have been
developed to define sigmassociations. Among them, the “Swiss School” [5, 10, 11]
proceeds with a preliminary delimitation of the study area in uniform OGUs
(landscape units), in which only one kind of potential vegetation would likely
occur. This exhaustive approach allows a simultaneous mapping of the vegetation
landscape of the study area.
The great advantage of the symphytosociological relevé is that it can include
vegetation units at any level, from very broad units at the level of the phytosoci-
ological class to very detailed units, corresponding to fine variations of a given
association (variant, facies), fragmentary aspect, or basal unit in the sense of
Kopecký and Hejný [12]. In this respect, such relevés can distinguish particularities
within the landscape and small differences in its use. Compared to detailed mapping
of the vegetation, there is a trade-off between precision at the unit level and that of
the location, a kind of Heisenberg’s uncertainty principle. The symphytosocio-
logical relevé is not limited regarding the precision of the vegetation units but
cannot offer precision in their location within the area of the relevé, in as much as
the surface occupied by the vegetation unit is very small. Conversely, fine-scale
mapping allows great precision of location but only down to the resolution per-
mitted by the scale; hence, not enabling the mapping of very tiny patches.
In this contribution, we are interested in comparison over a 32-year interval
(1982 versus 2014) of the vegetation landscape of a section of the Aletsch region,
which is a UNESCO World Heritage Centre located around the Aletsch Glacier in
the Valais Alps (Switzerland). In the years 1981/82 (hereafter 1982), the vegetation
landscape was studied along an elevation transect from 700 m up to 2230 m. Since
then, several environmental changes co-occurred in this region:
Symphytosociology, a Tool for Landscape Monitoring … 269

(1) The traditional land use with pasturing decreased strongly in the 1980s, espe-
cially in the upper part of the study area.
(2) The mean annual temperature increased by 1.5 °C during the 32-year interval.
(3) Deer population, which was scarce at the beginning of the 1980s, increased
drastically in the 1990s, especially on the north side of the Aletschwald Nature
Reserve, due to decreased human activity. The deer population fluctuates
between 100 and >200 individuals on less than 10 km2.
In order to see if these environmental changes had already produced a diachronic
change at the landscape level, especially around the tree line, the upper part of the
transect was resurveyed during the years 2013/14 (hereafter 2014), and the land-
scape relevés of the two surveys were compared.

2 Methods

2.1 Location

The study area is located in the upper part of the right-hand side of the Rhone River
valley (46°22′ N, 8°02′ E; Fig. 1). The bedrock is siliceous and pertains to the
southern border of the Aar massive. An elevation transect of 8.5 km2 was inves-
tigated in 1981/82, from the bottom of the Rhone River at 700 m up to the crest at
the tree line at 2230 m (southeast side) and then down to 1600 m, to reach the Great
Aletsch Glacier (northwest side). The crest can be divided into a western part, with
the Riederhorn reaching 2230 m, a central part at Riederfurka (2065 m) and an
eastern part up to Hohflue (2227 m).
Until the 1950s, the southeast side of the study area was used intensively in
traditional farming, with cultivation up to the 1500 m, hay up to 1900 m, pasturing
at a higher elevation, and wood-cutting up to the tree line. Pasturing in the upper
part, above 2000 m, was maintained quite intensively until the end of the 1970s and
then extensively until the 2000s. Tourism developed in the 1960s, mainly skiing
activities during winter and hiking during summer. On the northwest side, there has
been intensive pasturing in the upper part of the eastern part together with
wood-cutting in the past. However, since 1933, the Aletschwald in the eastern part
was made a nature reserve from the glacier to the crest, and so human activity has
subsided, except hiking along strictly limited paths.

2.2 Symphytosociological Relevés

For the 1981/82 study [4, 12, 13], the area was delimited exhaustively in 1–8 ha
OGUs, with the aid of tatopographical map (1:10,000) and aerial views, based on
three main criteria of more or less homogenous (a) slope, (b) aspect and
270 J.-P. Theurillat et al.

6°E 7°E 8°E 9°E 10°E

GERMANY

FRANCE

AUSTRIA

47°N
Bern 47°N
SWITZERLAND

Aletsch

ITALY 46°N
46°N

6°E 7°E 8°E 9°E 10°E

Fig. 1 Location of the study area (Aletsch) on the right-hand side of the Rhone river valley
(Valais, Switzerland)

(c) vegetation structure. Along the whole elevation transect from the Rhone River to
the glacier, 289 OGUs were delimited according to the above criteria. The sym-
phytosociological relevés of the OGUs were performed with the help of the
topographic map and a Thommen altimeter. In each relevé, the following elements
were recorded and quantified (see Table 1 as an example):
(a) The vegetation units, according to a list of as detailed as possible units. The
units were established with the assistance of vegetation relevés, performed
during a preliminary survey. These are working units where physiognomy is
emphasised; thus, corresponding to subassociations, variants or facies of
already, formally described vegetation units and, otherwise, to undetermined
physiognomic, fragmentary or basal units (see [12]). In addition, some of the
vegetation units, usually the dominant ones, were subdivided further according
to the state of their development and visible disturbance (e.g., vegetation
normally developed, vegetation poorly developed). If needed, these stages
could be combined during data analysis. Each unit bears an identifier number
that is indicated in the relevé, which is used in the numerical analyses.
Examples of working units for mesothermophilous dwarf scrub dominated by
Juniperus communis ssp. alpina are as follows:
775 Vaccinio-Juniperetum typical ø
776 Vaccinio-Juniperetum typical °
777 Vaccinio-Juniperetum ‘sphagnetosum’
778 Vaccinio-Juniperetum ‘rhododendretosum’ ø
Symphytosociology, a Tool for Landscape Monitoring … 271

779 Vaccinio-Juniperetum ‘saxifragetosum aizoon’

780 Vaccinio-Juniperetum ‘dryopteridetosum spinulosae’
782 Vaccinio-Juniperetum ‘festucetosum variae’
RH1 Vaccinio-Juniperetum ‘callunetosum’
RH37 Vaccinio-Juniperetum with Rubus idaeus
RH46 Vaccinio-Juniperetum degraded by herbicide
RH49 Vaccinio-Juniperetum with Juniperus sabina
RH52 Vaccinio-Juniperetum typical var. Epilobium angustifolium (some-
how ruderal)
RH53 Vaccinio-Juniperetum typical var. Urtica dioica

The numbers 775–782 correspond to units established during the 1982 survey;
numbers RH1–RH53 represent units added during the resurvey in 2014; ø
designates normally developed, and ° denotes poorly developed.

Table 1 Relevés in 1982 and 2014 of the operational geographic units (OGUs) dominated by the
mesophilous cembran pine forest. Manual classification: see text. The changes mainly attributable
to deer between 1982 and 2014 are highlighted in light grey. Data related to geomorphology, land
use, human activities, phanerophytic structure and isolated trees have been omitted
272 J.-P. Theurillat et al.

(b) The main vegetation structure (trees, shrubs, dwarf shrubs, herbs – both closed
and open herbaceous vegetation) was evaluated separately and quantified in the
field. This approach helps to control the quantification of each single vegetation
unit; the sum of the cover of all the vegetation units of a given structure must
correspond to the cover of the main vegetation structure to which they belong.
Isolated trees were recorded separately and placed in two categories of height
(0.5–2.5 m and > 2.5 m).
(c) The phanerophytic vegetation (> 0.5 m height) was divided further into <
2.5 m and > 2.5 m isolated broadleaf trees, < 2.5 m and > 2.5 m isolated
coniferous trees, coniferous patches, coniferous forests and native shrubs.
(d) The predominant land use (mowing, pasturing, cultivation, fallow, manure).
(e) Geomorphological features (e.g., terrain roughness, screes, aquatic elements).
(f) All anthropogenic activities (e.g., types of constructions, types of roads and
other kinds of transport, ski runs).
Except for surfaces up to 10 m2, the presence of all the above features was
quantified in the OGUs in proportion of their occurrence, according to height
classes of cover (1 =  10 m2, 2 =  1%, 3 = 1–5%, 4 = 5–10%, 5 = 10–25%,
6 = 25–50%, 7 = 50–75%, 8 = 75–90%, 9 = > 90%). A manual classification
(tables of relevés) produced seven main landscape units.

2.3 Resurvey

For the 2014 study, 61 OGUs with mean elevation above 2000 m, covering 1,736
km2, were re-sampled with the help of a Garmin 60 GPS and a 1:10,000 ortho-
photo. The delimitations of the OGUs in 1980 were previously reported on the
orthophoto (Fig. 2). Where needed, the delimitations were corrected in the field and
were fixed with additional GPS points. The same vegetation units as those for the
1982 survey were used, with new units added when needed.
Symphytosociology, a Tool for Landscape Monitoring … 273

Fig. 2 Example of operational geographic units (OGUs) reported on the orthophoto for the 2014
survey. OGUs were delimited in 1980 on a 1:10,000 topographic map, according to slope, aspect
and vegetation structure

2.4 Numerical Analyses

Data from both the 1982 and 2014 studies were analysed numerically in performing
Ward D’s cluster analysis, McQuitty’s principal component analysis (PCA), and
canonical correspondence analysis (CCA) with the VEGAN package in R (https://
github.com/vegandevs/vegan), using a rank transformation of 1–9 for the nine
classes of the cover data. The analyses were made by lumping the states “vegetation
normally developed” and “vegetation poorly developed” into a single unit.
Lumping is a conservative approach that minimises the differences between the
OGUs, to get closer to the manual classification.

2.5 Nomenclature

If not otherwise mentioned, the nomenclature of the species followed Aeschimann

et al. [1]. Trees and shrubs: cembran pine (Pinus cembra; larch (Larix decidua);
spruce (Picea abies L.); green alder (Alnus viridis); Swiss willow (Salix helvetica);
Juniperus nana Willd. = J. communis ssp. alpina Celak.
274 J.-P. Theurillat et al.

3 Results

3.1 Classification of the 1982 Relevés

The symphytosociological relevés of the 1982 survey were analysed manually [12].
For the upper part of the transect, above 2000 m, seven main vegetation landscape
units and three main sub-groups were identified within the 61 OGUs. These seven
landscapes (between brackets the more importantly as diagnostic associations
determining the landscape type when they are dominant in the OGU) were clas-
sified as follows:
A: landscape of cliffs (Asplenio-Primuletum hirsutae, Festucetum variae s.l.)
B: landscape of boulders (Cryptogrammetum crispae)
C: landscape of thermophilous cembran pine forests (Cotoneastro integerrimae-
Pinetum cembrae)
D: landscape of spruce forests (Homogyno alpinae-Piceetum abietis)
E: landscape of mesophilous cembran pine forests (Larici deciduae-Pinetum
cembrae) with a rocky variant (Ea, Riederhorn) and a gentler variant (Eb,
Aletschwald)
F: landscape of thermophilous dwarf heaths (Junipero nanae-Arctostaphyletum
uvae-ursi) with a warm and a cold variant
G: landscape of mesophilous heaths s.l. (Vaccinio-Juniperetum nanae,
Vaccinio-Rhododendretum ferruginei) with the sub-groups
G1: mesothermophilous dwarf heaths (Vaccinio-Juniperetum nanae, Geo
montani-Nardetum strictae) with a warm, a crest, and a cold variant
G2: dwarf heaths with Calamagrostis villosa (Vaccinio-Juniperetum nanae,
Calamagrostietum villosae s.l.)
G3: cryophilous dwarf heaths (Vaccinio-Rhododendretum, Empetro-
Vaccinietum).
The manual classification was globally validated by Ward’s cluster analysis,
performed on the 1982 relevés (Fig. 3). The main divisions of the cluster analysis
were also reinforced with PCA (not shown).
The principal division in the cluster analysis separated the OGUs of the forested
landscape where no pasturing occurs (cluster 1) from the OGUs of the non-forested
landscape submitted to pasturing (cluster 2). The group with OGUs not submitted to
pasturing (cluster 1) was divided further into two geographic-ecological clusters,
the first (cluster 1.1) corresponding to the western part of the study area with rockier
slopes (Riederhorn), and the second (cluster 1.2) associated with the less rocky
northern aspect of the eastern part of the study area (Riederfurka, Aletschwald). The
western cluster 1.1 was subdivided further into spruce forests (D) (cluster 1.1.1) and
cembran pine forests (cluster 1.1.2), with a division occurring in cluster 1.1.2
between two OGUs of rocky thermophilous cembran pine forests (C) (cluster
1.1.2.1) and less rocky OGUs with thermophilous cembran pine forests and
mesophilous cembran pine forests (E) (cluster 1.1.2.2). The eastern cluster 1.2
Symphytosociology, a Tool for Landscape Monitoring … 275

Fig. 3 Ward’s cluster analysis (Ward D) of the 1982 relevés. A–G: former manual classification
(see text). Cluster 1: forested operational geographic units (OGUs) with no pasturing. Cluster 2:
non-forested OGUs submitted to pasturing

mainly contained OGUs of the landscape of the cryophilous dwarf heaths (G3)
(cluster 1.2.1) and mesophilous cembran pine forests (E) (cluster 1.2.2).
Cluster 2, where pasturing occurs, was divided into two parts. Cluster 2.1 cor-
responded to heaths that were still more or less pastured, and cluster 2.2 resembled
OGUs with a relatively lower pasturing pressure. Cluster 2.1 was then divided into
four geographic-ecological sub-clusters. On the one hand, the south of the eastern
part of the study area (2.1.1), with a further subdivision into thermophilous dwarf
heaths (F) (cluster 2.1.1.1) and mesophilous dwarf heaths (G1) (cluster 2.1.1.2);
and, on the other hand, the mesophilous dwarf heaths (G1) of the south side of the
central part of the study area that were further subdivided into an upper (cluster
2.1.2.1) and lower part (cluster 2.1.2.2). Cluster 2.2 contained OGUs with a
comparatively lower pasturing pressure. This was a mixture of mesophilous heaths
(G1), heaths with Calamagrostis villosa (G2), thermophilous dwarf heaths (F),
boulders (B) and a cliff (A). This was divided further, on the one hand, into a
sub-cluster of the eastern part of the Riederhorn (cluster 2.2.1.1) and a sub-cluster
of the southwest part of the Riederhorn (cluster 2.2.1.2); and, on the other hand,
into a sub-cluster of thermophilous dwarf heaths in the eastern part (cluster 2.2.2.1)
plus a mixed sub-cluster 2.2.2.2, containing the unique OGUs occupied with a cliff.
The most important units of the vegetation landscape obtained by the manual
classification could be recognised as sub-groups in the numerical classification,
although the divisions are not always clear-cut. For the landscape units recognised
by manual classification, cluster analysis showed the following:
(1) The OGUs containing mesophilous cembran pine forests (E) were separated
geographically. This finding can be explained because, on the rocky slopes on
276 J.-P. Theurillat et al.

the north side of the Riederhorn, a mixture of mesophilous and thermophilous

cembran pine forests existed (Fig. 3, cluster 1.1.2).
(2) The OGUs of the cryophilous dwarf heaths (G3) were grouped with the OGUs
of the mesophilous cembran pine forests located on the north side (Fig. 3,
cluster 1.2) rather than with other OGUs of dwarf heaths because they did not
contain swards with Nardus stricta.
(3) The OGUs of the cold variant of the thermophilous heaths (Fb) formed a
sub-cluster (Fig. 3, cluster 2.1.1.1) within the mesothermophilous heaths (G1)
(Fig. 3, cluster 2.1) because they were located at higher elevation and thus also
contain mesophilous heaths.
(4) The unique OGU of cliffs (A) was classified with the warm variant of the
thermophilous heaths (F) (Fig. 3, cluster 2.2.2) because the OGU of the cliffs
contained thermophilous heaths.
(5) The three OGUs of the boulders (B) did not form a cluster because their
delimitation was not restricted solely to boulders.
(6) The sub-group G2 was not recognised as a group because the
Calamagrostietum villosae s.l. is a complex unit and because several other
kinds of heaths also occur in the five OGUs assigned manually to this land-
scape. Three out of five of these OGUs were part of a heterogeneous cluster
(Fig. 3, cluster 2.2.1) containing OGUs of the Riederhorn, mainly with
mesophilous heaths and some boulders.

3.2 Changes Between 1982 and 2014

3.2.1 Cluster Analysis

Ward’s cluster analysis on the relevés from both 1982 and 2014 (Fig. 4) showed the
same clustering structure as the cluster analysis of the relevés in 1982. In a first step,
the forested OGUs (cluster 1) were separated from the non-forested OGUs (cluster
2). Instead of having a geographical subdivision in cluster 1, however, the OGUs
with spruce forest (D) now formed a separate cluster (1.1), and all the OGUs with
cembran pine forests (C, E) were in the same cluster (1.2.2), except the cryophilous
heaths (G3) (cluster 1.2.1). In the non-forested cluster 2, the subdivision had not
changed, with a cluster (2.1) for the more or less pastured OGUs and another cluster
(2.2) for the non-pastured OGUs with cliffs (A), boulders (B), thermophilous heaths
(F) and mesophilous heaths with Calamagrostis villosa (G2).
There was a change between clusters 1 and 2 for two OGUs (D.188, G1.269),
and in one OGU (F.238) between clusters 2.1 and 2.2. Changes between the higher
sub-clusters occurred for OGU G3.264 in cluster 1.2, and OGU G1.236 in cluster
2.2.1. At a lower level, a change occurred for 17 OGUs, of which seven with
thermophilous heaths (F) were placed in clusters 2.1.2 and 2.2.2 and three with
boulders (B) placed in cluster 2.2.
Symphytosociology, a Tool for Landscape Monitoring … 277

Fig. 4 Ward’s cluster analysis (Ward D) of the relevés from both surveys (1982, 2014). A–G:
former manual classification (see text). Cluster 1: forested operational geographic units (OGUs).
Cluster 2: non-forested OGUs

On the other hand, when taking only the relevés from 2014 (Fig. 5), there was a
definite change in the main clustering, with a first division between the more or less
pastured OGUs (cluster 1) and the non-pastured OGUs (cluster 2), instead of
between the forested and non-forested OGUs. This behaviour shows a convergence
between the forested OGUs and the non-forested OGUs that are no longer pastured.
In addition, the non-pastured OGUs of cluster 2 were not subdivided into forested
and non-forested OGUs. There was one cluster (2.1) with the OGUs of the
Aletschwald Reserve, plus two adjacent OGUs with a recolonisation by mesophi-
lous cembran pine forest; and another cluster (2.2) with a sub-cluster 2.2.1 that was
about the same as cluster 2.2 in Fig. 3, for the relevés in 1982. Nonetheless,
sub-cluster 2.2.1 was without most of the OGUs of the thermophilous dwarf heaths
(F) (Fig. 3, cluster 2.2.2.1) and included, instead, most of the OGUs of the ther-
mophilous cembran pine forest (C). The other sub-cluster 2.2.2 contained mainly
the OGUs of the spruce forests (D) and the remaining OGUs of the mesophilous
cembran pine forests (E).

3.2.2 Canonical Correspondence Analysis (CCA)

The five classes of vegetation structure (trees > 2.5 m, shrubs 0.5–2.5 m, dwarf
shrubs < 2.5 m, closed herbaceous vegetation, and open herbaceous vegetation)
were seen as driving factors for change in the OGUs between 1982 and 2014
(Fig. 6). The modification in each category between 1982 and 2014 was statistically
significant. Changes in the cover of trees, dwarf shrubs and herbs were the main
factors of change. Globally, when looking at the position of the group centroids, a
centripetal convergence trend of the relevés appeared in 2014, relative to the relevés
in 1982, except for the boulders (B) and the mesophilous heaths with
278 J.-P. Theurillat et al.

Fig. 5 Ward’s cluster analysis (Ward D) of the 2014 relevés. A–G: former manual classification
(see text). Cluster 1: operational geographic units (OGUs) with more or less pasturing. Cluster 2:
OGUs with no pasturing

Calamagrostis villosa (G2). The convergence was also confirmed by cluster

analysis of the relevés in 2014 at the highest level (Fig. 5).
There was a marked increase in the tree cover of the non-forested OGUs (A, B,
F, G1, G2), as exemplified by the change in the average centroid. A decrease
occurred in the cover of the dwarf shrubs in the spruce forests (D), the mesophilous
cembran pine forests (E), the warm variant of the thermophilous dwarf heaths
(F) and the mesophilous heaths with Calamagrostis villosa (G2). The cover of the
herbaceous vegetation increased in the spruce forests (D) and mesophilous cembran
pine forests (E) but decreased in the warm variant of the mesothermophilous dwarf
heaths (G1). The cryophilous dwarf heaths (G3) appeared stable, besides an
increase in the herbaceous vegetation, in one OGU.
The convergence was substantiated by comparing the CCA performed separately
on the relevés from 1982 and those from 2014 (not shown). For the 1982 relevés,
the four main subdivisions resulting from the cluster analysis (Fig. 3) could be
individualised on the CCA, with forested and cryophilous OGUs on one side
(clusters 1.1 and 1.2; C, D, E, G3) and more or less pastured OGUs on the other
side (clusters 2.1 and 2.2; A, B, F, G1, G2). For the 2014 relevés, only the two
principal subdivisions, pastured OGUs (cluster 1; F, G1) versus non-pastured
OGUs (cluster 2; A, B, C, D, G2, G3; Fig. 5), could be identified.
Symphytosociology, a Tool for Landscape Monitoring … 279

Fig. 6 Canonical correspondence analysis (CCA) of the relevés from both surveys (1982, 2014).
The driving factors are the five classes of vegetation structure (trees > 2.5 m, shrubs 0.5–2.5 m,
dwarf shrubs < 2.5 m, closed herbaceous vegetation, and open herbaceous vegetation)

3.3 Landscape Change

Comparing the change of land cover between 1982 and 2014, one can see that 34
OGUs (56%) showed a cover change greater than 10%, either in tree cover (>
2.5 m), dwarf shrubs (< 0.5 m) or herbaceous vegetation s.l. Proportionally, the
forest-dominated OGUs were the more stable, with 9 out of 19 (47%) showing no
change between 1982 and 2014 (Fig. 4). The modification in tree cover corre-
sponded mainly to an increase in patches of spruce and spruce forests at the lowest
elevations, in five OGUs on the southeast side; as well as an increase in meso-
philous cembran pine patches and forests in four OGUs on the northwest side,
outside the nature reserve. The change in dwarf-shrub cover did not show any clear
pattern for any given vegetation units. However, at the lowest elevation, there was
mainly a decrease in thermophilous heaths and an increase in mesophilous heaths.
At the highest elevation, there was mainly a decrease in mesothermophilous heaths
and an increase in mesophilous heaths. There was no pattern in the changes of the
herbaceous vegetation. Most of the modifications occurred with the Calamagrostis
villosa meadows and the Nardus stricta swards, with either an increase or a
decrease of their cover.
Cembran pine dominated the regeneration of young trees (0.5–2.5 m) outside the
forested area. A regeneration density of more than 10 individuals/ ha was observed
in 43 OGUs (70%), of which 20 OGUs had a density between 10 and 29 (33%), 17
between 30 and 59 (28%), and 7 OGUs with more than 60 individuals, reaching up
280 J.-P. Theurillat et al.

to 154 individuals (data not shown), respectively. Regeneration occurred at all

elevations, with 19 OGUs located at mean elevation below 2100 m. By compar-
ison, a density of more than 10 individuals/ ha of spruce occurred in only 18 OGUs
(29.5%); two thirds of these were located at mean elevation higher than 2100 m.
Young spruce reached a maximum of 45 individuals/ ha.

3.4 Thermophilisation

Among the differences in the vegetation units, an increase in some thermophilous

units was noticeable, particularly for the OGUs at the lowest elevation. These units
corresponded mainly (1) to a new presence or an increase of shrubs from warmer
conditions (plant communities with Juniperus sabina, shrubby J. communis
ssp. alpina (> 0.5 m), J. communis var. intermedia, Rosa spp., Amelanchier ovalis,
Cotoneaster integerrima and shrubby Populus tremula); (2) to increase of ther-
mophilous swards (Festucetum variae s.l., Carici sempervirentis-Brachypodietum
rupestris); and (3) to increase in the thermophilous dwarf heath
(Junipero-Arctostaphyletum uvae-ursi s.l.). A sign of thermophilisation also
appeared in a few cases, with a distinct decrease or even the disappearance of some
cryophilous units, mostly small patches of snowbed vegetation (Alchemillo
pentaphyllae-Salicetum herbaceae s.l.). Evident signs of thermophilisation (posi-
tive and negative) emerged in 18 OGUs, which were also visible, but less impor-
tantly, in 27 other OGUs. In total, 74% of the OGUs displayed some sign of
thermophilisation. No thermophilisation was apparent in most of the OGUs con-
taining the mesophilous cembran pine forests, except two OGUs on steep, rocky
slopes.

3.5 Impact of the Deer

The presence of deer was evident in all OGUs, with frequent incidence of drop-
pings, primarily in 26 OGUs with paths. Strong impact by deer on the vegetation
existed in two zones. One was the Aletschwald Nature Reserve on the north side,
with the OGUs of mesophilous cembran pine (Table 1) and where human activity is
limited to restricted hiking; the other location was on the southeast side, in the
OGUs of spruce forests where there is hardly any human activity. The main impact
was from trampling, affecting the dwarf-shrub vegetation, mostly Rhododendron
ferrugineum and Juniperus communis ssp. alpina communities. The result was a
reduction in the mesophilous and mesothermophilous heaths (G1), and the occur-
rence of more or less highly degraded patches of these vegetations.
In the nature reserve, especially in the OGUs with mesophilous cembran pine
forests (Table 1), the degradation and destruction of Rhododendron ferrugineum
and Juniperus communis ssp. alpina dwarf shrubs led to local increase in the clonal
Symphytosociology, a Tool for Landscape Monitoring … 281

grass Calamagrostis villosa, both as a substituting meadow and as the understory of

mesophilous cembran pine forest. In some resting places, even ruderal vegetation
with Urtica dioica had developed. There was also trampled vegetation, with Poa
supina along the paths. The browsing had strongly decreased or, in some OGUs,
even eliminated the small patches of green alder, Swiss willow and tall herbs, with
Adenostyles alliariae and Peucedanum osthruthium. In the spruce forests, trampling
and browsing also led to the degradation and destruction of Rhododendron fer-
rugineum and Juniperus communis ssp. alpina in the openings and an increase in
herbaceous vegetation, either with Festuca rubra in the mesophilous parts or with
Nardus stricta swards. In contrast, on the southeast part of the study area, the
durability of some extensive swards with Nardus stricta, despite cessation of
pasturing, could have resulted from browsing by deer.

4 Discussion

Comparison of the symphytosociological relevés made in 2014 with those made in

1982 verified that the landscape around the tree line remained globally the same at
the level of the plant communities. In the conservative cluster analysis (Fig. 4), only
five OGUs (8%) changed their classification at the level of the higher clusters, and
17 (28%) at the level of the lower clusters (Fig. 4). However, when comparing
photos of the southeast side taken in the 1980s versus those of today, one can see
distinctly that the tree cover has increased, notably for spruce at the lowest ele-
vations. The importance of the structural transformation of the landscape was
revealed by comparing the cover of the five main categories of vegetation, with
56% of the OGUs showing a change greater than 10%. Since vegetation in the
non-forested units was changed more, we are inclined to think that the main driver
of the observed differences is the abandonment of the traditional pasturing that had
shaped the landscape near tree line for centuries. The noticeable decrease of ther-
mophilous heaths and increase in mesophilous heaths at lowest elevation, albeit
counterintuitive under climatic warming of 1.5 °C, are consistent with a decrease in
pasturing pressure, increased shade produced by increasing spruce, and the cessa-
tion of the manual Rhododendron ferrugineum and Juniperus nana dwarf shrubs,
which can now re-grow on the southeast part of the study area. The same reasoning
applies to the decrease in mesothermophilous and increase in mesophilous heaths in
the upper OGUs in the eastern part. Therefore, pasturing was a strong structuring
element of the landscape, and maintaining it would be important for the stabilisation
of the landscape in the face of climate change [7]. In the absence of pasturing, it is
known that deer can substitute for cattle to maintain some kinds of swards, with
browsing by 100 deer being about equivalent to that of 30 cattle.
On the other hand, the importance and rapidity of the modifications observed
near tree line cannot be explained only by the abandonment of pasturing.
A simultaneous acceleration factor is also needed. The increase in the mean annual
temperature by 1.5 °C since the beginning of the 1980s is probably the explanation,
282 J.-P. Theurillat et al.

especially for the development of the spruce forests in the lowest OGUs. Such a
temperature increase corresponds to a theoretical rise by the tree line of 270 m.
Therefore, the non-forested OGUs around 2000 m are now distant from a climax
equilibrium that would correspond to the spruce forest, namely, the speed at which
spruce forest can now develop. A warmer vegetation period is also reflected by
signs of thermophilisation in 74% of the OGUs, with increase or emergence of
shrubs and swards at relatively lower elevation in warmer conditions, and disap-
pearance of cryophilous vegetation in the fragmentary snowbeds in the OGUs of the
cryophilous heaths (G3) that occurred on the northern side, in the upper part of the
nature reserve.
The nature reserve, in which no human activity has occurred since its creation in
1933, would have been an opportunity to observe whether the climax vegetation of
the mesophilous cembran pine forests (E) would show some resistance to the
temperature increase of 1.5 °C. However, the strong presence of deer degraded the
dwarf heaths by trampling; caused the disappearance of some vegetation types,
such as green alder, by browsing; and caused ruderalisation along paths and in
resting places. Deer presence also favoured the expansion of the grass
Calamagrostis villosa, both by trampling and browsing (Table 1). Although there
was no sign of thermophilisation in the corresponding OGUs, which could be
considered a sign of stability against the current warming, the long-term impact of
deer could be a local accelerator of a deeper change in these OGUs. Ironically, in
the absence of top predators, like wolf, the large, concentrated deer population,
hardly regulated by selective shooting, could drive future vegetation of the nature
reserve back to a state resembling the situation that existed 100 years ago, when the
emblematic mesophilous cembran pine forest was pastured by cattle, sheep and
goat, and submitted to wood-cutting [8].

5 Conclusion

Symphytosociological relevés are an effective and potent tool for detailed landscape
monitoring, especially with the existing facilities offered by geo-referencing that
allow selecting precise standard plots in a given area. Comparison of symphy-
tosociological relevés is more powerful than comparison of vegetation maps, even
at very fine scale, which cannot reach the level of detail afforded by symphy-
tosociological relevés because of scale limits.
Although the plant communities have not been modified profoundly at the
landscape level during the last 30 years, the symphytosociological resurvey in 2014
revealed a steady trend towards increases of dwarf shrubs and tree cover. Therefore,
there is a need for some management, in the face of the ongoing climate change and
increasing deer population, to maintain the attractiveness of a traditional landscape
shaped by millennia of human activity in places like the Aletsch World Heritage
site.
Symphytosociology, a Tool for Landscape Monitoring … 283

Acknowledgements This research has been supported by the Swiss National Science Foundation
within the Man and Biosphere project (FNRS 4.331.2.79.55 to Claude Béguin) and the Service des
forêts, des cours d’eau et du paysage of the canton Valais (support 2012 to Jean-Paul Theurillat).

References

1. Aeschimann D, Lauber K, Moser DM, Theurillat JP (2004) Flora alpina. Paris, Berlin
2. Béguin C, Hegg O (1975) Quelques associations d’associations (sigmassociations) sur les
anticlinaux jurassiens recouverts d’une végétation naturelle potentielle (essai d’analyse
scientifique du paysage). Doc Phytosociol 9–14:9–18
3. Béguin C, Hegg O, Camenisch M, Theurillat JP (2016) Jean-Marie Géhu et ses liens avec la
Suisse: la symphytosociologie s.l. et le sauvetage des documents de la Station Internationale
de Géobotanique Méditerranéenne et Alpine. Doc Phytosociol (Numéro Spécial) 8:109–119
4. Béguin C, Theurillat JP (1984) Landschaftsökologische Studie in der Region Aletsch (MAB
6) nach einer modifizierten symphytosoziologischen Methode. Verh Ges Ökol 12:149–157
5. Chalumeau A, Bioret F (2013) Méthodologie de cartographie phytosociologique en Europe:
approches symphytosociologique et géosymphytosociologique. Synthèse bibliographique.
Institut de Géoarchitecture, Université de Bretagne Occidentale, Brest
6. Géhu JM (1977) Le concept de sigmassociation et son application à l’étude du paysage
végétal des falaises atlantiques françaises. Vegetatio 34:117–125
7. Henne PD, Bigalke M, Büntgen U, Colombaroli D, Conedera M, Feller U, Frank D, Fuhrer J,
Grosjean M, Heiri O, Luterbacher J, Mestrot A, Rigling A, Rössler O, Rohr C, Rutishauser T,
Schwikowski M, Stampfli A, Szidat S, Theurillat JP, Weingartner R, Wilcke W, Tinner W
(2018) An empirical perspective for understanding climate change impacts in Switzerland.
Reg Environ Change 18:205–221
8. Hess E (1933) La Forêt d’Aletsch, monument naturel. Bull Murith Soc Valais Sci Nat 51:78–
111
9. Rivas-Martínez S (1976) Sinfitosociologia, una nueva metodologia para el estudio del vegetal.
Anal Inst Bot Cavanilles 33:179–188
10. Schwabe A (1991) Perspectives of vegetation complex research and bibliographic review of
vegetation complexes in vegetation science and landscape ecology. Excerpta Bot Sect B
Sociol 28:223–243
11. Schwabe A (1997) Sigmachorology as a subject of phytosociological research: a review.
Phytocoenologia 27:463–507
12. Theurillat JP (1992) Etude et cartographie du paysage végétal (symphytocoenologie) dans la
région d’Aletsch (Valais, Suisse). Développement historique et conceptuel de la
symphytocénologie, niveaux de perception, méthodologie, applications. Beitr Geobot
Landesaufn Schweiz 68:1–384 & annexe
13. Theurillat JP (1992) Abgrenzungen von Vegetationskomplexen bei komplizierten
Reliefverhältnissen, gezeigt an Beispielen aus dem Aletschgebiet (Wallis, Schweiz). Ber
Reinhold-Tüxen-Ges 4:147–166
14. Tüxen R (1973) Vorschlag zur Aufnahme von Gesellschaftskomplexen in potentiell
natürlichen Vegetationsgebieten. Acta Bot Acad Sci Hungar 19:379–384
Ecology
Why Does Mediterranean Vegetation
Seem So Diverse?

Elgene O. Box

Abstract Worldwide patterns of biodiversity have been catalogued and mapped at

least since the 1990s, based partly on studies from much earlier. For plant biodi-
versity, construed most easily as species richness, theories on controlling factors
have also been summarized and analyzed by numerous authors, with particular
emphasis on relationships between plant species richness and ecosystem function in
general, especially primary productivity. A less studied aspect of plant diversity and
vegetation structure involves the basic ecological plant types and their morpho-
logical characters that make up the vegetation – and may also be related to function.
In the absence of complete data worldwide, models are required for global cov-
erage. The purpose of this study is to use an existing model of the climatic limits of
major plant growth forms to predict potential geographic patterns of richness in
basic plant forms and form characters, in order to understand why
Mediterranean-type regions, and southern Italy in particular, appear so rich. Highest
levels of total richness in potential plant forms and pheno-physiognomic characters
are predicted for tropical and some subtropical regions, but richness in coexisting
main structural elements is suggested to be high in the more open vegetation of
regions with Mediterranean-type climates, perhaps the Mediterranean region
proper. With global warming some of this structural richness may be lost.

Keywords Climatic envelopes Form richness of vegetation Main structural



complex Mediterranean-type regions Physical vegetation structure Plant form





characters Plant life forms PFORMS model Southern Italy

1 Introduction

It has been observed by many that biodiversity, in many taxonomic groups, including
vascular plants, increases from high latitudes toward the tropics – albeit not mono-
tonically on continental west sides (e.g. [19, 34, 37, 48, 55, 60, 66, 67, 75]).

E. O. Box (&)
Department of Geography, University of Georgia, Athens, GA 30602, USA

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 287
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_15
288 E. O. Box

Worldwide statistics on biodiversity have been catalogued, among others, by

Groombridge [31], Mittermeier et al. [49], and Olson and Dinerstein [52], and pat-
terns of plant species richness have been mapped by Barthlott et al. [4]. Theories on
the factors that control plant species richness have also been suggested by many
authors (e.g. [2, 32, 38, 42, 43, 77], cf [24]) and have been summarized, among others,
by Shmida and Wilson [62], Diamond [22], Palmer [53], Pausas and Austin [54], and
Evans et al. [25]. Relationships between plant species richness and ecosystem
function have been explored by, among others, Brophy et al. [18], Hooper et al. [35]
and Midgley [46], with an especially intense discussion regarding the relationship
between plant species richness and primary productivity (e.g. [47, 73, 74]). A novel
and perhaps quite useful idea that plant species richness may be controlled by light
rather than productivity or other factors was provided by Gilbert [28].
Tropical regions appear to have the greatest diversity in species and perhaps
other biological aspects (e.g. taxonomic groups, trophic webs, ecological niches).
Transitional areas may also be quite diverse, however, to the extent that they
contain elements from both adjacent regions. Regions with Mediterranean-type
climates are transitional between the temperate and tropical zones, and their floras
contain elements of both (e.g. [5, 33, 56]). In North America, a third component, the
so-called Madro-Tertiary geoelement, arose in the developing region of
Mediterranean-type climate (s.l.) between the Arcto-Tertiary and the
Neotropical-Tertiary geofloras, and has been ascribed to mid-continental drying [3].
Hotspots of diversity in the mediterranean regions have been noted by Médail and
Quézel [44, 45] and Skeels and Cardillo [63] (cf [21, 59]). The tropical region may
also be richest in different basic plant types. An early world map of the potential
co-existence of different plant growth forms as major structural elements in vege-
tation stands, however, suggested high numbers in regions with Mediterranean-type
and other warm, subhumid climates [7].
In mediterranean climates, the usual response of vegetation along gradients from
wetter to drier conditions starts, at the wetter end, with low sclerophyll forests
(Quercus species around the Northern Hemisphere), with moderate-sized leaves
and deep root systems that permit the trees to remain active during the dry summer
(e.g. [71]). Intermediate levels of precipitation involve broad areas of sclerophyll
scrub (maquis, macchia, matorral, chaparral, kwongan, etc.), also with deep roots
but with smaller, hard but still evergreen leaves. But this response to increasing
summer dryness may be discontinuous toward the dry end, with the increasingly
smaller, harder sclerophylls replaced suddenly, at least partly, by plants with larger,
softer but deciduous leaves, where soil water is insufficient for continued plant
activity in summer. These more drought-adapted plants are largely dwarf shrubs,
with shallow roots and soft summer-deciduous leaves that permit only a short
springtime growing season. These different responses to varying degrees of summer
dryness lead to the coexistence of great variety in plant structures, foliage types and
seasonality patterns in relatively open vegetation stands, in which dominance
through greater shade tolerance becomes less relevant.
Why Does Mediterranean Vegetation Seem So Diverse? 289

Major reasons for the historical importance of the Mediterranean region include
not only its crossroads location between Europe, Africa and Asia, but also the
diversity of its landscapes and their resources. Already the ancient Greeks saw
vegetation as the basis for recognizing and delimiting different regions ([70], cf
[69]). Diversity, in this sense, means not only species richness but also the variety
of different landscapes, plant types and their products. Anyone can appreciate the
diversity of sizes, shapes and seasonal patterns of plant in Mediterranean cultural
landscapes. But a closer look also reveals a high apparent degree of structural
complexity, i.e. diversity, in natural vegetation, based on the large number of plant
forms coexisting in the mostly open vegetation not dominated by closed forest
overstoreys. Diversity in plants can be viewed in many ways: taxa, morphological
attributes, functional “traits”, more generalized ecological plant types, etc. Global
coverage and scrutiny may only be possible for those aspects that can be gener-
alized for modeling, which means mainly generalized basic plant types, their
structural attributes, and the overall structure (complexity) of the vegetation stands
that they build. Such basic plant types can be related to whole, recognizable
organisms (as opposed, for example, to functional diversity) and are perhaps the
only aspects that can be modeled adequately to provide global [terrestrial] coverage.
As a result, only richness in plant forms and characters will be treated in this study.
‘Diversity’, as used herein, will refer to the main physical structure of vegetation
stands, as represented by the number and mix of the different larger plant forms that
build the vegetation structure, i.e. the ‘main structural elements’. The Russian
literature called these main structural contributors ‘edificators’ (Эдификaтopы, e.g.
[1]), and the word has been used occasionally in Western ecology. More com-
monly, these have been called the ‘main structural elements’ (e.g. [27]; [17] p. 112;
[51] p. 30; [14] p. xxv; [39] p. 613). Open vegetation stands appear more diverse
because more different shapes and sizes of plants can be seen. Consider, for
example, what a bird seeking shelter or a temporary perch might see when
approaching from the outside.
High diversity of coexisting plant types in Mediterranean vegetation was first
suggested by a world map of form diversity predicted from a global-scale climatic
envelope model of plant types [7]. The high diversity in Mediterranean areas was
emphasized again by a recent assessment of potential diversity in plant form
characters, based on an updated version of the world envelope model and a larger
number of sites [16]. Both studies found high form diversity in areas that are
subhumid to only moderately humid, generally warm (i.e. with little or no frost),
somewhat seasonal, and have wet and dry seasons. Such conditions often produce
more open vegetation, as in deserts and grasslands. In these situations, and in polar
climates, the vegetation is usually dominated by only a few plant forms, although
some deserts (e.g. the Sonoran) but may appear surprisingly form-rich (but still not
have a large number of forms in total). This leaves scrub vegetation, perhaps
especially that of the drier Mediterranean-type climates, as potentially the most
form-diverse. But why does the Mediterranean region proper seem more diverse
than other regions of Mediterranean-type climate?
290 E. O. Box

The purpose of this study is to use an existing model (PFORMS) of the climatic
limits of major plant growth forms to predict potential geographic patterns of
richness in basic plant forms and form characters, in order to understand why
Mediterranean-type regions, and southern Italy in particular, appear so rich. In
keeping with the title of this book, there is a tool involved, namely a climate-based
general model of basic plant types and their environmental relationships.

2 Modeling Plant Forms and Characters

The idea of basic ecological plant types is not new and grew out of a need to
understand general ecological relationships without having to treat large numbers of
different species individually. Plant types defined entirely by structure, such as
deciduous broad-leaved trees or stem-succulents, had been called ‘growth forms’
(Wuchsformen) by Drude, based on the Hauptformen of Humboldt [36], the 54
types of Grisebach [30], and his own 55 postulated forms [23]. Structural types
interpreted as ecological adaptations to environmental conditions were called ‘life
forms’ (Lebensformen) by Warming [72] and could be interpreted as basic eco-
logical types, grouping taxa with similar ecological requirements and morpholog-
ical responses to similar environmental conditions. The first modern system of plant
forms was designed for temperate zones [58], but the approach was later extended
to the tropics (e.g. [41, 61]). A comprehensive, global classification of plant life
forms is given as Appendix A in Mueller-Dombois and Ellenberg [50]. The pos-
sibility of form change over the course of a lifespan was incorporated in the concept
‘living form’ (жизнaя фopмa) in the Russian literature (see summaries by [26, 68]).
The currently popular term ‘plant functional type’ (PFT) seems to have originated
just before the large IIASA workshop on ‘global’ vegetation change” (April 1988;
see preface in [65], cf [78]). Initially the concept of PFTs developed along quite
different pathways (see [12]), including the “function follows form” approach of
Box [7] but also approaches that made no reference to form whatsoever (see also
review by [64]). Most simply, PFTs have been defined as “functionally similar plant
types” ([12], cf [29]) or as “groups of species that use the same resources and
respond to the environment in a similar way” [54].
Since plant metabolism, transpiration and other processes operate through the
size, texture and seasonality of plant photosynthetic (and some other) surfaces,
basic plant types must be described by morphological (pheno-physiognomic)
attributes. The most important of these seem to be:
– Permanence – mainly woody vs non-woody, plus plant architecture (branching,
etc.)
– Size – relative to other plants
– Seasonality patterns – but not just evergreen vs deciduous
– Texture of the photosynthetic surface: “hard” or “soft” (gas exchange), shade
tolerance, etc.
Why Does Mediterranean Vegetation Seem So Diverse? 291

Consequently, the original model focused on these attributes to describe basic plant
types [7, 9].
The basic approach consists of three submodels [7, 16]. The first is a set of
structurally defined “basic plant types” (growth forms) defined by form characters.
The six characters used to represent the basic plant forms are structural type (e.g.
tree, shrub, forb), relative plant size, type of leaf or other photosynthetic organ (e.g.
phyllode, green woody stem), relative leaf size, seasonality habit, and consistency
of the photosynthetic surface (e.g. sclerophyllous, malacophylous, ligneous).
A sample of the plant types and their characters, as coded in the model, is given in
Table 1.
The second submodel is a set of climatic variables hypothesized to represent the
effective climatic environment (and for which data can be obtained readily
worldwide). These variables include monthly, seasonal-extreme, and annual tem-
perature and precipitation, plus an annual moisture index (precipitation divided by
potential evapotranspiration). Each site also includes absolute minimum tempera-
ture, either recorded (if over enough years to be reasonable) or estimated (often
better) based on the difference between mean monthly minimum and absolute
minimum at nearby sites where it has been recorded reliably (cf [16]).
The plant forms and climatic variables are linked by a third submodel, which
consists of a climatic “envelope” for each plant type, defined by apparent tolerance
limits for that plant type relative to each of the climatic variables. The limiting
values in these envelopes involve commonly accepted values where possible, such
as cardinal temperatures (e.g. [40], cf [76]), as well as empirical values obtained
from worldwide comparisons. Envelopes for a few plant types prominent in
mediterranean vegetation are shown in Table 2, which also defines the climatic
variables used. Other details of the model have been described elsewhere [7, 9, 16]
and require too much space to repeat here.
The model, now called PFORMS, has been updated and improved from time to
time, in particular by 1) by improving the estimator for potential evapotranspiration
(cf [8]); 2) by adding absolute minimum temperature as a limiting factor (cf [11]);
and 3) by recognizing shade tolerance as an important attribute of leaf structure (cf
[13, 17]). Each of these improved model predictions significantly and made the
results more reliable globally. The model has also been improved by adding more
plant forms and by making the whole set more geographically balanced, especially
by including major understorey forms in all regions. There are now 115 such plant
forms in the model (see list in Box and Fujiwara [17]). Finally, the climatic
envelopes have been re-calibrated repeatedly, based largely on field experience, for
better fit to the actual and potential geographic ranges of the various plant types.
A sample PFORMS result, for Palermo (Sicily), is shown in Table 3, along with
naturally occurring local species corresponding to the forms predicted.
292 E. O. Box

Table 1. Sample of plant forms and characters in the PFORMS model

Each plant type is a structurally defined growth form, defined by the six characters shown but also
interpreted tacitly as an adaptation to environmental conditions, i.e. each form is also a ‘life form’
sensu Drude and others (see main text). For example, Mediterranean Shrubs are shrubs (multiple
main stems) of “normal” size, with broad but small leaves which are evergreen but drought
adapted sclerophylls, e.g. Rosmarinus officinalis or Phillyrea spp. Among the 115 plant types
included, almost all combinations of the six form characters (most important in boldface) are
unique. Abbreviations:
Struc (structural type): T = tree, Tp = pole tree, RT = tuft tree, S = shrub, SS = stem-succulent, G =
graminoid, F = forb, Th = cryptogam (not shown: arborescents, semi-shrubs, ferns, vines,
epiphytes)
Size (relative plant size): T = tall/large, N = normal (medium), S = small, D = dwarf
Ltype (green organ): B = broad leaf, N = needle, F = frond, L = linear leaf, G = graminoid leaf (not
shown: green woody stem)
Lsize (relative leaf size): L = large, N = normal (medium), S = small, V = very small
Surf (green surface): S = sclerophyllous, L = laurophyllous (shade tolerant), C = coriaceous, M =
malacophyllous (Mr = reinforced), P = pubescent (mainly soft), F = fleshy (succulent) (not shown:
ligneous)
Habit (seasonality): EG = evergreen, SE = semi-evergreen, RG = raingreen, SG = summergreen
(not shown: ephemeral)
Why Does Mediterranean Vegetation Seem So Diverse? 293

Table 2. Climatic envelopes for selected Mediterranean-climate plant forms, with climatic
variables used and their limiting values

These sample envelopes are from the PFORMS model used herein. A climatic envelope consists of
a set of upper and lower limiting values of the [climatic] envelope variables, of which there are
eight:
BT = annual biotemperature (sum of above-freezing monthly means, divided by 12 (°C)
Tmax = mean temperature of the warmest month (°C)
Tmin = mean temperature of the coldest month (°C)
Tabmin = absolute minimum temperature (°C)
MIy = annual moisture index (precipitation/potential evapotranspiration)
Pmax = average precipitation of the wettest month (mm)
Pmin = average precipitation of the driest month (mm)
PmTmx = average precipitation of the warmest month (mm)
Asterisks (****) indicate unspecified (open-ended) and presumably unimportant limits.
Occurrence of a plant form is hypothesized to be possible only between the limiting values for
all envelope variables. See also Box [8-10, 16, 17].

3 Methodology

Estimation and mapping of plant form and character diversity, in

Mediterranean-type regions and worldwide, followed essentially the same steps as
in the global study ([16]: if commercially blocked, please write to the author). This
methodology consisted of five steps:
1. Develop an appropriate, large, geographically balanced world data-base of site
climatic values
2. Predict the potential plant-form occurrences, using the climatic-envelope model,
at each climatic site
3. Place the predicted plant forms at each site into the standard layers recognized in
vegetation stands
294 E. O. Box

Table 3. Predicted PFORMS model results for Palermo (Sicily)

The plant forms predicted for Palermo (northwestern Sicily) by the PFORMS climatic-envelope
model are shown in the left-hand column, organized by traditional stand layers: T1 = canopy trees,
T2 = understorey tree layer, S1 = upper shrub layer, S2 = lower shrub layer, H = ground layer, M =
moss layer. Vines and epiphytes are placed in the T2 layer. Topo-specific forms not used in the
richness analysis are shown in brackets. The closest limiting factor in the envelope design is shown
in the second column (e.g. MIy for moisture index, see Table 2 for abbreviations), along with the
proximity to closest envelope limit (0–1 scale; u = upper value limiting) (see also Box [8, 9, 16]).
With predicted maximum woody cover of 62%, this result can be interpreted as open vegetation
dominated by Mediterranean-type shrubs and dwarf shrubs (sclerophyll and summergreen), with
scattered smaller trees and arborescents – plus exotics such as Eucalyptus and cacti. Note that, in
Mediterranean-type climates, summer-deciduous forms may be considered raingreen.
Actual plant taxa occurring in and near Palermo are shown at right, juxtaposed with the plant
forms that they represent. Non-native but perhaps well naturalized taxa are shown in parentheses;
probably not naturalized but widely planted ornamental taxa (including some taxa from the famous
local botanical garden) are shown in brackets. Question marks indicate missing information or, for
particular taxa, unknown occurrence or naturalization status. Unpredicted but widely occurring
taxa (mainly on special microsites) include Pinus spp., plus species that are hard to classify, such
as the tall, tussocky cane grass Ampelodesmus mauritanicus and the forb Eryngium campestre,
which though sclerophyllous could also be considered raingreen, since it does generally dry out
rather completely in late summer. Some of these unpredicted forms were rejected by only one
envelope variable; see discussion of these “near misses” later in the main text. Matches of plant
taxa to particular plant forms were made based on field experience (four visits to Palermo) and
field notes, supplemented by images and verbal descriptions from numerous websites, including
Wikipedia. Species names and verification of native or naturalized status were based largely on the
checklist of the flora of Sicily by Raimondo et al.[57] and supplemented by field notes on naturally
occurring species.
Palermo is representative of model results and potential accuracy at many if not most sites studied.
It is used here as an example because: 1) it is representative; 2) it is where the form-richness results
in the latter part of this paper were first presented; and 3) it illustrates a basic premise of this study,
namely that subhumid warm-temperate climates, such as the Mediterranean type in southern Italy,
have unusually high richness in co-existing plant types, co-occurring in open vegetation stands (cf
Figure 1, from Box 1981[8], page 100).
Why Does Mediterranean Vegetation Seem So Diverse? 295

4. Count and synthesize occurrences of forms and the individual form characters
5. Map the resulting patterns.
Each site in the data-base has long-term average (pre-1990) values of the monthly,
seasonal, and annual variables, plus absolute minimum temperature (see Table 2).
The envelope model PFORMS was run initially on 1866 such sites worldwide, but
more sites were added for mapping, deleting some coastal and island sites and
adding many more sites inland, for a total of 2889 sites.
Forms predicted by the model are normally displayed by synusiae but were
re-assigned (program PFdiv) to the usual standard layers recognized in vegetation
stands, i.e. the emergrent, canopy, subcanopy, shrub, low-shrub, herb (ground) and
moss-lichen layers, as shown in Table 3. Each plant form has a single character
value for relative plant size (e.g. large, medium, small), representing a maximum
potential that is not realized everywhere. Placement of forms into stand layers was
thus adjusted for the particular site, based on the climatic moisture balance as a
suggestion of actual stand height and strata. For example, ‘Tall Sclerophyll Trees’
(potentially in a super-tree layer) would only occur as shorter trees if a site is too
dry. As a result, one can distinguish closed versus open stands and understand stand
structure more easily.
The numbers of plant forms predicted to occur, and the numbers of different
values for the individual form characters (i.e. structural type, seasonality, etc.), were
then counted for each site and summed for world totals (also by program PFdiv).
The individual sites richest in plant forms and characters were also determined and
listed (by PFdiv). The site counts were mapped, over the world’s land areas, by a
distance-squared-decay contour algorithm (program PIXMAP) and colored (pro-
gram PIXPAINT) by assigning color codes from Adobe Photoshop, which did the
final mapping (see details in [16]). All programs except Photoshop were written in
FORTRAN by the author.
Finally, for the current study, the site and map results were analyzed regionally,
focusing especially on richness within the Mediterranean region, in comparison
with worldwide patterns and with other regions of Mediterranean-type climate, i.e.
California, central Chile, the Cape region of South Africa, and the west-facing
coastal areas of southwestern and southeastern Australia. In order to understand the
apparent diversity of Mediterranean vegetation, special attention is paid to the
groups of main plant forms (main structural elements) that provide the main
physical structure of vegetation stands. This ‘main structural complex’ consists
essentially of those forms that receive full sunlight and are not in a shady under-
storey. For closed forests this means only the canopy trees; for open forests, both
the canopy and the tree understorey (T2) layer are involved. For woodlands and
scrub, all woody forms are included, plus larger stem succulents, forbs and grasses.
In Table 3, for example, because the vegetation predicted for Palermo is somewhat
open (estimated cover 62%), the main structural complex includes the T1, T2, S1
and S2 layers, for a total of 23 forms out of the total of 35 predicted to occur. Vines,
epiphytes and cryptophytes (s.l.) are always excluded from this calculus.
296 E. O. Box

4 Results Worldwide

Results are presented as tallies of plant-form richness for particular sites and regions
(based on 1866 sites) and as world maps (based on 2889 sites). These are first
results, so there are some obvious errors, especially the bad spatial interpolation in
some parts of the maps, such as northern Scandinavia, southwestern Arabia, the
Mackenzie Valley of northwestern Canada, and Chukotka, where too few climate
sites were available.
The total number of plant forms predicted for sites worldwide, and interpolated
for mapping, is shown in Fig. 1 (from [16]). Total potential form richness appears
highest in much of non-Amazon Brazil and other humid warm-temperate to sub-
tropical parts of South America; southern Mexico and northern Central America;
southern China; northeastern Australia; small equatorial and larger
humid-subtropical parts of Africa; and northernmost New Zealand. These areas
involve mostly tropical regions with wet and dry seasons, smaller equatorial areas
(mainly in Africa), and more subtropical to warm-temperate humid areas. A bit
lower but still relatively high form richness is also predicted for areas with
Mediterranean-type climates, including Italy, submediterranean Galicia (north-
western Iberia), central Chile and southern Australia – but not California. High
potential plant-form richness is also predicted in southern Britain, but this was
precluded by the Pleistocene glaciations.
These worldwide results are summarized by region and richness levels in
Table 4. Overall, the two areas potentially richest in plant forms seem to be
non-Amazon tropical and subtropical Brazil and subtropical southern China. Africa

Fig. 1. Total Number of Plant Forms Predicted by the PFORMS Model.

The most forms predicted at any site was 56-61 (see Table 5).
Legend: Dark brown = 1-4; Light brown = 5-7; Dark red = 8-9; Red = 10-11;
Light orange = 12-14; Yellow = 15-17; Light green = 18-19; Medium green = 20-24;
Dark green = 25-29; Blue = 30-39; Purple = > = 40; Black = no forms predicted
Why Does Mediterranean Vegetation Seem So Diverse? 297

Table 4. Predicted plant-form richness in world regions (excluding Mediterranean)

Shown by region are the numbers of modeling sites (out of 1866) at which the total number of
plant forms predicted fell into the respective deciles. For example, Brazil had 2 sites with 60 or
more forms predicted, 6 with 50–59 forms, 14 with 40–49 forms, and 24 with 30–39 forms
predicted. The regions richest in plant-form potential seem to be subtropical Brazil and subtropical
China.
298 E. O. Box

as a whole is also relatively rich in potential forms, with Australia and especially
tropical Asia and Meso-America further down the list. The everwet equatorial
regions of South America and the East Indies are predicted to be less form-rich
because both light-demanding forms and forms adapted to shorter growing seasons
are largely precluded in the shady environments of tropical rainforests, which are
strongly dominated by evergreen shade-tolerant trees and understorey forms.
Remaining areas of Europe, most of the USA, and much of temperate East Asia are
predicted to be still less form-rich, dominated largely by summergreen forms. Form
richness is predicted to be lowest in the driest and coldest areas but also low in
some cool-temperate humid areas (e.g. Canada, Manchuria).
Predicted form richness in different stand layers was treated in Box [16], with
overstorey form richness predicted to be highest in warm subhumid to quite dry
areas with no forest canopy. Overstorey form richness was also predicted to be high
in (sub)tropical wet-dry areas but lower in areas of closed forest, especially tropical
rainforests and Northern Hemisphere forests, where summergreen deciduous trees
or boreal conifers dominate strongly. Shrub-layer form richness was generally
predicted to be higher, throughout both the everwet and wet-dry tropics, and in both
humid and summer-dry warm-temperate areas. Less diverse shrub layers are sug-
gested for quite dry areas, and no shrub layer at all was predicted [correctly] for the
vast permafrost areas of Siberia, where Larix pole trees have only lichen-moss
ground layers between, with occasional dwarf shrubs. Species richness is generally
highest in herb layers, but forest areas herein were generally predicted to have
relatively few forms, with significantly more only in the tropics and subtropical
southern China, and in relatively steep transitions to arid areas. Large areas of more
or less uniform understorey form richness, at 8–12 forms, were predicted over much
of temperate to subboreal North America, eastern Europe and northern China.
Similarly large areas of 15–20 predicted plant understorey forms are predicted for
tropical wet-dry and warm-temperate to subtropical humid areas. The areas of
greatest understorey form richness are predicted to occur in humid subtropical to
warm-temperate areas of southern China, and to a lesser extent in southeastern
Africa and eastern to southeastern Brazil.
Geographic patterns are probably more important than site results, since indi-
vidual sites may involve artifacts such as predicted occurrences just outside or
inside envelope boundaries. Predicted form richness at the richest sites worldwide is
compared with that of the world’s regions of Mediterranean-type climate in Table 5.
Three sites, two in Brazil and one in Zimbabwe, had 60 or 61 plant forms predicted
(out of 115). Southern China, however, had the most sites with at least 50 forms
predicted. Almost all these locations are tropical or subtropical, but not equatorial.
Indeed the map (Fig. 1) also suggests a general reduction of form richness in
equatorial areas, where the lack of a dry season precludes opportunities for plant
forms adapted to shorter growing seasons. On the other hand, the predicted
form-rich sites in Table 5 all occur in climates with driest-month precipitation
below 30 mm (except one site in southern Brazil). The actual climate types
involved are mostly tropical wet-dry in Brazil, Mesoamerica and southern Africa;
and humid subtropical to warm-temperate, with colder winters, in southern China
Why Does Mediterranean Vegetation Seem So Diverse? 299

Table 5. Predicted plant-form richness at the richest sites versus richer sites in Mediterranean
climates

Predicted plant-form richness is shown for the richest sites in the world data-base, ordered by
decreasing richness; and (bottom) for richer representative sites in regions with
Mediterranean-type climate (ordered by latitude). Boldface type indicates relatively high or low
values in each column. None of the richest sites has a Mediterranean-type climate (see last
column), although Jonkershoek is perhaps submediterranean. Among the Mediterranean-type sites,
less summer-dry sites are often richer in forms, including some submediterranean sites.
Abbreviations: Lat = latitude; Elev = elevation (m); Tmax, Tmin = mean temperature of
warmest/coldest month respectively; Py = annual precipitation, Pmin = precipitation of driest
month; MIy = annual Py/potential evapotranspiration, Forms = number of plant forms predicted by
the PFORMS model

(plus Hlabisa in southeastern South Africa). Only one location (Pietermaritzburg)

has annual precipitation below 1000 mm, which is less than annual PET. The only
Mediterranean-type sites in the list are submediterranean Jonkershoek (55 forms,
Western Cape, South Africa) and truly Mediterranean Messina (53 forms, on the
strait between peninsular Italy and Sicily).
What do these potentially richest areas have in common? Initially, it seems that
they all:
– are relatively warm but with a cooler season, i.e. tropical or subtropical but not
equatorial;
– have mild winter (coolest month < 19 °C, permitting occurrence of some
temperate forms);
– are in generally moist climates;
– have seasonal precipitation, but with at least 10 mm (on average) even in the
driest month; and.
– may be at a range of elevations (not > 1800 m).
300 E. O. Box

Overall, this suggests relatively long growing seasons, with evergreen or seasonally
green closed-canopy forests (sometimes with many epiphytes). Elevation (and
higher latitude) result in cooler but still mild winters up to a point, and this may
permit some temperate forms to occur without excluding frost-sensitive tropical
forms. These very richest sites are not an exact match to the mapped geographic
patterns, which show wide areas of high predicted form richness in more wet-dry to
generally subhumid warm areas.
Predicted numbers of structural types, leaf types, seasonality types and surface
types among the plant forms predicted are shown as separate world maps in Fig. 2
(each with a different color scheme!). The most structural types (upper left of Fig. 2)
are predicted for both everwet and wet-dry tropical areas, as well as summer-dry
areas, before dropping off fairly abruptly in the transition to the temperate zone.
Again, California seems to be the outlier among the world’s mediterranean areas.
Fewest structural types are predicted in non-wooded arid and polar areas, but
numbers are also low in forest areas throughout the higher-latitude continental areas
of the Northern Hemisphere. The only broad areas of predicted intermediate rich-
ness in structural types are in areas of temperate deciduous forest.
The most types of leaves or other photosynthetic organs (upper right of Fig. 2)
are predicted in tropical wet-dry areas, areas of Mediterranean-type climate (except
California), and some more humid warm-temperate areas, including eastern
Australia, and southeastern Africa and South America. Again, erroneous high
richness is predicted in southern Britain. Richness in photosynthetic organs is
predicted to be low in Northern high latitudes and drier continental areas, and
lowest in arid and polar areas. Questionable results include the lower values in the
southeastern Amazon (yellow) and northwestern Australia (light green).
The potential number of different seasonal foliation patterns represented by the
plant forms predicted is shown in the lower left of Fig. 2. High-latitude Northern
areas involve mostly just two potential patterns: evergreenness among conifers and
tundra dwarf shrubs, and summergreen deciduousness among the broad-leaved
trees, larger tundra shrubs (e.g. Salix), and graminoids. The highest predicted
number of seasonality patterns, unlike the top-row panels, is in drier rather than
wetter tropical and subtropical areas, including areas of Mediterranean-type climate
(this time including California) but also more humid warm-temperate to subtropical
climates. Fewer seasonal habits are suggested in the equatorial and wetter wet-dry
areas, where most forms are evergreen or raingreen.
Different photosynthetic surface types (lower right of Fig. 2) include sclero-
phylls, shade-tolerant laurophylls and soft (deciduous) malacophylls, as well as the
“harder” surfaces of succulents, phyllodes and green woody stems. Fewer types are
predicted throughout almost all of the Northern Hemisphere, except for the
Mediterranean area and humid subtropical to tropical (wet-dry) southern Asia. The
only Southern Hemisphere areas predicted to have few surface types are in extreme
southern South America and the cool-subtropical Atacama and Namib desert areas,
plus some of extreme southern Australia. The most co-occurring surface types are
predicted for tropical wet-dry, Mediterranean-type, and humid warm-temperate to
subtropical areas. Fewer surface types are predicted in temperate grasslands but
Why Does Mediterranean Vegetation Seem So Diverse? 301

Fig. 2. Number of Plant Form Characters represented among the Plant Forms Predicted The four
panels represent:
a) plant structural types (upper left), such as trees, shrubs, stem-succulents, graminoids;
b) foliage types (upper right), including broad leaves, needles, and leafless (e.g. succulents);
c) seasonal foliation habits (lower right), including evergreen, summergreen, and ephemeral; and
d) types of the photosynthetic surfaces (lower right), such as sclerophyllous, coriaceous, soft
(malacophyllous), and woody (photosynthetic stems).
Structural Types (upper left): maximum of 13 at any site
Dark red = 12-13; Red = 11; Orange = 10; Light orange = 9; Yellow = 8; Light green = 7;
Medium green = 6;
Dark green = 5; Light brown = 4; Dark brown = 1-3; Black = no forms predicted;
Foliage Types (upper right): maximum of 6 at any site
Red = 6; Orange = 5; Yellow = 4; Green = 3; Dark green = 2;
Light brown = 1; Black = no forms predicted Intervening colors result from interpolation;
Seasonal Habits (lower left): maximum of 8 at any site
Dark red = 8; Red = 7; Orange = 6; Yellow = 5; Green = 4;
Dark green = 3; Light brown = 2; Dark brown = 1; Intervening colors result from interpolation
Black = no forms predicted;
Surface Types (lower right): maximum of 7 at any site
Dark red = 7; Red = 6; Orange = 5; Yellow = 4; Green = 3;
Light brown = 2 Dark brown = 1; Intervening colors result from interpolation Black = no forms
predicted

only slightly lower than in Northern forest areas, both of which are dominated by
malacophylls.
Two apparent anomalies are of particular interest, since their results, which occur
on multiple maps, were not expected within the context of generally understandable
302 E. O. Box

patterns. These two areas are California (mediterranean) and submediterranean

Atlantic Europe, from (southern) Britain to Galicia. Although present-day repre-
sentation is low by climatically potential evergreen forms in Atlantic Europe, one
can find palms, numerous conifers, other evergreens, and even eucalypts in local
botanical gardens. California is different in that it covers a moderately large area on
a large continent, and is thus more “subcontinental”, with perhaps more frequent
and severe droughts (and desiccating downslope winds), than the other regions of
Mediterranean-type climate. Perhaps it is for this reason that summergreens and
conifers play a large role, as also in the temperate rainforest of northwestern North
America, which is dominated by tall conifers, unlike the broad-leaved evergreen
rainforests of the world’s other regions of marine west-coast climate.
If one takes a zonal view of world bioclimates and vegetation, sensu Walter for
example (cf [15]), one can sort the zones into those that are everwet (no dry season,
such as equatorial); those that are always dry (no wet season, i.e. deserts); one that
is always cold (polar); and those that have both a wet season and a dry season.
These last are perhaps the most interesting, since these permit vegetation responses
ranging from forest (wettest) through scrub and shrubland (mid-range of precipi-
tation) to sparse steppe or (dwarf) scrub at the dry end. In the wet-dry tropics, scrub
is largely a degradation stage, and much former woodland and even forest has been
converted permanently to savanna. This leaves only the generally subhumid to
slightly humid Mediterranean-climate regions with large areas of climatically more
or less stable, open maquis, chaparral, kwongan or other scrub vegetation. Scrub is
by definition a mix of woody plant forms, in which no single form is dominant and
form diversity should thus be relatively high. In addition, the transitional geo-
graphic positions of the various Mediterranean-type regions permit potential mixing
of tropical, subtropical and warm-temperate evergreen elements with temperate
summergreen and some tropical raingreen elements.

5 Why Does the Mediterranean Seem So Rich?

The maps of predicted form and character richness (Figs. 1 and 2), the overview of
predicted richness in regions (Table 4) and the brief overview of sites worldwide
with greatest predicted form richness (Table 4) all suggested the greatest numbers
of potential plant forms in tropical (but not equatorial) to subtropical regions with
significant precipitation seasonality. One major advantage leading to high form (and
species) richness in the frost-free tropics involves the appearance of numerous
epiphyte and other frost-sensitive forms, even if some of these “additional” forms
are only in understoreys and do not contribute greatly to overall vegetation struc-
ture. The areas suggested as richest in plant forms generally have the climatic
potential for closed forest, even if only seasonally closed (i.e. deciduous). These
results, however, also suggested fairly high potential form richness in at least some
regions with Mediterranean-type climate.
Why Does Mediterranean Vegetation Seem So Diverse? 303

Table 6. Predicted plant-form richness for representative sites in Southern Italy, plus Montpellier

Numbers of plant forms predicted to occur, as well as near misses, are shown for sites from central
to southern Italy, plus submediterranean Montpellier for comparison. The “main mix” (right) is the
number of woody and other larger forms predicted, i.e. the ‘main structural elements’ (see main
text), those that provide the main structure of the vegetation. Their fraction of the total number of
forms is shown in the last column. Highest values in most columns are shown in boldface. Highest
form richness generally occurs at less dry sites, but only Salerno and montane Floresta are truly
wet for Mediterranean conditions and should have closed forest overstoreys. For the other sites, all
with potentially more open vegetation, more than half of the forms predicted would contribute
significantly to the overall stand structure. This is what makes the vegetation of
Mediterranean-climate regions so structurally diverse. Messina, with the most total and
main-mix forms predicted, is subhumid (P/PET < 1.0) but does have relatively high summer
humidity and precipitation. With 53 forms, Messina is the only Mediterranean site that approaches
the richness level of the world’s form-richest sites (see Table 5). The large number of near misses
at Montpellier reflects its submediterranean position, where many forms may be precluded only
barely by one aspect of winter temperature. Abbreviations: Elev = elevation; Py, P = annual
precipitation (mm); Pmin = precipitation of driest summer month (mm); PET = annual potential
evapotranspiration (mm).

In order to understand the apparent structural complexity (form richness) of

mediterraanean vegetation, we must look at what I will call the ‘main structural
complex’ at each location, i.e. those larger forms (main structural elements, vide supra),
that receive full sunlight and are not submerged in shady understoreys. This main
structural complex is what builds the physical structure of a vegetation stand and gives
mediterranean vegetation its characteristic appearance of diversity. Even non-forest
vegetation may develop a fairly dense overstorey, as in some chaparral, maquis, ma-
torral, macchia, etc.; but with disturbance, the more diverse mix of potential plant
forms is revealed. This can perhaps be understood best by looking at the Palermo
example (Table 3). For Palermo, 35 plant forms were predicted to be possible in the
climatic-potential natural vegetation. Of these 35 (but omitting the vines, epiphytes and
cryptic succulents), 23 of the remaining 32 forms can be considered part of the main
structure (i.e. layers T1, T2, S1 and S2). Even without the three predicted forms that do
not occur at Palermo (no species examples shown in the righthand column), this is a
high percentage. This is why Mediterranean vegetation seems so diverse.
Corresponding values for total forms predicted and those in the main structure
(“main mix”) are shown in Table 6 for various sites in southern Italy. Here also one
304 E. O. Box

sees high percentages for the main-mix forms, except at the much wetter sites,
Salerno and Floresta, which potentially should have closed forest. The results in
Table 6 also suggest that potential form richness is generally greater if sites are
sub-humid (i.e. potentially have open vegetation) and have more summer rainfall
(as represented by Pmin). Up to the point of potential forest canopy closure (i.e. P/
PET around 1.0), there is also a less clear suggestion of increasing total form
richness (but not a higher main-mix percentage). Overall annual water balance
(annual P and P/PET) appears to be less important. At global scale there was some
suggestion that diversity in general decreases along gradients of increasing stress. In
Mediterranean climates, the main stress factor is summer dryness, except on the
poleward margins where significant winter frost reduces vitality and growing
seasons.
Another aspect of the model results, shown also in Table 6, involves the number
of “near misses”, i.e. those plant forms that were not predicted to occur because
they were rejected by only one of the screening variables. Many of these near-miss
forms, especially those rejected by a moisture variable, may nevertheless occur on
special microsites in natural situations but especially in cultural landscapes. The
near misses at Palermo included the following plant forms:
- Limited by winter cold (Tmin): Bottle Trees, Bottle Palms
(with water-storing trunks)
Leaf-Succulent Shrubs
(e.g. Crassula)
- Limited by insufficient cold (Tmin): Cold-Suffrutescent Shrubs
(e.g. Artemisia)
- Limited by overall dryness (P/PET): Tropical Evergreen Microphyll Trees
(e.g. Meliaceae)
Tropical Xeric Needle Trees
(e.g. Casuarina)
Sub-Mediterranean Needle Trees
(e.g. Cedrus)
Temperate Needle Trees
(pines, e.g. Pinus sylvestris)
Laurophyll Small Trees
(e.g. Laurus nobilis)
Temperate Evergreen Small Trees
(e.g. Ilex aquifolium)
Palmiform Tuft-Trees and Treelets
(i.e. palms)
Mesic Stemgreen Arborescents
(e.g. Spartium junceum)
Tropical Evergreen Shrubs
(perhaps facultatively deciduous)
Mesic Semi-Shrubs
- Limited by lack of aridity (P/PET): Ephemeral Desert Forbs
Mat-Forming Thallophytes (i.e. mosses)
- Limited by summer dryness (Pmin): Summergreen Arborescents
Why Does Mediterranean Vegetation Seem So Diverse? 305

Fig. 3. Mediterranean Sections of the Form and Character Maps of Figures 1 and 2. The four
panels represent: a) total numbers of plant forms predicted (upper left); b) number of plant
structural types (upper right); c). seasonal foliation habits (lower left); and d) textures of the
photosynthetic surfaces (lower right). See Figures 1 and 2 for color legends.

The forms limited by low winter temperatures can often be found near buildings
and are certainly represented in the Palermo Botanical Garden, except for
cold-winter Artemisia. Most of the forms limited by overall (annual) or summer
dryness also occur on more favorable microsites, especially the palms, needle trees,
and leafless stemgreen Spartium junceum. Only the desert ephemerals, which
generally require arid climates with only short wet periods, are probably not rep-
resented in the Palermo Botanical Garden. For Palermo there were 18 such near
misses predicted, and the numbers are even higher for most of the other sites in
Table 6. Indeed, these numbers are generally a bit higher than for sites in most
world regions with other climate types.
Structural richness results not only from the number of coexisting plant forms at
a place but also from the number of different plant structural types, types of pho-
tosynthetic organ, and seasonality patterns found among the different plant forms
possible. World patterns for such ‘character richness’ were shown in Fig. 2, but it is
difficult to see details of individual regions on such world maps. Albeit with coarse
resolution, the Mediterranean portions of the maps in Figs. 1 and 2 are reproduced
in Fig. 3. From these regional sections it is possible to see more clearly the sug-
gestion of especially high plant character richness in central and southern Italy (all
four maps in Fig. 3) and some other regions for particular characters.
The total number of plant structural types (trees, shrubs, forbs, etc.) predicted to
be possible in the potential vegetation is especially high in southern Italy and still
relatively high in the southern Iberian Peninsula, the near-coastal Levant, and some
306 E. O. Box

Table 7. Predicted plant-form richness in Mediterranean regions of the World

The range of the total numbers of plant forms predicted at representative sites in each region are
shown in the NF column. For each region, the site with the most forms predicted is listed first,
followed by other form-rich sites. For example, Messina (eastern Sicily) was predicted to be the
richest site in the Western Mediterranean, with 53 plant forms predicted, followed by Porto with
43 forms. With 53 forms, Messina is the only truly Mediterranean site in the world data-base
(n=2889) that approaches the potential richness of the richest sites in Table 5. The Mediterranean
region proper appears to be potentially the richest of the world’s five regions of
Mediterranean-type climate. California seems to have fewer plant forms than the other regions,
although the core mediterranean region in central Chile is also low. (Temuco and Albany are both
more humid submediterranean.)
Abbreviations: n = the total number of sites in each region; NF = number of forms predicted
(range).

coastal areas around the Aegean Sea. These are all coastal or near-coastal areas that
receive somewhat more precipitation and humidity in the otherwise quite dry
Mediterranean summer. Predicted richness in seasonality habits is generally highest
in the same areas, although here the largest area of highest predicted richness seems
to have shifted to the [drier] southeastern quadrant of Spain. For richness in types of
leaf or other photosynthetic surfaces (sclerophyll, malacophyll, etc.), southern Italy
is again predicted to have the highest levels, followed this time, though, by the
windward, more humid Atlantic side of the Iberian Peninsula. Taken together, these
patterns suggest a special “Garden of Eden” position for southern Italy, as was
suggested to this author by the very first predictive results in 1981 (cf [7], p. 100; or
see [16], Fig. 1).
The Mediterranean region is compared with the world’s other regions of
Mediterranean-type climate in Table 7, which shows the range of total plant forms
predicted at a representative set of sites in each region. These results (and the maps)
suggest high form richness in the Mediterranean region proper, perhaps especially
Why Does Mediterranean Vegetation Seem So Diverse? 307

in the western Mediterranean but only slightly less in the eastern Mediterranean.
The predicted form richness then decreases more significantly in the drier Levant
and on into the more continental inland Middle East (e.g. Iraq, Iran) – with the
notable exception of Rasht (40 forms), in the relatively wet climatic strip just south
of the Caspian Sea. The Mediterranean region may be more form-rich than the other
four comparable regions because it is not restricted to a narrow ocean-coastal strip
on a large continent, which may result in drier summers. The Mediterranean Sea is
in just the right latitude to provide precipitation throughout much of its inland
extent. The Levant and inland Middle East are drier, and apparently thus less
diverse, because they are farther from the winds that bring moisture from the sea.
North Africa is less form-rich because it is nearer the core zone of subtropical high
pressure and is simply more arid overall.
The Australian and South African mediterranean regions are especially
species-rich and are predicted here also to have high form richness, but both are in
reality dominated by shrub forms, with few trees and almost no deciduous elements.
California (including the northern part of Baja California) is predicted to be less
form-rich than the other regions, perhaps because it is generally drier in summer
(see Table 5). This is curious since California is usually regarded as one of the
world’s main hotspots of species richness. California is the only one of the world’s
mediterranean regions that extends far inland without a large source of water (such
as the Mediterranean Sea). This greater continentality means colder winter tem-
peratures (at least inland) as well as less and more irregular precipitation. This may
explain partly the greater importance of summergreen forms and conifers in western
North America. The dry Middle East is somewhat comparable in this respect.
Finally, we might also look at submediterranean regions, which already have
mild enough winter temperatures to permit sclerophyllous and other evergreen
broad-leaved plants but are not yet so summer-dry as to preclude summergreens.
This suggests high potential form richness – but not necessarily. For the main
vegetation structure (cf Table 6), the summer-drier true mediterranean climate
produces more open vegetation with more forms contributing to stand structure,
even if some water-demanding summergreen forms are lost. And even though
winters are relatively mild in the submediterranean zone, they do not permit the
truly subtropical or tropical forms, such as larger stem-succulents, most palms,
hot-desert shrubs, many tropical deciduous (raingreen) plants, and most xeric
arborescents. So, the true mediterranean zone, which does pick up these forms, may
indeed have greater potential form richness than the submediterranean.

6 Conclusion

To begin with, these are only results from a necessarily imperfect model – some
will say playing with numbers and abstractions. At best such models can do no
more than suggest potentials, which may or may not be worth looking into further.
At least, though, the model herein does provide greater resolution of vegetation into
308 E. O. Box

constituent plant types than do other global models, and it is based as far as possible
on accepted climatic limits, such as cardinal temperatures. This model is based on
the general logic that form follows function (and vice versa), i.e. that there are
general reasons why drier climates tend to have smaller plants with smaller, often
harder leaves: i.e. that there is a strong relationship between plant
pheno-physiognomy and plant function.
One might normally expect highest values of both plant-form and character
richness in the tropics, lowest values in polar and arid areas, perhaps low values
also in boreal areas, and perhaps quite variable results in some oceanic areas (cf
[20]). Taken together, the results shown herein suggest:
– greater potential richness (at least for plant forms) in areas with mild or no real
winter;
– greatest potential richness in climates with both a wet season and a dry season;
and
– depressed potential richness over large high-latitude land masses and in the
everwet tropics.
The geographic regions predicted to be richest in different plant forms seem to have
two features in common: they are mostly warm, with marginal frost (i.e. not
equatorial), and have some degree of dryness, either a wet-dry seasonality pattern or
an overall subhumid character (see fuller list in [16], Fig. 3).
The results also suggest the potential for many co-existing forms in the main
vegetation structure, and fairly high potential richness overall, in climates that are:
– warm-temperate to subtropical (with little or no frost), permitting
evergreen-summergreen overlap;
– subhumid to moderately humid, permitting somewhat open vegetation, not
closed forests;
– seasonal (dry and wet seasons), fostering the co-occurrence of evergreen and
raingreen forms; and
– not too arid, where richness decreases.
Richness also seems enhanced in such situations by some amount of precipitation in
the dry season. This description fits Mediterranean-type climates well, and the result
is what gives the vegetation of Mediterranean-type climates its characteristic
appearance of diversity.
Along the main gradients of increasingly favorable climatic conditions (in-
creasing simultaneous warmth and wetness), there seems to be a general pattern in
the model results, as perhaps also in nature, of increasing potential richness in plant
forms. This suggests that new forms appear before old forms drop out, and that the
greatest potential richness should thus appear in transitional zones. As the climate
permits new forms, though, these usually appear first as ornamentals or in other
cultivations before they appear in natural vegetation, for example the widely rec-
ognized ‘laurophyllization’ of the vegetation south of the Alps, by which evergreen
broad-leaved garden species have escaped and become naturalized. This also
contributes to the overall impression of high diversity in Mediterranean landscapes,
which are of course largely cultural landscapes now.
Why Does Mediterranean Vegetation Seem So Diverse? 309

Along broad-scale geographic gradients, increasing stress may lead to more

ecological niches for plant species, with less chance for dominance by individual
species. As the stress becomes harsher, though, species diversity usually decreases.
In mediterranean climates, the most significant stress factor is often summer dry-
ness. But before this point is reached, do the increases in stress, perhaps species
richness, and species niches lead also to more plant forms? The model results
suggest perhaps the opposite. Is there some “intermediate stress hypothesis” for
plant forms, and do these results reflect patterns in floristic richness or other patterns
that seem worth pursuing?
Apparent “hotspots” of plant diversity have been mapped by Groombridge et al.
([31], Fig. 15.1) and by [4]. Hotspots in general have been catalogued in more detail
by Groombridge et al. ([31], Tables 15.1, 15.6), including the 241 Centres of Plant
Diversity identified at that time by the Plant Conservation Programme of the
International Union for the Conservation of Nature. The greatest concentrations of
overall diversity hotspots are in Southeast Asia, Meso-America (and south to
Ecuador), and to a slightly lesser extent from the eastern Mediterranean eastward
into Iran, and in various parts of central to southern Africa. The table of endemism
hotspots for higher plants (Groombridge et al. [31], Table 15.1) lists 18 hotspots,
most of which are also in the tropics, including everwet regions (e.g. western
Amazonia, northern Borneo, the Colombian Chocó, and peninsular Malaysia). Four
of the world’s five regions of Mediterranean-type climate are included, but not the
Mediterranean region itself. Some areas with higher levels of predicted form
richness on multiple maps seem to coincide with both these compilations, including
subtropical to tropical southern China and Southeast Asia, areas of
Mediterranean-type climate (except California), central to southern Mexico, parts of
New Zealand, and various parts of Australia. Southern China and Meso-America
have long been suggested as areas of great diversity and speciation (e.g. Barthlott
et al. [4]). High levels of taxonomic radiation also occurred in Australia and New
Zealand, but these have been attributed partly to small taxonomic stocks radiating
on long isolated but climatically diversifying land masses.
Possession of a seasonal but overall humid climate has already been suggested as
a common factor among the sites with the very most total plant forms predicted.
A relatively mild winter (or none at all) seems important, but the existence of drier
periods during the year seems even more important for the development of high
potential plant form and character richness. Some drier areas with greater richness
may be seen as “marginal” climates, i.e. transitions between wet and dry, which
immediately suggests these as natural cradles for adaptation and evolution.
With global warming, the Mediterranean region is expected to become generally
drier, as the subtropical high-pressure belt moves northward into southern Europe.
A similar fate is expected in the world’s other regions of Mediterranean-type cli-
mate, and California in particular has already seen historic drought in recent years.
Since mediterranean vegetation is already largely open vegetation, greater dryness
would suggest less apparent diversity in vegetation stands rather than any gain from
the more open vegetation that would accompany greater dryness. Water-demanding
summergreen forms may be lost the most, and there would be only small gains
310 E. O. Box

among desert forms. Mediterranean zones are expected to advance poleward,

upward, and perhaps inland – but perhaps not in southern Australia and South
Africa, where there is no more land to advance into. The relatively large, inland
submediterranean zone in Italy may be reduced but may also move northward
across the Alps. For that portion of the Middle East that receives significant winter
rainfall and can be called Mediterranean, the prognosis is more difficult. Indeed the
inland climate of Iran, in particular, is somewhat unusual already, with winter-rain
and summer-rain floristic elements mixing, without a large representation of the
sclerophyll element (e.g. [6]).
Of course, with environmental change, plant characters may also change as the
plant species adapt to the new conditions. The greatest potential for change in form
characters probably involves seasonal habits, since these can be quite “plastic”,
(facultative rather than obligatory) in many taxa, even under current conditions,
especially in the tropics and subtropics. Leaf consistency and size may also change
in some taxa, with leaves perhaps becoming “harder” or smaller or both with
increasing dryness. Least changeable is probably the overall plant architecture, i.e.
the structural type.

References

1. Alekseenko MI (1968) Edifictors and dominants of the plant communities of chalk outcrops in
the basin of the Northern Donets River. Biol. Nauka v Univers. I Ped. Inst. Ukrainy za 50 Let,
pp 84–85. Kharkov University (in Russian)
2. Auerbach M, Shmida A (1987) Spatial scale and the determinants of plant species richness.
Trends Ecol Evol 2:238–242
3. Axelrod DI (1958) Evolution of the Madro-Tertiary geoflora. Bot Rev 24(7):433–509
4. Barthlott W, Lauer W, Placke A (1996) Global distribution of species diversity in vascular
plants: towards a world map of phytodiversity. Erdkunde 50:317–327+map
5. Blondell J, Aronson J, Bodiou J-Y, Boeuf G (2010) The Mediterranean region: biological
diversity in space and time, 2nd edn. Oxford University Press, Oxford
6. Blumler MA, Plummer JC (2015) Deciduous woodlands in the near eastern fertile crescent,
and a comparison with California. In: Box EO, Fujiwara K (eds) Warm-temperate deciduous
forests around the northern hemisphere, pp 267–276
7. Box EO (1981) Macroclimate and plant forms: an introduction to predictive modeling in
phytogeography. Tasks for vegetation science, vol 1, p 258. Dr. W. Junk BV, The Hague
8. Box EO (1986) Some climatic relations of the vegetation of Argentina, in global perspective,
pp 181–216. Veröff. Geobot. Inst. ETH, Stiftung Rübel (Zürich)
9. Box EO (1987) Plant life forms in Mediterranean environments. Annali Di Botanica 45(2):7–
42
10. Box EO (1995) Factors determining distributions of tree species and plant functional types.
Vegetatio 121:101–116
11. Box EO (1995) Climatic relationships of the forests of East and South-East Asia. In: Box EO
et al (eds) Vegetation science in forestry: global perspective based on forests ecosystems of
East and Southeast Asia. Kluwer, Dordrecht, pp 23–55
12. Box EO (1996) Plant functional types and climate at the global scale. J. Veg. Sci. 7:309–320
Why Does Mediterranean Vegetation Seem So Diverse? 311

13. Box EO (1997) Bioclimatic position of evergreen broad-leaved forests. In: Island and
high-mountain vegetation: biodiversity, bioclimate and conservation, pp 17–38. Proceedings,
annual IAVS meeting, Tenerife, April 1993. Universidad de La Laguna, Servicio de
Publicaciones, Tenerife (Canary Islands, Spain).
14. Box EO (ed) (2016) Vegetation structure and function at multiple spatial, temporal and
conceptual scales, p 578. Springer, Cham
15. Box EO (2016b) Global bioclimatic zonation. In: Box EO (ed) Vegetation structure and
function at multiple spatial, temporal and conceptual scales, pp 3–52. Springer
16. Box EO (2019) Form and character diversity of potential world vegetation. Flora 254. https://
doi.org/10.1016/j.flora.2019.03.001
17. Box EO, Fujiwara K (2005, 2013) Vegetation types and their broad-scale distribution. In: van
der Maarel E (ed) Vegetation ecology, pp 106–128, van der Maarel E, Franklin J (eds) 2nd
edn., pp 455–485+plate 15.1 (color). Wiley, London
18. Brophy C, Dooley A, Kirwan L, Finn JA, McDonnell J, Bell Th, Cadotte MW, Connolly J
(2017) Biodiversity and ecosystem function: making sense of numerous species interactions
in multi-species communities. Ecology 98(7):1771–1778
19. Crame JA (2001) Taxonomic diversity gradients through geological time. Divers Distrib
7:175–189
20. Crawford RMM (2008) Plants at the margin. Cambridge University Press, Cambridge, p 478
21. Davis SD, Heywood VH, Hamilton AC (1994) Centres of plant diversity: a guide and strategy
for their conservation. Vol. 1: Europe, Africa, Southwest Asia and the Middle East. World
Wildlife Fund and World Conservation Union, Cambridge
22. Diamond J (1988) Factors controlling species diversity: overview and synthesis. Ann.
Missouri Bot. Gard. 75:117–129
23. Drude O (1896) Deutschlands Pflanzengeographie. Handbuch Dt. Land-. u. Volkskunde,
Engelhorn, Stuttgart, p 502
24. Erwin DH (2009) Climate as a driver of evolutionary change. Curr Biol 19:R575–R583
25. Evans KL, Warren PH, Gaston KJ (2005) Species-energy relationships at the macroecological
scale: a review of the mechanisms. Biol. Rev. 80:1–25
26. Fekete G, Szuko Lacza J (1970, 1971) A survey of plant life-form systems and the respective
research approaches. Annales Hist Nat Musei Nation Hungarici Pars Botanica 62:115–127,
63:37–50
27. Gavilan RG (2003) Does global warming pose a threat to Mediterranean biodiversity?
Bocconea 16(1):379–395
28. Gilbert N (2011) Plant biodiversity theory debunked. Nature. https://doi.org/10.1038/news.
2011.553. Accessed 22 Sept 2011
29. Gillison AN (2013) Plant functional types and traits at the community, ecosystem and world
level. In: van der Maarel E, Franklin J (eds) Vegetation ecology, pp 347–386
30. Grisebach ARH (1872) Die Vegetation der Erde nach ihrer klimatischen Anordnung, Leipzig
31. Groombridge B (ed) (1992) World conservation monitoring centre. Global biodiversity: status
of the earth's living resources, p 585. Chapman & Hall, London
32. Hawkins BA, Field R, Cornell HV, Currie DJ, Guégan J-F et al (2003) Energy, water, and
broad-scale geographic patterns of species richness. Ecology 84(12):3105–3117
33. Heywood VH (1995) The Mediterranean flora in the context of world biodiversity. Ecol
Mediterr 20:11–18
34. Hillebrand H (2004) On the generality of the latitudinal diversity gradient. Am Nat 163:192–
211
35. Hooper DU, Chapin FS, Ewel JJ et al (2005) Effects of biodiversity on ecosystem functioning:
a consensus of current knowledge. Ecol Monogr 75(1):3–35
36. von Humboldt A (1806) Ideen zu einer Physiognomik der Gewächse, p 28. F. G. Cotta,
Tübingen. Reprinted 1957 by Akad, p 19. Verlagsgesellschaft, Leipzig
37. von Humboldt A (1808) Ansichten der Natur mit wissenschaftlichen Erlaeuterungen. J. G.
Cotta, Tübingen
38. Huston M (1979) A general hypothesis of species diversity. Am Nat 113:81–101
312 E. O. Box

39. Keith DA (ed) (2017) Australian vegetation, 3rd edn. Cambridge University Press, Cambridge
40. Larcher W (1973) Limiting temperatures for life functions in plants. In: Precht H, et al.
(eds) Temperature and life. Springer-Verlag, Berlin
41. Lebrun J (1966) Les formes biologiques dans la végétation tropicale. Mém Soc Fr 1966:166–
177
42. MacArthur RH, Levins R (1967) The limiting similarity, convergence and divergence of
coexisting species. Am Nat 101:377–385
43. MacArthur RH, Wilson EO (1967) The theory of island biogeography. Princeton University
Press, Princeton
44. Médail F, Quézel P (1997) Hot-spots analysis for conservation of plant biodiversity in the
Mediterranean basin. Ann Mo Bot Gar 84:112–127
45. Médail F, Quézel P (1999) Biodiversity hotspots in the Mediterranean basin: setting global
conservation priorities. Consv. Biol. 13(6):1510–1513
46. Midgley GF (2012) Biodiversity and ecosystem function. Science 335:174–175
47. Mittelbach GG, Steiner CF, Scheiner SM, Gross KL, Reynolds HL et al (2001) What is the
observed relationship between species richness and productivity? Ecology 82:2381–2396
48. Mittelbach GG, Schemske DW, Cornell HV, Allen AP, Brown JM et al (2007) Evolution and
the latitudinal diversity gradient: speciation, extinction and biogeography. Ecol Lett 10:315–
331
49. Mittermeier RA, Gil PR, Mittermeier CG (1997) Megadiversity: earth’s biologically richest
nations. Conservation International, Washington
50. Mueller-Dombois D, Ellenberg H (1974) Aims and methods of vegetation ecology. Wiley,
New York, p 547
51. Nakhutsrishvii G (2013) The vegetation of Georgia (South Caucasus). Springer, Heidelberg
52. Olson DM, Dinerstein E (1998) The global 200: a representation approach to conserving
earth’s most biologically valuable ecoregions. Conserv Biol 12:502–515. World Wildlife
Fund, Washington
53. Palmer MW (1994) Variation in species richness: towards a unification of hypotheses. Folia
Geobotan Phytotax 29:511–530
54. Pausas JG, Austin MS (2001) Patterns of plant species richness in relation to different
environments: an appraisal. J Veg Sci 12:153–166
55. Pianka ER (1966) Latitudinal gradients in species diversity: a review of concepts. Am Nat
100:33–46
56. Quézel P (1985) Definition of the mediterranean region and origin of its flora. Dr Junk W,
Dordrecht
57. Raimondo FM, Domina G, Spader V (2010) Checklist of the vascular flora of Sicily. Quad
Bot Amb Appl 21:189–252
58. Raunkiaer C (1934) The life forms of plants and statistical plant geography. Clarendon,
Oxford
59. Reid WV (1998) Biodiversity hotspots. Trends Ecol Evol 13:275–280
60. Rohde K (1992) Latitudinal gradients in species diversity – the search for primary cause.
Oikos 65:514–527
61. Schnell R (1970–1977) Introduction à la Phytogéographie des Pays Tropicaux, vols 4 (see vol
II: chapter 3). Gauthier-Villars, Paris
62. Shmida A, Wilson MV (1985) Biological determinants of species diversity. J Biogeogr 12:1–
21
63. Skeels A, Cardillo M (2017) Environmental niche conservatism explains the accumulation of
species richness in Mediterranean-hotspot plant genera. Evolution 71(3):582–594
64. Smith TM, Shugart HH, Woodward FI, Burton PJ (1993) Plant functional types. In:
Solomon AM, Shugart HH (eds) Vegetation dynamics and global change. Chapman & Hall,
New York, pp 272–292
65. Solomon AM, Shugart HH (eds) (1993) Vegetation dynamics and global change. Chapman &
Hall, New York
Why Does Mediterranean Vegetation Seem So Diverse? 313

66. Stehil FG, Douglas RG, Newell ND (1969) Generation and maintenance of gradients in
taxonomic diversity. Science 164:947–949
67. Stevens GC (1989) The latitudinal gradient in geographical range: how many species coexist
in the tropics. Am Nat 133:240–258
68. Szuko-Lacza JS, Fekete G (1969, 1972) A survey of plant life-form systems and the
respective research approaches. Annales Hist Natur Musei Nat Hungarici Pars Botanica
61:129–139, 64:53–62
69. Thanos CA (2005) The geography of Theophrastus’ life and of his botanical writings. In:
Karamanos, AJ, Thanos, CA (eds) Biodiversity and natural heritage in the Aegean, pp 113–
131
70. Theophrastus (1916, 1926) Enquiry into plants (history of plants), vols I–II. In: Hort AF
(ed) Ancient Greek with English translation. Harvard University Press
71. Walter H (1968, 1973) Die Vegetation der Erde in öko-physiologischer Betrachtung, 3rd edn.,
vol 1, p 744 and vol II, p 1002. Fischer-Verlag, Stuttgart
72. Warming E (1895) Plantesamfund: Grunträk af den ökologiske Plantgeografi. København.
English version 1909: oecology of plants, p 422. Humphrey Milford & Oxford University
Press, Oxford
73. Whittaker RJ (2010) Meta-analyses and mega-mistakes: calling time on meta-analysis of the
species richness-productivity relationship. Ecology 91(9):2522–2533
74. Whittaker RJ, Willis KJ, Field R (2001) Scale and species richness: towards a general,
hierarchical theory of species diversity. J Biogeogr 28:453–470
75. Willig MR, Kaufman DM, Stevens RD (2003) Latitudinal gradients of biodiversity: pattern,
process, scale and synthesis. Ann Rev Ecol Evol Syst 34:273–309
76. Woodward, I. 1987. Climate and Plant Distribution. Cambridge University Press. 174 pp.
77. Wright DH (1983) Species-energy theory: an extension of species-area theory. Oikos 41:496–
506
78. Wullschleger S, Epstein HE, Box EO, Euskirchen ES et al (2014) Plant functional types in
earth system models: past experiences and future directions for application of dynamic
vegetation models in high-latitude ecosystems. Ann Bot 114:1–16. https://doi.org/10.1093/
aob/mcu077
Coastal Dune Vegetation Zonation
in Italy: Squeezed Between
Environmental Drivers and Threats

Alicia T. R. Acosta

Abstract The present work provides an overview of recent studies in coastal dune
systems performed by the Plant Ecology Laboratory, Roma 3 University (Italy),
highlighting major findings. In this review major habitat types along the sea-inland
environmental gradient are described, emphasizing their ecological value in terms
of biodiversity and landscape heterogeneity. A synthesis of major threats is also
presented. Occupying transition zones between terrestrial and marine ecosystems,
coastal dune systems are characterized by strong environmental gradients which
determine the coexistence of different plant communities in a relatively small area.
Despite the presence of numerous habitats which are considered priorities for
international conservation goals, coastal dune ecosystems have undergone consis-
tent transformations and are currently considered to be highly endangered. The
understanding of major threats is a crucial step toward developing effective man-
agement strategies to safeguard coastal dune biodiversity, in the hope of preserving
the functionality and uniqueness of these fragile ecosystems.

Keywords Coastal vegetation Coastal dune systems Dune species Threatened

ecosystems

1 Introduction

By occupying the transition between terrestrial and marine ecosystems, coastal

dunes represent particularly interesting environments both from an ecological and a
landscape perspective. The complex, compressed zonation of plant communities
along the sea-inland gradient is, in fact, one of the most interesting features of
Mediterranean sandy shores [1, 10, 24, 25, 43, 46, 48], giving rise to a unique
biodiversity in terms of both number of different habitats and species composition
[49]. Mediterranean coastal areas have been the scene of a wide variety of human

A. T. R. Acosta (&)
Department of Sciences, Rome 3 University, V. le Marconi 446, 00146 Rome, Italy
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 315
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_16
316 A. T. R. Acosta

activities for millennia. However, anthropic pressure intensified greatly over the
course of the twentieth century, with remarkable land-use changes for urbanization,
harbor and industrial centers, roadways, and increased tourist activities, just to cite a
few examples. Particularly after World War II, various related human pressures
caused severe losses in the extent of dune systems, the same factors still threatening
these communities to this day. It is widely acknowledged that coastal dune habitats
are currently in a critical conservation condition in the whole Mediterranean Basin.
In fact, coastal sandy ecosystems are currently pinpointed as being vulnerable to
high, widespread biodiversity loss, making them among the most endangered
habitats at European level [25, 26]. The continuous degradation of coastal dune
systems is clearly connected to the increasing intensification of some human
activities, such as the recreational use of beaches. Especially during the warm,
sunny summer months, many coastal areas attract a great number of local visitors
and international tourists. Thus, human activities may also represent a serious
threat, even though at the same time, they provide many coastal areas with their
main source of income. The understanding of these threats is a crucial step toward
developing effective management strategies to safeguard coastal dune biodiversity.
If these ecosystems are managed according to a thorough, sustainable conservation
strategy, human activities could also represent a long-lasting economic income for
local populations.
The present work brings together recent published studies on coastal dune
systems performed by the Plant Ecology laboratory, Roma 3 University (Italy).
Thus, here I present an outline of the ‘state of the art’ of our studies in coastal dunes
by summarizing major findings. This review is not intended as a comprehensive
synthesis of the studies conducted in coastal dunes in Italy, but an overview of our
previous works.

2 Coastal Dune Vegetation Zonation: Environmental

Gradients, Habitats and Species Diversity

On sandy coastal habitats, factors related to substrate and to wind action vary along
the sea-land ecotone, forming a marked directional stress and natural disturbance
gradient. Environmental factors shift along this gradient, concomitantly with plant
communities. It is interesting to note that the gradient is similar both in the
Mediterranean and in oceanic coastal ecosystems [18], with analogous trends along
the dune profile [21]. However, we could also highlight some peculiarities of
Mediterranean systems, such as the relatively minor importance of wind-related
variables as compared to soil properties. In particular, organic matter and grain-size
variability are closely correlated with the distribution of plant communities along
the coastal zonation [22].
In Mediterranean ecosystems, vegetation is organized along a well-known
zonation ranging from pioneer annual communities on the drift line to
Coastal Dune Vegetation Zonation in Italy ... 317

Fig. 1 Simplified representation of the vegetation zonation in Mediterranean coastal dune

ecosystems. Vegetation is organized along a well-known zonation ranging from pioneer annual
communities on the drift line to Mediterranean scrubs on fixed dunes. EUNIS Habitats (see
Table 1) were identified by different letters: A B1.1 Sand beach driftlines, B B1.3 Shifting coastal
dunes, C B1.4 Coastal stable dune grassland (grey dunes), D B1.6 Coastal dune scrub.

Mediterranean scrub on fixed dunes. This zonation is observed not only in vascular
plants but also in other taxonomic groups, such as bryophytes [32]. Habitats of
conservation interest can be identified based on the presence of characteristic,
diagnostic plant species, a list of which has been standardized across Europe in the
EUNIS classification of habitats (EUNIS B1; European Environment Agency
2008). Moreover, several dune plant assemblages have been included in the
Annex I of the Habitats Directive [9] (Table 1, Fig. 1).
Acosta et al. [2] analyzed coastal dune vegetation zonation in relation to dune
morphology. Later, Bazzichetto et al. [7] described the morphological profile of
each coastal dune habitat along the zonation, using morphological variables such as
elevation, slope, curvature and sea distance. By integrating field-collected vegeta-
tion data and remotely sensed imagery, these authors found that habitats occur
differentially, and that dune elevation and sea distance are key factors in shaping
habitat occurrence along local gradients. Moreover, species richness varied along
the sea-to-land gradient, reaching its peak in the intermediate sector of the vege-
tation zonation. Nevertheless, species richness seems to differ significantly only
between the upper beach and all the other habitat types [4]. The upper beach is a
very selective environment with extreme maritime conditions. In fact, very few,
highly specialized vascular plant species can survive here, with specific morpho-
logical and physiological traits usually interpreted as a response to salt, drought and
unstable substrate. Outside the drift line, coastal dune plant communities tend to be
perennial, although annual grasslands are an interesting exception. Pioneer psam-
mophilous plants have evolved a variety of adaptations that allow survival, growth
and reproduction under relatively harsh conditions. Proceeding inland, vegetation is
318 A. T. R. Acosta

Table 1. EUNIS, Nature 2000 and main phytosociological Alliances.

progressively less exposed to the harsh maritime environment and gradually less
tolerant to salt spray, strong winds and sand burial [1]. A significant variation in the
degree of specialization was also observed along the vegetation zonation: habitats
closer to the sea and under greater stress held higher proportions of specialized
species and had higher average specialization levels [13]. Regarding fine-scale
community assembly, Santoro et al. [40] highlighted how community assemblage
patterns were strongly linked to the sea-land environmental gradient. In pioneer
foredune communities, an aggregated assemblage was observed. In contrast, pro-
gressing along the zonation, assemblage patterns first shifted to random and then
became segregated in the fixed dunes. These results are consistent with the “Stress
Gradient Hypothesis” [8], which predicts that along an environmental gradient,
facilitation drives the assembly of plant communities where conditions are harsh,
whereas competition is predominant where conditions are less severe. In addition,
Conti et al. [17] highlighted the importance not only of the environmental condi-
tions along the already well-known sea-land gradient, but also of environmental
heterogeneity at fine local scale, suggesting that micro-abiotic filtering might play a
more important role than previously anticipated.

3 Major Threats

3.1 Land Use Changes and Landscape Fragmentation

Most studies on coastal dune vegetation changes have been conducted using
remotely sensed data, generally agreeing that the composition and spatial patterns
of coastal landscapes have been drastically modified by human activities over the
last 50 years. In particular, the expansion of artificial areas is an important driver of
this transformation, together with afforestation and the loss of land to new
Coastal Dune Vegetation Zonation in Italy ... 319

agricultural activities [28]. As on other Mediterranean coasts, agricultural and

afforestation lands tend to be concentrated in areas of the inner sectors of the coastal
zonation, mostly affecting fixed dunes. Therefore, it seems that the typical envi-
ronmental gradient of coastal ecosystems mentioned before, which constrains the
natural vegetation zonation, also shapes human activities such as agriculture and
afforestation. In contrast, urban expansion involves the whole landscape and seems
less constrained by this gradient, as no preferences for any particular natural dune
cover type have been observed [28].
It seems that landscape patterns and transformations affect taxonomic diversity
more moderately than do local abiotic and biotic site conditions, although a slight
decline on highly disturbed coastal landscapes was observed, mainly in patches
sharing borders with artificial areas [30]. Surprisingly, the effect of the landscape
spatial pattern on functional diversity was stronger: a generalized increase in
functional diversity was seen in human-dominated contexts, but it decreased in
those areas that have undergone high urbanization rates in the last 60 years. In this
way, even if a short-term enrichment in functional traits related to human presence
was evident, functional diversity could be negatively affected in the long term.
For floristic changes, Del Vecchio et al. [19] analyzed changes in plant species
composition of coastal dune habitats over a 20-year period in a sector of the
Adriatic coast. Subsequently, Prisco et al. [35] assessed major changes between the
late 1960s and the present day at national level, taking advantage of the extensive
spatial and temporal coverage of a coastal dune vegetation data-base [34].
Unexpectedly, results in protected areas showed positive trends in species cover but
without clear evidence of climatic preferences, at least when using Ellenberg’s
indicator values [35]. A possible explanation is that the recent incorporation of
Natura 2000 sites into the protected-area network may have limited the effects of
urbanization and beach exploitation activities on coastal habitats, thus promoting
species cover. However, it is also worth noting that large stretches of Italy’s
coastlines remain unprotected and are at risk of further biodiversity declines.
Regarding recent erosion/accretion processes, positive trends in vegetation cover
were also observed by using permanent vegetation transects, but only at stable or
accreting sites [36]. In particular, the cover of the key species Ammophila arenaria
increased at sites where the shoreline is accreting, so underlining that highly rele-
vant plant species of mobile dunes can develop effectively without erosion pro-
cesses and with relatively low anthropic disturbance. As could have been
anticipated, negative trends were observed at sites with coastal erosion [36].

3.2 Trampling

Tourism is a growing activity with great economic potential, in terms of both

employment and income. The greatest growth in tourism is occurring in the
sub-sector of coastal and marine tourism, while the Mediterranean region plays host
320 A. T. R. Acosta

to around 33% of the world’s tourism industry [5]. Among direct threats, increasing
numbers of visitors may also mean increased trampling and the consequent
degradation of coastal dune ecosystems [5]. A four-year experiment highlighted
how fencing can be an effective method for promoting the passive recovery of
coastal dune vegetation, in terms of both species cover and species richness, as well
as their spatial organization [41]. This capacity for rapid recovery is, in fact, a
typical feature of many coastal dune plants, which are often well adapted to high
levels of stress and disturbance [12, 51] and are thus able to recover quickly once a
source of disturbance has been limited. Therefore, fencing demonstrates a high
potential as a passive recovery strategy for improving the conservation status of
degraded dune habitats.

3.3 Alien Species

Invasive species constitute a major threat to biodiversity, and their impact on

natural and semi-natural habitats has been the focus of many ecological studies. In
Europe, coastal dunes are among the habitats most invaded by alien species [14].
Expansion of alien species has characterized many coastal dune systems of the
whole Mediterranean basin, leading to drastic changes in floristic composition.
Indeed, native species of coastal areas are in a fragile equilibrium with their natural
environment, and as such they are extremely sensitive to human-driven modifica-
tions [13].
At regional scale, human population density was the most important factor
negatively related to native species richness in central Italy. On the other hand, alien
species richness was strongly related to climatic factors, especially to precipitation
regimes (more aliens in areas with more precipitation). Secondarily, alien intro-
ductions appear to be related to recent urban sprawl and associated gardening [11].
In addition, fragmentation of natural habitats in the landscape (cf. [23]) may directly
enhance the invasibility of ecological communities, thus reducing the extent of
natural habitats and increasing isolation and the decline of ecological quality of the
remaining fragments. The resulting fragmented pattern has the potential to boost
habitat colonization by alien species from the surrounding matrix. Malavasi et al.
[29] found evidence that fragmentation of natural coastal habitats enhances the
number of aliens present in each vegetation patch. This increase in patch invasion
was observed in correspondence with the recent expansion of artificial areas closer
to natural habitats, confirming the existence of consistent links between
human-driven landscape changes and the presence and distribution of alien species
on coastal dunes. As in other human-modified landscapes, recent construction
directly on the dunes, road development and urban sprawl all seem to have pro-
moted alien invasion. In addition, human structures probably increase the propagule
pressure of alien species, although adjacency to other natural dune patches may also
facilitate their spread. In fact, patches closer to other natural patches tend to host a
higher average number of aliens, most likely because other invaded natural patches
Coastal Dune Vegetation Zonation in Italy ... 321

may facilitate alien establishment elsewhere in the landscape. Some aliens widely
present on Mediterranean coasts, such as Carpobrotus spp. and Agave americana,
with extensive vegetative growth, lead to the formation of extensive mats that may
facilitate their further expansion into other neighboring vegetation patches.
At local scale, results support the hypothesis that variation in abiotic conditions
can explain differences in invasibility, although propagule pressure seems to be the
main determinant of alien species occurrence and spread [12]. Thus, the major role
played by propagule pressure cannot be ignored, as it remains the main determinant
of alien species occurrence and spread even along a strong natural environmental
gradient [12].
In central Italy 8.5% (57 species) of the registered total vascular flora was alien,
but only few such species were considered invasive, such as the genus Carpobrotus
(C. acinaciformis/C. edulis), Agave americana, Erigeron canadensis (=Conyza
canadensis), Oenothera biennis and Ambrosia coronopifolia [3]. It is worth
emphasizing that most alien species came from America, but there are some sig-
nificant differences in alien distribution along the Adriatic and Tyrrhenian coasts.
Actually, significantly higher rates of alien species of African origin (mainly
tropical or from the Mediterranean fynbos) were found on the Tyrrhenian side,
while higher rates of extra-tropical American and Eurasian species were found on
the Adriatic. In fact, the genus Carpobrotus (from the Cape Region of South Africa)
and Agave americana (from Mexico), probably the most diffused invasive aliens
along the Tyrrhenian coast, were quite rare on the Adriatic. On the other hand,
Oenothera biennis and Ambrosia coronopifolia (from extra-tropical North
America) were the most diffused invasive aliens on the Adriatic coast.
The Genus Carpobrotus
Currently, invasion by Carpobrotus spp. constitutes one of the most severe threats
to the Mediterranean basin’s coastal plant communities [44, 50]. In Italy two dif-
ferent species have been identified, Carpobrotus edulis and C. acinaciformis,
although important taxonomical problems still remain unsolved [45]. This succulent
perennial taxon reaches high cover values and is able to form monospecific pop-
ulations which seriously endanger the survival of many dune habitats, from shifting
dunes with Ammophila arenaria to fixed dunes with Juniperus spp.
Adult Carpobrotus plants have been recognized to have an undesirable direct
impact on native plants in several areas of its secondary distribution range, with
negative effects on germination, survival, growth and reproduction of the victimized
native species [6, 15, 33, 50]. Several studies have shown that Carpobrotus invasion
ultimately affects patterns of native species diversity [27, 42], confirming that its
successful establishment probably operates through the replacement and exclusion
of native species, rather than coexistence. In conjunction with the decline in taxo-
nomic diversity, Jucker et al. [27] reported a parallel loss in functional diversity as
Carpobrotus abundance increased, this suggests that the invasive species is some-
how acting as a filter in the process of native community assembly by preferentially
excluding species with specific life-history traits. In particular, on the basis of a
plant-trait analysis, Carpobrotus appears to exclude closely related and ecologically
322 A. T. R. Acosta

similar taxa preferentially [27], and ruderal and nitrophilous species may substitute
for the typical native dune species [31]. Moreover, it has been shown how this alien
invasion influences the fine-scale assembly of native coastal communities, by
leading to a shift to randomness in community structure [40].
Carpobrotus has also been involved in many indirect negative effects on the
invaded native coastal ecosystems. Actually, it has been recognized as a major
driver of shifts in soil conditions and disruption of soil geochemical processes [33,
39]. In particular, it seems that in coastal dune pioneer habitats, Carpobrotus
invasion is likely to affect soil physico-chemical and biological processes [39].
These soil modifications could ultimately inhibit germination and affect the survival
of the specialized native dune species.
Beach Litter
Among various threats, marine litter seems to be one with a global distribution and
probably one of the most recognized pollution problems in the world’s oceans.
Litter washed ashore is called “beach litter” and is one of the most obvious signs of
marine litter pollution [37]. Sandy shores are generally considered important sinks
for floating debris, which, after stranding, generally becomes trapped in sand or
might be blown farther inland. In central Italy, plastic and polystyrene are the most
common litter items, in line with findings which confirm plastic as the most
widespread marine litter worldwide [37]. Moreover, there are differences in the
spatial distribution of the beach litter along the coastal zonation. A relationship
between beach-litter distribution and the sea-land gradient was also found: coastal
habitats on the upper beach, shifting dunes and the transition to the fixed dunes
were similarly threatened by beach litter, with the highest levels of litter accumu-
lation. On the other hand, fixed dune habitats were less influenced by beach litter. It
seems that fixed-dune habitats may act as a natural barrier protecting the inner part
of the dune system from marine litter dispersion [37]. Results on the various
potential sources of litter (urban areas, rivers and beach users) suggested that solid
pollution along coastal ecosystems was significantly related to the nearby presence
of urban areas. In fact, urban areas seem to be the main potential source of beach
litter, and rivers were probably the main, but not the only, carriers of litter items
from their origin to the marine environment [38].
Overall, accumulation of large amounts of beach litter is considered an aesthetic
loss to the value of the coastal environment, with potential adverse effects on tourist
activities. For this reason, cleaning beaches is one of the most important manage-
ment issues for coastal planning and is usually carried out by mechanical equip-
ment. Unfortunately, mechanical techniques allow no distinction between beach
litter and biological resources such as plants, animals, and organic debris naturally
deposited on shorelines and often removes the latter completely [37]. Therefore, the
presence of beach litter should be considered an indirect threat to biotic commu-
nities, promoting changes in the dune morphology and erosion processes. In par-
ticular, this threat is mostly relevant in those coastal areas characterized by intense
recreational use of beaches.
Coastal Dune Vegetation Zonation in Italy ... 323

4 Conclusions

Although the great diversity of Italian coastal dune habitats is now widely recog-
nized, many threats should also be emphasized. The consequences of these threats
are habitat and species loss, fragmentation, the final outcome sometimes being the
complete destruction of coastal dune systems. The remarkable “artificialization” of
coastal areas is probably one of the most serious concerns in Mediterranean coun-
tries. Coastal dune plant species are, to a large degree, highly specialized in coping
with the severe constraints imposed by the coastal environment along the sea-land
gradient: wind erosion, sand burial, aridity and incoherent soils. However, they are
not able to cope with the combination of human-induced environmental changes
outlined in this review. The high sensitivity of coastal plant species to such changes
leads to increasing concern about imminent biodiversity loss (both taxonomical and
functional) in coastal dune habitats. In particular, human impact is likely to affect the
most specialized coastal plants more strongly, while more generalist and alien plants
might be able to survive, leading to an overall biotic homogenization process among
communities. The critically poor conservation conditions make these habitats those
with the highest risk level; as such they require further research and special, and even
urgent, management measures in the hope of preserving the functionality and
uniqueness of these fragile ecosystems for future generations.

Acknowledgements This review is the outcome of the intensive work of many people. There are
so many that I could fill a long list. I would not like to forget anyone, but will just mention some of
them: Manuele Bazzichetto, Marta Carboni, Luisa Conti, Maria Laura Carranza, Silvia Del
Vecchio, Stefania Ercole, Francesca Izzi, Tommaso Jucker, Marco Malavasi, Irene Prisco,
Gianluca Poeta, Riccardo Santoro, Marta Gaia Sperandii and Angela Stanisci.

References

1. Acosta ATR, Stanisci A, Ercole S, Blasi C (2003) Sandy coastal landscape of the Lazio region
(Central Italy). Phytocoenologia 33(4):715–726
2. Acosta ATR, Ercole S, Stanisci A, Pillar VDP, Blasi C (2007) Coastal vegetation zonation
and dune morphology in some Mediterranean ecosystems. J Coast Res 236:1518–1524
3. Acosta ATR, Carranza ML, Di Martino L, Frattaroli A, Izzi CF, Stanisci A (2008) Patterns of
native and alien plant species occurrence on coastal dunes in Central Italy. In:
Tokarska-Guzik B, Brock JH, Brundu G, Child L, Daehler CC, Pyšek P (eds) Plant
Invasions: human perception, ecological impacts and management. Backhuys Publishers,
Leiden, pp 235–248
4. Acosta ATR, Carranza ML, Izzi CF (2009) Which habitat contributes best to plant species
diversity in coastal dunes? Biodivers Conserv 18:1087–1098
5. Acosta ATR, Jucker T, Prisco I, Santoro R (2013) Passive recovery of Mediterranean coastal
dunes following limitations to human trampling. In: Martinez LM, Gallego-Fernandez JB,
Hesp PA. Restoration of coastal dunes, pp 187–198. Springer, Cham
6. Affre L, Suehs CM, Charpentier S, Vilà M, Brundu G, Lambdon P, Traveset A, Hulme PE
(2010) Consistency on the habitat degree of invasion for three invasive plant species across
Mediterranean islands. Biol Invasions 12(8):2537–2548
324 A. T. R. Acosta

7. Bazzichetto M, Malavasi M, Barták V, Acosta ATR, Moudrý V, Carranza ML (2018)

Modeling plant invasion on Mediterranean coastal landscapes: an integrative approach using
remotely sensed data. Landscape Urban Plan 171:98–106
8. Bertness M, Callaway RM (1994) Positive interactions in communities. Trends Ecol Evol
9:191–193
9. Biondi E, Blasi C, Burrascano S, Casavecchia S, Copiz R, Del Vico E, Galdenzi D,
Gigante D, Lasen C, Spampinato G, Venanzoni R, Zivkovic L (2009) Italian Interpretation
Manual of the 92/43/EEC Directive Habitats, https://vnr.unipg.it/habitat/index.jsp
10. Biondi E, Brugiapaglia E, Allegrezza M, Ballelli M (1989) La vegetazione del litorale
marchigiano (Adriatico centro-settentrionale). Colloq Phytosociol 19:429–460
11. Carboni M, Thuiller W, Izzi F, Acosta ATR (2010) Disentangling the relative effects of
environmental versus human factors on the abundance of native and alien plant species in
Mediterranean sandy shores. Divers Distrib 16(4):537–546
12. Carboni M, Santoro R, Acosta ATR (2011) Dealing with scarce data to understand how
environmental gradients and propagule pressure shape fine-scale alien distribution patterns on
coastal dunes. J Veg Sci 22(5):751–765
13. Carboni M, Zelený D, Acosta ATR (2016) Measuring ecological specialization along a
natural stress gradient using a set of complementary niche breadth indices. J Veg Sci 27
(5):892–903
14. Chytrý M, Pyšek P, Wild J, Pino J, Maskell LC, Vilà M (2009) European map of alien plant
invasions based on the quantitative assessment across habitats. Divers Distribs 15(1):98–107
15. Conser C, Connor EF (2009) Assessing the residual effects of Carpobrotus edulis invasion,
implications for restoration. Biol Invasions 11(2):349–358
16. Ciccarelli D (2015) Mediterranean coastal dune vegetation: are disturbance and stress the key
selective forces that drive the psammophilous succession? Estuar Coast Shelf S 165(5):247–
253
17. Conti L, de Bello F, Lepš J, Acosta ATR, Carboni M (2017) Environmental gradients and
micro-heterogeneity shape fine-scale plant community assembly on coastal dunes. J Veg Sci
28:762–773
18. Doing H (1985) Coastal foredune zonation and succession in various parts of the world.
Vegetatio 61:65–75
19 Del Vecchio S, Prisco I, Acosta ATR, Stanisci A (2015) Changes in plant species composition
of coastal dune habitats over a 20-year period. AoB PLANTS 7:plv018
20. European Commission (2013) Interpretation Manual of European Union Habitats. EUR28
21. Feola S, Carranza ML, Schaminée JHJ, Janssen JAM, Acosta ATR (2011) EU habitats of
interest: an insight into Atlantic and Mediterranean beach and foredunes. Biod Conser 20
(7):1457–1468
22. Fenu G, Carboni M, Acosta ATR, Bacchetta G (2013) Environmental factors influencing
coastal vegetation pattern: new insights from the Mediterranean Basin. Folia Geobot 48
(4):493–508
23. Forman RTT, Godron M (1986) Landscape Ecology. Wiley, New York
24. Gerdol R, Piccoli F (1984) Sand dune vegetation in the Po Delta (N Italy). Ecologia
Mediterranea 10(3–4):119–131
25. Janssen JAM, Rodwell JS, Criado MG, Gubbay S, Arts GHP, Haynes T, Nieto A, Sanders N,
Landucci F, Loidi J, Ssymank A, Tahvanainen T, Valderrabano M, Acosta ATR,
Aronsson M, Arts G, Attorre F, Bergmeier E, Bijlsma RJ, Bioret F, Biţă-Nicolae C,
Biurrun I, Calix M, Capelo J, Čarni A, Chytrý M, Dengler J, Dimopoulos P, Essl F, Gardfjell,
Gigante D, Giusso del Galdo G, Hájek M, Jansen F, Jansen J, Kapfer J, Mickolajczak A,
Molina JA, Molnár Z. Paternoster D, Piernik A, Poulin B, Renaux B, Schaminée JHJ,
Šumberová K, Toivonen H, Tonteri T, Tsiripidis I, Tzonev R, and Valachovič M (2016)
European Red List of Habitats. Part 2. Terrestrial and freshwater habitats. European Union.
https://doi.org/10.2779/091372.
Coastal Dune Vegetation Zonation in Italy ... 325

26. Genovesi P, Angelini P, Bianchi E, Dupré E, Ercole S, Giacanelli V, Ronchi F, Stoch F

(2014) Specie e habitat di interesse comunitario in Italia: distribuzione, stato di conservazione
e trend. ISPRA, Serie Rapporti, 194/2014
27. Jucker T, Carboni M, Acosta ATR (2013) Going beyond taxonomic diversity: deconstructing
biodiversity patterns reveals the true cost of iceplant invasion. Divers Distrib 19:1566–1577
28. Malavasi M, Santoro R, Cutini M, Acosta ATR, Carranza ML (2013) What has happened to
coastal dunes in the last half century? A multitemporal coastal landscape analysis in Central
Italy. Landscape Urban Plan 119:54–63
29. Malavasi M, Carboni M, Cutini M, Carranza ML, Acosta ATR (2014) Landscape
fragmentation, land-use legacy and propagule pressure promote plant invasion on coastal
dunes: a patch-based approach. Landscape Ecol 29(9):1541–1550
30. Malavasi M, Conti L, Carboni M, Cutini M, Acosta ATR (2016) Multifaceted analysis of
patch-level plant diversity in response to landscape spatial pattern and history on
Mediterranean dunes. Ecosystems 19:850–864
31. Malavasi M, Santoro R, Cutini M, Acosta ATR, Carranza ML (2016) The impact of human
pressure on landscape patterns and plant species richness in Mediterranean coastal dunes.
Plant Biosyst 150(1):73–82
32. Murru V, Marignani M, Acosta ATR, Cogoni A (2018) Bryophytes in Mediterranean coastal
dunes: ecological strategies and distribution along the vegetation zonation. Plant Biosyst 152
(5):1141–1148
33. Novoa A, González L, Moravcová L, Pyšek P (2013) Constraints to native plant species
establishment in coastal dune communities invaded by Carpobrotus edulis: implications for
restoration. Biol Conserv 164:1–9
34. Prisco I, Carboni M, Acosta ATR (2012) VegDunes – a coastal dune vegetation database for
the analysis of Italian EU habitats. Biodiv Ecol 4:191–200
35. Prisco I, Carboni M, Jucker T, Acosta ATR (2016) Temporal changes in the vegetation of
Italian coastal dunes: identifying winners and losers through the lens of functional traits.
J Appl Ecol 53(5):1533–1542
36. Prisco I, Stanisci A, Acosta ATR (2016) Mediterranean dunes on the go: Evidence from a
short-term study on coastal herbaceous vegetation. Estuar Coast Shelf S 182:40–46
37. Poeta G, Battisti C, Acosta ATR (2014) Marine litter in Mediterranean sandy littorals: spatial
distribution patterns along central Italy coastaldunes. Mar Poll Bull 89:168–173
38. Poeta G, Conti L, Malavasi M, Battisti C, Acosta ATR (2016) Beach litter occurrence in
sandy littorals: the potential role of urban areas, rivers and beach users in central Italy. Estuar
Coast Shelf S 181:231–237
39. Santoro R, Jucker T, Carranza ML, Acosta ATR (2011) Assessing the effects of Carpobrotus
invasion on coastal dune soils. Does the nature of the invaded habitat matter? Community
Ecol 12(2):234–240
40. Santoro R, Jucker T, Carboni M, Acosta ATR (2012) Patterns of plant community assembly
in invaded and non-invaded communities along a natural environmental gradient. J Veg Sci
23(3):483–494
41. Santoro R, Jucker T, Prisco I, Carboni M, Battisti C, Acosta ATR (2012) Effects of trampling
limitation on coastal dune plant communities. Envir Manag 49:534–542
42. Santoro R, Carboni M, Carranza ML, Acosta ATR (2012) Focal species diversity patterns can
provide diagnostic information on plant invasions. J Nat Conserv 20(2):85–91
43. Sburlino G, Buffa G, Filesi L, Gamper U, Ghirelli L (2013) Phytocoenotic diversity of the
N-Adriatic coastal sand dunes – the herbaceous communities of the fixed dunes and the
vegetation of the interdunal wetlands. Plant Sociol 50(2):57–77
44. Sintes T, Moragues E, Traveset A, Rita J (2007) Clonal growth dynamics of the invasive
Carpobrotus affine acinaciformis in Mediterranean coastal systems: a non-linear model. Ecol
Model 206(1):110–118
45 Suehs CM, Affre L, Médail F (2004) Invasion dynamics of two alien Carpobrotus (Aizoaceae)
taxa on a Mediterranean island: I. Genetic diversity and introgression. Heredity 92(1):31–40
326 A. T. R. Acosta

46. Stanisci A, Acosta ATR, Ercole S, Blasi C (2004) Plant communities on coastal dunes in
Lazio (Italy). Annali Di Botanica 4:115–128
47. Suehs CM, Affre L, Médail F (2004) Invasion dynamics of two alien Carpobrotus
(Aizoaceae) taxa on a Mediterranean island: II reproductive strategies. Heredity 92(6):550–
556
48. Vagge I, Biondi E (1999) La vegetazione delle coste sabbiose del Tirreno settentrionale
italiano. Fitosociologia 36(2):61–95
49. van der Maarel E (2003) Some remarks on the functions of European coastal ecosystems.
Phytocoenologia 33:187–202
50. Vilà M, Tessier M, Suehs CM, Brundu G, Carta L, Galanidis A, Lambdon P, Manca M,
Médail F, Moragues E, Traveset A, Troumbis AY, Hulme PE (2006) Local and regional
assessments of the impacts of plant invaders on vegetation structure and soil properties of
Mediterranean islands. J Biog 33(5):853–861
51. Wilson B, Sykes MT (1999) Is zonation on coastal sand dunes determined primarily by sand
burial or salt spray? A test in New Zealand dunes. Ecol Lett 2:233–236
Constraints on Evergreen Broad-Leaved
Forests in the Southeastern United
States

Elgene O. Box and Kazue Fujiwara

Abstract In a humid, mild-winter warm-temperate climate, as occurs across most

of the southeastern US coastal plain, one would normally expect to find “laurel
forests” dominated by temperate-zone evergreen tree taxa, as in other
warm-temperate regions. Instead, on the sandy uplands of the geologically young
coastal plain, one finds a topogenic, landscape-scale mosaic of wetlands, forests,
open woodlands and scrub, some of it maintained by recurring fires. As a result, the
potential over the coastal plain is not extensive forest but rather, where soil,
topography and fire permit, a slow progression to woods dominated by coriaceous
or harder-leaved evergreen broad-leaved trees, with somewhat open canopies and a
greater role for deciduous trees, even at maturity. Colder winters to the north
constrain most evergreen forest types, but there are also subtler, non-climatic
constraints. The purpose of this paper is to describe briefly the main evergreen
broad-leaved forest types and their dynamics, and to evaluate the climatic and
non-climatic factors that constrain their distributions, especially within the context
of local climatic warming and drying.

Keywords Warm-temperate climate Humic sand Topogenic vegetation





Limiting temperatures Laurophyll Sclerophyll Evergreen forest succession

Fire cycle

1 Introduction

Spreading live oaks (Quercus virginiana), dripping with Spanish moss (Tillandsia
usneoides) and complemented by fragrant but primitive-looking Southern magno-
lias (Magnolia grandiflora), all evergreen, constitute the character forests and
settlement vegetation of the original American South, i.e. the US southeastern

E. O. Box (&)
Department of Geography, University of Georgia, 210 Field Street, Athens, GA 30602, USA
K. Fujiwara
Yokohama City University, Yokohama, Japan

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 327
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_17
328 E. O. Box and K. Fujiwara

coastal plain, which extends from southern New Jersey to south Florida and west to
coastal Texas. Although other forests may be evergreen and broad-leaved, the term
‘evergreen broad-leaved forest’ is commonly used for forests dominated by
extra-tropical evergreen broad-leaved trees, which constitute the “zonal” vegetation
of humid warm-temperate climates, generally understood as having no regular dry
season and mild but not frost-free winters [12, 15, 16], cf [49, 58, 66, 71, 74, 121,
128, 136, 160–162]. Such forests are found in humid warm-temperate to subtropical
parts of East Asia, southern Brazil, eastern Australia, and New Zealand. In the
southeastern US, on the other hand, the potential for characteristic evergreen
broad-leaved forests is restricted to the warm-temperate climate of the coastal plain
and is clearly separated from the subtropical forests of southernmost Florida by
intervening non-forest areas and by the cessation of frost [32].
That part of the world most similar physiographically to eastern North America
is East Asia [6, 53], cf [54], where humid warm-temperate areas originally had
“laurel forests” with mesomorphic, glossy, shade-tolerant laurophylls, called “lu-
cidophylls” by Kira [75], cf [150]. These warm-temperate forests extended south-
ward from the Tokyo area, southernmost Korea and the Chang Jiang (Yangtze)
estuary near Shanghai, through subtropical southeastern China [142–144, 174]; the
Okinawa islands [102]; and northern Taiwan [26, 65, cf 145]. Most of this forest
was lost long before ecological studies began, and the secondary deciduous forests
were interpreted until recently as the potential natural vegetation. Laurel forests
were finally recognized as the true potential natural vegetation, based on small
remnants of original or very similar forest that remained around temples, shrines,
and even some farmhouses [101, 104, 106]). Such laurel forests are dominated by
Persea (= Machilus, Lauraceae) and laurophyllous Quercus and Castanopsis
(Fagaceae), along with laurophylls from families such as Theaceae and
Myrsinaceae [44, 142, 144, 174], [46], cf [118, 119]. It was also recognized by Kira
[72], however, that deciduous forests could persist within the warm-temperate zone
in some interior parts of Japan (with colder winters). Such forests were called
‘warm-temperate deciduous forests’, occur also in China, and were treated in detail
by Box and Fujiwara [16].
The US southeastern coastal plain is one of the world’s largest regions in which
the main substrate, geologically young sand, does not match the prevailing climate,
which would normally produce zonal red-yellow podzolic clay soils. The coastal
plain is delimited inland by the Fall Line, a slight escarpment (at which European
explorers met the first river rapids) which still presents areas of nutrient-poor
sandhills left from old oceanic shorelines. Most of the coastal plain is covered by
relatively fine sand with a significant humic content (over limestone in Florida).
This is not a bad substrate for plant growth but, in combination with the low-lying
topography and warm climate, produces sharp vertical gradients between dry
conditions above and wet, sometimes saturated conditions not far below the surface,
where groundwater may be high, at least seasonally. This translates also into steep
horizontal gradients, between distinctly hydromorphic vegetation in lower-lying
areas and distinctly xeromorphic vegetation on areas only a few meters or only
decimeters higher [122, 168, 169].
Constraints on Evergreen Broad-Leaved Forests ... 329

As in East Asia, the original forests of the US South were lost early, to plantation
agriculture and to Northern shipbuilding [173]. In the US there was no tradition of
preserving original forest, but some large evergreen broad-leaved trees, usually live
oaks, have remained where agriculture was not possible, suggesting that such trees
may have been important in the original forest. Due to recurring fire, both natural
and man-induced (e.g. [3, 79]), forests probably never covered the coastal plain
even where topography permitted. As a result of topography, substrate and fire, the
“natural” vegetation of the southeastern coastal plain is a mosaic of deciduous
forest, pinelands, sclerophyll live-oak woods and scrub, various kinds of wetlands,
and agricultural land covers—albeit with increasing wooded cover throughout the
1900s due to land abandonment, except in Florida [127, cf 141, 60]. Some areas of
evergreen broad-leaved forest were recognized fairly early (e.g. [48, 59, 84]), but
most current forests on the coastal plain involve a variable mix of evergreen and
deciduous broad-leaved trees called ‘Southern Mixed Hardwoods’ [126], which
was eventually accepted as the potential climax forest type on the coastal plain (cf
[80, 108]).
With global warming, this and more completely evergreen forest types could
perhaps expand northward. What factors would constrain this expansion? Critical
factors include not only the role of fire, nutrient requirements, and substrate limi-
tations or affinities but also attributes of potential main structural tree species. Much
information on evergreen broad-leaved taxa and woods is well known to local
naturalists but needs to be published and examined within the context of geographic
zonation, environmental constraints, and global warming. The purpose of this paper
is to describe briefly the structure and dynamics of the main evergreen broad-leaved
forest types of the southeastern US, and to examine constraints on their distributions
and their prospects under global warming.

2 Data and Methods

The basic methodology of this study is to evaluate environmental conditions near

the geographic limits of native evergreen broad-leaved forests and woodlands, and
to explore the potential dynamics of these forests under environmental change,
especially global warming and perhaps drying. Constraints on these forests are both
climatic and substrate-related, so data are needed for both aspects, as well as for
representative vegetation stands.
For vegetation, sites were sought to represent the full variation in evergreen
broad-leaved forest types. This involved fieldwork in Florida, Georgia, and the
Carolinas but extending west to the Hill Country of central Texas [14]. Forest
stands were described by the relevé method of Braun-Blanquet [19], see [45], cf
[112]. This method provides complete three-dimensional descriptions of stand
composition and structure, and species cover and clustering (sociability), by layer.
A relevé quadrat in a southeastern forest is typically about 25  25 m and requires
about one hour to execute. The data-base for evergreen forests was supplemented,
330 E. O. Box and K. Fujiwara

beginning in 1988, by relevés obtained during a three-year Japan-funded project to

inventory the vegetation of eastern North America [105]. The resulting combined
data-base involves about 100 relevés for evergreen broad-leaved or related forests.
Climatic data include long-term mean monthly temperature and precipitation at
sites across the Southeast, but also extreme temperature minima. In particular, the
‘absolute minimum temperature’, i.e. the coldest temperature that can occur, was
suggested quite early as an important limiting factor, especially for evergreen
broad-leaved trees [20, 131, 132]. Even where measured, though, records of
absolute minima are often not long enough to capture the true absolute minimum.
The global geography of absolute minima is quite different from that of cold-month
means, due to land-sea configurations, monsoon systems, etc. (Box 1995a). Thus
one cannot simply assume a predetermined offset from cold-month mean temper-
ature (as done by [125], cf [89]). On the other hand, reasonable estimates of
absolute minima can be made based on the local difference between long-term
absolute minima and the cold-month mean (program TXTRAP, cf [13,9]). The site
values of absolute minimum temperature used herein are the lower of the measured
value or the estimate.
Topography and substrate properties come from existing maps (e.g. [40, 41, 156,
67, 120]) but were corroborated by data recorded at the relevé sampling sites. The
physical properties recorded at data sites are qualitative but include topographic
position, general soil texture, and soil humic and moisture content (at sampling
time).

3 The Coastal Plain Environment

Winter temperatures on the coastal plain involve January means near or only a bit
below 10°C (higher in Florida), usually light frost, but absolute minima generally not
below about −15°C (see Table 1). This makes the coastal plain a warm-temperate
climate, not “subtropical” as erroneously claimed by various unfortunate variants of
Köppen’s terminology (cf [12], p. 11). In addition to mean temperatures, Table 1
also shows, for sites across the Southeast, the absolute minimum temperature, the
number of consecutive warm months (mean at least 10°C); average precipitation for
the year and the driest month; and the climatic moisture index (i.e. annual precipi-
tation divided by potential evapotranspiration). The sites in Table 1 are listed from
north to south, in five sub-regions. The Interior South includes inland sites on the
piedmont or uppermost coastal plain, with winters that permit only deciduous for-
ests. The Warm-Temperate Transition includes sites with absolute minima only a bit
below −18°C. Sites in the true Warm-Temperate region generally have cold-month
means of at least 8°C and absolute minima above or not far below −15°C. Most of
Florida is also warm-temperate but shows higher mean minima and all 12 months
above 10°C. Finally, data are also shown for subtropical Miami, where the
cold-month mean is above 18°C and the absolute minimum is −2.8°C, which per-
haps occurred only once.
Constraints on Evergreen Broad-Leaved Forests ... 331

Table 1 Climatic data for representative sites across the US South

Abbreviations are as follow: Lat = latitude, Tmax = mean temperature of the warmest month, Py =
average annual precipitation, Elev = elevation (m), Tmin = mean temperature of the coldest month,
Pmin = avg. precip. of driest month, Tabmin= absolute minimum temperature, Lwarm = number
of months above 10°C (mean), MIy = Py/PETy, where PETy = avg. annual potential
evapotranspiration. All temperatures are in °C and all precipitation amounts in millimeters.

Temperatures in the Southeast are compared in Table 2 with those at comparable

sites in East Asia and Europe. Milder winters extend farther north in Europe, as on
all continental west sides. For east sides, at comparable latitudes and geographic
position, mean winter temperatures are generally higher in eastern North America
but absolute minima are lower than in East Asia [9]. In East Asia the monsoon
system insures consistently low winter temperatures, but the east-west orientation of
most mountain ranges blocks the extremely cold air masses from further north.
North America has much less winter-monsoon effect, since it has the Gulf of
Mexico to moderate winter cold; and it has no east-west mountains that affect the
east. So winter means are higher and extremes are lower, especially in the south-
eastern US. It is for this reason that evergreen broad-leaved forests extend further
north in East Asia (see map in Fujiwara and Box [46], p. 305; cf [137]).
332 E. O. Box and K. Fujiwara

Table 2 Minimum temperatures at comparable humid sites around the Northern temperate zone

The sites are listed from north to south, are mostly near-coastal (except with asterisk), and are
paired as far as possible by comparable geographic position (i.e. distance from a coastline). The
top number at each site is the coldest-month mean temperature; the number directly under this is
the absolute minimum temperature, i.e. the lowest ever measured (or expected, indicated by ‘e’,
see Data and Methods). Temperatures for some additional US sites are given in Table 1.
Constraints on Evergreen Broad-Leaved Forests ... 333

Fig. 1 Climatic Moisture Balance in the Southeastern USA. This moisture balance is estimated as
average annual precipitation divided by annual potential evapotranspiration (PET). By this
measure, the wettest areas on the southeastern coastal plain are along the northern coast of the Gulf
of Mexico and the mid-Atlantic coast (New Jersey to North Carolina). The driest areas are in
central to south-central Florida and in Texas, where precipitation decreases rapidly westward as
Gulf moisture becomes less available to onshore winds. Note that most of South Carolina, southern
Georgia, and northern Florida is also relatively dry, as is the area of the sharp gradient from
Louisiana into east Texas. Annual PET is estimated in the southeastern USA by the method of
Holdridge [64], which appears to be relatively accurate at warm-temperate latitudes [8]
Climatic Moisture Balance in the Southeastern USA. This moisture balance is estimated as average
annual precipitation divided by annual potential evapotranspiration (PET). By this measure, the
wettest areas on the southeastern coastal plain are along the northern coast of the Gulf of Mexico
and the mid-Atlantic coast (New Jersey to North Carolina). The driest areas are in central to
south-central Florida and in Texas, where precipitation decreases rapidly westward as Gulf
moisture becomes less available to onshore winds. Note that most of South Carolina, southern
Georgia, and northern Florida is also relatively dry, as is the area of the sharp gradient from
Louisiana into east Texas. Annual PET is estimated in the southeastern USA by the method of
Holdridge [64], which appears to be relatively accurate at warm-temperate latitudes [8]

The other major climatic factor for vegetation is available moisture, depicted
here by the climatic moisture index (see above), as mapped in Fig. 1. In Louisiana
and Arkansas, and westward into eastern Texas and Oklahoma, there is a sharp
westward reduction in rainfall as less moisture becomes available to prevailing
winds coming over land rather than the Gulf of Mexico. The pattern in Fig. 1 also
suggests greater moisture stress in southern Georgia, South Carolina and especially
central to south-central Florida, where much of the landscape is indeed more open.
Forests in Florida are often topographically circumscribed and referred to as
334 E. O. Box and K. Fujiwara

hammocks. Precipitation in the Southeast is highly variable from year to year, and
relative to long-term seasonal averages. This is due partly to the inherent variability
in the shape and position of the jet stream over North America, but perhaps even
more due to El Niño events.
The southeastern climate is conditioned greatly by the surrounding ocean and
Gulf of Mexico, which moderate winter temperatures (Fig. 2). As a result, the
warm-temperate climate does not reach far inland, beyond the coastal plain, except
through some of the broader river valleys, especially the Mississippi. The
warm-temperate climate also does not extend to subtropical south Florida, where
frost essentially disappears and forests are dominated rather suddenly by subtropical
taxa [32], cf [93, 115]. Compared with other regions of humid warm-temperate
climate, that of the southeastern US is more subcontinental, i.e. it shows relatively
high temporal variability in both temperature and precipitation. The effect of this
variability on vegetation is exacerbated by topography and substrate.
The substrate across the southeastern US coastal plain is largely humic fine sand,
dotted with deep sterile sandhills representing old dune systems; moist depressions,
often with shallow peat accumulations; and broad moist bottomlands along

Fig. 2 Winter Temperatures in the Southeastern USA. The map shows the mean temperature of the
coldest month (Tmin, in January), in degrees Celsius. Greller [56] also mapped the 5ºC isotherm for
mean nighttime minimum in January (running between the 10°C and 14ºC Tmin isotherms) and
related this to the northern boundary of his evergreen broad-leaved forest zone. At global scale, even
short-term exposure to temperatures below about −15ºC has been identified as potentially lethal to
temperate-zone evergreen broad-leaved tree species (e.g. Sakai and Larcher [134]), and this isotherm
generally follows the poleward limits of warm-temperate evergreen broad-leaved forests worldwide
(e.g. Woodward [173]). In the Southeast, this −15ºC isotherm for absolute minimum temperature
runs near the 10ºC isotherm for Tmin. Harder-leaved trees, such as live oaks and southern magnolia,
may tolerate short-term exposure to slightly lower temperatures, perhaps as low as about −18ºC, a
value which may extend even to areas near the Gulf of Mexico coastline (e.g. Tallahassee)
Constraints on Evergreen Broad-Leaved Forests ... 335

tributaries of larger rivers (e.g. [27, 67, 169, 96, 120]). The lower-lying areas may
have swamp forests (Taxodium, Nyssa, Chamaecyparis, etc.), fire-maintained
‘pocosin’ shrublands [140], and other wetlands or near-wetlands. Mesomorphic
laurophyll “bay” forests may occur in moist depressions with peat accumulations,
while xeromorphic scrub (with Pinus, deciduous and evergreen Quercus, etc.)
dominates the highest, driest (and nutrient-poorest) deep-sand areas (e.g. [27]).
A few areas do have relatively extensive richer soils. One is the loess hills of
Louisiana, with silt-loam soils over loess [146], cf [35]. Another is the Apalachicola
bluff area of western Florida, where limestone is near the surface and nutrient
conditions may be dramatically better [59]. Such areas can support more meso-
morphic laurophyll trees as well as more nutrient-demanding deciduous forests
[36], cf [108]. All of this means that one must attempt to distinguish immediately
between ‘zonal’ vegetation, i.e. that corresponding to the humid warm-temperate
climate, on the one hand, and ‘azonal’ vegetation determined by unusual edaphic,
topographic, or pyrogenic factors. On the southeastern coastal plain, azonal vege-
tation may cover a larger area than does the potential zonal vegetation. Such
topographic contrasts are especially apparent in the climatically drier parts of the
region (central Florida through the Carolinas) but occur even in the wettest areas as
well (western Florida through southern Louisiana).
Fire, ignited by lightning strikes or by humans, is the joker in all this and is
fostered by the often dry conditions above the surface [86, 117, 25, 78]. Climate
determines the amount of fuel available and the length and severity of dry periods,
but weather regulates the moisture content of dead fuels and the ignition and spread
of wildfires. Wildfires burn off remaining or incipient broad-leaved forests on the
uplands and maintain their replacing pine-palmetto savannas on lower flats [27].
Pollen analyses suggest that fire and the fire-adapted pine vegetation, including the
characteristic Pinus palustris-Aristida stricta open woodlands, expanded rapidly to
cover large areas of the coastal plain about 5000–6000 years ago [166, 167].

4 Evergreen Broad-Leaved Tree Taxa

The evergreen broad-leaved woody flora of the southeastern US is not large, as

compared with comparable parts of East Asia (cf [46]). This may be due to bio-
geographic history and the small size, throughout evolutionary time, of the
near-coastal, milder-winter region in the southeastern US. The evergreen
broad-leaved trees native to the southeastern US are listed in Table 3 and fall into
two distinctly different groups, namely trees with: 1) hygromorphic, shade-tolerant
dark green laurophylls (Lorbeerblätter, [128], cf [73, 74, 157], as in laurel forests
elsewhere), or 2) distinctly harder, xeromorphic leaves with a wide range of shade
tolerance (but less than true laurophylls). The main laurophylls are the three “bay”
species: Persea borbonia (Lauraceae, “red bay”, plus almost identical P. palustris),
Magnolia virginiana (“sweet bay”), and Gordonia lasianthus (Theaceae, “loblolly
bay”), so called because they are slightly aromatic and were used by early settlers as
336 E. O. Box and K. Fujiwara

Table 3 Main evergreen broad-leaved woody taxa in the temperate US South

Laurophylls are darker, glossy green and shade-tolerant, as in the “laurel forests” of East Asia,
southern Brazil—and tropical rainforests. Coriaceous leaves are pliable but usually thicker, and are
usually somewhat less shade-tolerant. Sclerophylls are hard (rigid), resist water loss better, and are
usualy not shade-tolerant, although Q. virginiana does regenerate well under its own canopy.
Lauro-sclerophylls are an unusual combination of hardness and shade-tolerance, both of which
lead to slower growth rates.
Constraints on Evergreen Broad-Leaved Forests ... 337

cooking substitutes for ‘bay leaves’ (Laurus nobilis). There are also two prominent
but somewhat puzzling lauro-sclerophylls, Magnolia grandiflora and Ilex opaca
(much like European I. aquifolium), which have quite hard but dark, shade-tolerant
leaves, thus grow slowly, but can reach canopy height with time.
The most important and widely occurring tree is Quercus virginiana (live oak),
whose spreading crown, long lifespan (to 300 years or more), and intermediate
shade tolerance permit it to grow up and dominate, where topography and history
permit, across most of the coastal plain. Q. virginiana is evergreen, not
semi-evergreen, and dominates forests that are warm-temperate, not subtropical or
only maritime, as characterized erroneously by Thyroff et al. [153] in an otherwise
useful documentation of live oak’s phenotypic plasticity (e.g. sclerophyllous to
semi-sclerophyllous; cf [111]). Q. virginiana may dominate from maritime forests
([7, 17, 170], pp 185) to broad coastal-plain uplands (e.g. [115, 170], pp 188) and
west to the more continental but still warm-temperate Hill Country of central Texas,
where it hybridizes with very similar Q. fusiformis [158], cf [113, 152]. Q. vir-
giniana also extends into south Florida, where it has long been seen as a “suc-
cessional” tree in subtropical hammocks (e.g. [1], and even into western Cuba. This
may be an example of the contention by Loehle [92] that trees often grow well at
their southern range limits and are limited there by competition rather than
physiology.
Two other important oaks may deserve the descriptor ‘semi-evergreen’, namely
Q. laurifolia, a bottomland species that does lose many leaves in the northern part
of its range (near the Fall Line); and Q. nigra, a widespread “tardily deciduous” tree
that may become essentially evergreen at the southern end of its range in central
Florida [68, 113]. The other important evergreen oak is Quercus hemisphaerica, a
fast-growing, mainly upland secondary species (red-oak group), which was long
confused with Q. laurifolia (white-oak group) [42], cf [154, 155]. The most scle-
rophyllous oaks are small-tree Q. geminata and shrubby Q. myrtifolia, which may
dominate clearly on deeper, nutrient-poor sandhills and sandy flats. Other
arborescent evergreen hollies include coriaceous Ilex vomitoria and I. cassine, both
of which can be quite abundant locally. Somewhat similar ecologically are the
coriaceous-leaved understorey trees Vaccinium arboreum and Lyonia ferruginea.
These evergreen broad-leaved species and their ecology are described in various
local identification manuals, the best of which are by Godfrey [52], Kurz and
Godfrey [82], Hunt [68] and Muller [113] (but see also [29, 38], and Lakele and
Wunderlin [87]).
The occurrence of distinctly hygromorphic and xeromorphic evergreens results
from the strong moisture gradients in the sandy landscape. The relative shade
tolerance and growth rates of laurophyll and sclerophyll evergreens, and deciduous
trees, are significant for vegetation dynamics on sufficiently mesic sites, where all
three forms can occur. The laurophylls are the true counterparts of the “laurel
forest” taxa of East Asia and elsewhere, and are found on the US southeastern
coastal on sites that remain moist or at least mesic throughout the summer, such as
shallow depressions (see [170], p 74). These laurophyll species are more
shade-tolerant than deciduous species or other evergreen species, and thus can form
338 E. O. Box and K. Fujiwara

permanent “bay forests” where permanent moisture availability permits. The oaks
and other broad-leaved evergreens grow more slowly (except Q. hemisphaerica,
e.g. [51]) and are potential dominants on the humic sands across most of the coastal
plain. The most sclerophyllous oaks (Q. geminata and shrubby Q. myrtifolia)
dominate on deep, nutrient-poor sandhills (e.g. [27]).

5 Evergreen Broad-Leaved Forest Types

As abandoned upland areas grow back in humid warm-temperate climates, in the

absence of fire, they pass through a succession from initial colonizers, to tall
grasses, then light-demanding pines, then more shade-tolerant deciduous trees,
which can invade once the pines have provided a less harsh, shadier, more humid
microclimate. Where winters are not too cold, the deciduous trees can then be
replaced at least partly by broad-leaved evergreen trees. The basic hypothesis is
that, eventually, shade-tolerant laurophyll trees or other evergreen broad-leaved
trees such as oaks, with intermediate shade tolerance, will replace deciduous trees
with higher light requirements, as in East Asia [11, 47]. The logic of shade toler-
ance as a mechanism of succession in temperate forests was suggested early by
Cowles [31] and explained further by Bazzaz [4], but without reference to ever-
green broad-leaved trees, which do not occur in their northern areas of the USA.
The shade-tolerance theory of succession was extended to include laurophyll trees
through experience in East Asia, especially the southern half of Japan, where the
replacement of deciduous forests (long thought to be the potential) by evergreen
laurel forests was documented by Miyawaki [100], albeit without explaining
shade-tolerance as the mechanism (cf [103, 107]). Evergreen trees have several
advantages over deciduous trees: earlier shoot growth and photosynthesis in spring
(e.g. [24]), other competitive advantages in forest canopies (e.g. [37, 62, 116],
Cavender-Bares et al. [23], [124]), greater shade tolerance, even in evergreen oaks
(e.g. Thyroff et al. [152]); the ability to regenerate under deciduous canopies as well
as under their own shade; and permanent occupation of canopy space. This is true
to some extent also for more xeromorphic Q. virginiana, which does regenerate
under its own, less dense canopy (many personal observations).
Although the coastal-plain climate suggests evergreen broad-leaved forests, this
potential was not recognized until recently by most American botanists and ecol-
ogists. Syntheses describing the biomes and other vegetation regions of North
America have treated the southeastern coastal plain variously as “longleaf-pine
forest” [135], “southeastern pine forest” [139], “southeastern evergreen forest” [18],
“southern mixed forest” [80], “pine forest” (e.g. [39]), “warm-temperate mixed
forest” [91] or “Vegetation of the Southeastern Coastal Plain” [27]. The term
“Southern mixed hardwoods” was enshrined by Quarterman and Keever [126] to
describe potential dominance by a varying mix of deciduous and evergreen
broad-leaved tree species, plus conifers (cf [108]). Only a few writers, mainly from
outside the US, recognized the region as potential “evergreen broad-leaved forest”
Constraints on Evergreen Broad-Leaved Forests ... 339

(e.g. [137]). Probably the first American to describe the evergreen broad-leaved
forest zone as such was Greller [55, 56, 57], who expanded Braun’s [18] region-
alization of the eastern US forests to include a semi-evergreen transition and a zone
of potential evergreen broad-leaved forest.
As a first step in clarifying the position of evergreen broad-leaved forests in the
Southeast, two attempts were made by the present authors to identify basic types of
such forests and related vegetation. The first attempt (see Table 4, from [14])
recognized five types of evergreen broad-leaved forest on humic sand, as well as
laurophyll “bay” forests in scattered moist depressions, oak-dominated floodplain
forests, and more open Q. geminata woodlands on deep sterile sand. The second
attempt (see Table 5, from Fujiwara and Box [46]) was based on phytosociological
tablework and recognized types by actual or potential dominant species, mainly Q.
virginiana but also successional Q. hemisphaerica, harder-leaved Q. geminata on
sterile sand, laurophylls in moist depressions, and Q. laurifolia in bottomlands and
floodplains. More floristic detail and consideration of zonal and pedogenic position
for mostly the same types was provided later by Greller [57], who also related his
types to those of other authors. It is not possible to treat all these types here, so we
take six main types, each of which is described briefly below, with some consid-
eration of dynamic status: zonal (climatic climax), topogenic, edaphic, or pyro-
genic. Examples are given in the Appendix (Relevé Tables A.1, A.2, A.3, A.4, A.5

Table 4 Evergreen broad-leaved forest types identified by Box and Fujiwara [14]

All taxa listed are evergreen except: Carya, Liquidambar, Liriodendron, Nyssa, Q. michauxii and
coniferous Taxodium, plus (mainly in understoreys) Acer, Fraxinus, Ulmus, Rosaceae and
Rhamnaceae. The Carya species is mainly C. glabra, but others may occur on the more deciduous
uplands. The main Nyssa species is N. biflora (plus N. sylvatica in drier uplands), but N. aquatica
may dominate in quite hydric situations. Taxodium is mainly T. distichum, but shorter T.
ascendens also occurs in some ponds.
340 E. O. Box and K. Fujiwara

Table 5 Evergreen broad-leaved woody communities identified by Fujiwara and Box [46]

All taxa listed are evergreen except: Quercus nigra, Q. michauxii, small tree Q. laevis; Q. texensis
and Q. sinuata var. breviloba in Texas; some Vaccinium spp., Acer rubrum, Carya glabra,
Liquidambar, Nyssa biflora, Lyonia mariana, subtropical Lysiloma latisiiqua, and coniferous
Taxodium. Relevés in central Texas were all near Austin and show Q. virginiana, but almost
indistinguishable Q. fusiformis occurs near there and further west. The communities are described
in more detail in [46].
Constraints on Evergreen Broad-Leaved Forests ... 341

and A.6) and illustrate most, if not all, of the common characteristics of these main
evergreen broad-leaved forest types.
Bay forests (also “bayheads” or “baygalls”; see Relevé Table A.1) are domi-
nated by shade-tolerant laurophyll trees and thus represent the only true counter-
parts to the “laurel forests” of East Asia. On the US coastal plain, though, they
occur only in small, moist depressions scattered very widely over the otherwise
sandy substrate. Such depressions often have a thin accumulation of peat, which
keeps them permanently moist through the hot summers (see [170], p 74). The main
tree taxa are a small sample from the main laurophyll genera of East Asia: Persea
borbonia, P. palustris, Magnolia virginiana, and Gordonia lasianthus. Of these, M.
virginiana is the most common and occurs also in floodplains and some more mesic
upland forests and woody savannas. The combined canopy and tree understorey of
bay forests is usually fairly dense and may also contain evergreen Ilex opaca, I.
cassine, and occasional remnant pines or deciduous trees. As a result, the lower
understorey is quite variable and consists largely of juveniles, other shade-tolerant
woody species such as laurophyllous Ilex spp. and Lyonia lucida, and sometimes
abundant ferns. Vines and epiphytes may occur but are not abundant. Bay forests
are most common from the Carolinas to central Florida. The examples are from the
southern part of this range, where Gordonia becomes more prominent.
Beech-Magnolia forests (see Relevé Table A.2) are co-dominated by deciduous
beech (Fagus grandifolia) but are included here because they, plus a variant with
co-dominant [evergreen] Ilex opaca, present the most luxurious development of
evergreen southern magnolia (Magnolia grandiflora). These forests were recog-
nized early [48, 123] as a potential climatic-climax forest type at least on certain
mesic upland sites on the southeastern coastal plain (see also [80]). Due to the
generally droughty sand substrate and the prevalence of fire, however, they are
found only in small, topographically protected areas with better soils. Based on old
land records from 1821, widespread forest cover by “primeval” beech-magnolia
forest was reconstructed by Delcourt and Delcourt [34] on the loess hills of
Louisiana; similar beech-magnolia forests have been described on mesic sites in the
Apalachicola bluffs area of western Florida [5, 36, 83]. The “purest” old-growth
stands are not species-rich, may have few epiphytes and little understorey, vines
(e.g. Smilax) only in lower layers, and may have only stately beech and magnolia in
canopies that can reach 40 m (e.g. [172]). Other forests co-dominated by Magnolia
grandiflora include some bluff forests on seepage slopes in southern Georgia
([170], pp 169) and the Magnolia grandiflora-Ilex opaca type of Veno [159] in
northern Florida.
The stability of beech-magnolia associations was questioned due to poor
regeneration by magnolia under its own dense canopy, although it does regenerate
with an admixture of other canopy trees [81], cf [43]. As a result, beech-magnolia is
now often considered as a variant of the highly variable but self-perpetuating
Southern Mixed Hardwoods, now accepted as the potential forest type because
continuing human disturbance insures greater variability in canopy and overall
diversity. Relevé Table A.2 shows an old-growth beech-magnolia forest in
Louisiana, protected from fire by its location in the bend of a river [172]. The other
342 E. O. Box and K. Fujiwara

two examples are from species-rich ‘southern mixed hardwoods’ forests in western
Florida, with beech-magnolia co-dominance. The first (Apalachicola Bluff) is on a
gentle “steephead” slope (cf [83]) and the other in a slight depression.
Live Oak-Magnolia forests (see Relevé Table A.3), i.e. a combination of
Quercus virginiana (live oak) and Magnolia grandiflora, were first described as a
consistent association by Daubenmire [33], cf [85], and are potentially the most
widely occurring evergreen broad-leaved forest type across the somewhat drier
uplands as well as on old coastal dunes and other drier (i.e. sandier) lowland areas.
Although live oak is the potential long-term dominant, slow-growing magnolia
remains in the canopy and understorey. The canopy may also include occasional
deciduous companions, especially Carya glabra and Q. nigra. Palmettos (Sabal or
Serenoa) may occur in understoreys, along with evergreen shrubs, but the overall
number of species may not be large.
The maritime version of this forest type, called ‘maritime forest’ since Bourdeau
and Oosting [7], is illustrated by the first example (Naval Live Oaks) in Relevé
Table A.3, which is from coastal dunes near Pensacola (western Florida). This
forest actually contains more magnolia than shown in the relevé. The best inland
example in Relevé Table A.3 is the last (Dudley Hammock, see photo in [170],
p 188), well protected on a military reservation in southeastern Georgia. The middle
example is from a sandy levee along the Mississippi River near its delta south of
New Orleans. In this variant there is little magnolia, perhaps due to the high
groundwater. Very species-poor live oak-magnolia forests were also found by the
authors on very low-lying areas along the Texas coast north of Corpus Christi.
Upland Oak forests (cf Relevé Table A.4) are mainly secondary and differ from
the previous type by having less live oak and magnolia, at least in their current
composition. They are usually dominated currently by Quercus hemisphaerica, a
pioneer evergreen oak that colonizes wide areas after forest cutting or other
wide-area disturbance (e.g. [51]). Such woods are found especially in various state
parks in Florida, where Q. hemisphaerica grew up quickly on abandoned land
acquired by the state in the first half of last century. If live-oak propagules are
available, and there is no fire, these forests may succeed to the live oak-magnolia
type described above and could cover wide upland areas; if not, due to ongoing
disturbance, they may remain for decades as a mix of co-dominant evergreen (Q.
hemisphaerica) and deciduous oaks (mainly Q. nigra and Q. falcata). Q. hemis-
phaerica lives only about 100 years, but in some cases its dominance in relatively
mature stands may continue for decades, as in the Kolomoki Mounds (southwestern
Georgia) example in Relevé Table A.4, which has no live oak at all (cf [51]). The
other examples, both from northern Florida, have live oaks but only in the canopy.
These forests continue west to southeastern Texas and can be fairly species-rich,
with a relatively consistent woody composition (as suggested by the three
examples).
Oak-Palm forests (cf Relevé Table A.5) combine live-oak (Q. virginiana)
canopies with often dense palmetto (Sabal palmetto) subcanopies, plus other oaks
and some subtropical understorey species (cf [147]). These forests are often cir-
cumscribed by topography into discrete hammocks, and occur mainly in central to
Constraints on Evergreen Broad-Leaved Forests ... 343

south-central Florida, in the transition from the temperate-zone flora to the north
and the subtropical flora to the south [32]. Live oak occurs mainly in the canopy,
while Sabal shows better regeneration. Tillandsia species are especially common,
as well as vines and other epiphytes, both warm-temperate and subtropical.
Subtropical species increase southward, but there may also be some admixture of Q.
laurifolia or Q. hemisphaerica (not both together), depending on site conditions.
The first and last examples in Relevé Table A.5 are from central Florida; the middle
example (Little Corkscrew, southwestern Florida) illustrates the southward increase
in subtropical epiphytes.
Floodplain and Bottomland forests in the South (cf Relevé Table A.6) occur in
many forms, some entirely deciduous, many semi-evergreen, and some mainly
evergreen. The most common evergreen broad-leaved dominant of floodplain for-
ests is Q. laurifolia, but Magnolia virginiana may be co-dominant or at least
abundant in many situations, along with various deciduous canopy and understorey
trees. Q. virginiana and Magnolia grandiflora are much less common in flood-
plains. Warm-temperate vines and epiphytes are common, while composition of the
herb layer depends strongly on topography and hydrology. The first example (Pearl
River), with strong dominance by Q. laurifolia, is from southeastern Louisiana (cf
[171]). The last two examples are from bottomlands surrounded by sandy uplands:
Tiger Creek in central Florida and the Beidler Forest near the Fall Line sandhill area
of central South Carolina, below Columbia. Despite the distance between these
examples, their tree composition is remarkably consistent.
Other evergreen broad-leaved woods on the coastal plain include, in particular,
the xeric upland oakwoods on the deep, sterile sand of old dunes. These oakwoods
are dominated by Q. geminata, a shorter, often visibly scrubby oak tree with
sclerophylls harder than those of Q. virginiana. There are also oak-pine sandhill
savannas with deciduous Q. laevis, and shorter oak scrub co-dominated by ever-
green Q. geminata, Q. myrtifolia and Q. chapmanii; these two sandhill types are
distinguished by fire regime [114].

6 Constraints and Vegetation Dynamics

Forests may be constrained by limits on their main tree species or on forest

development itself. The main evergreen broad-leaved trees are constrained first of
all by climatic conditions, in particular winter cold and increasing dryness westward
(and southward in Florida). For temperature, at least three possible limitations have
been suggested. Mean temperature of the coldest month was used by Köppen [76]
to delimit temperate climates, and the 0°C isotherm for mean January temperature
was shown by Hocker [63] to fall close geographically to the northern limit of
“southern pines”; also, the northern limit of laurel-forest trees along the northern
Pacific coast of Japan falls near the 0.5°C mean January isotherm. Due to variations
in continentality, though, there is no consistent relationship worldwide between
winter mean temperatures and the boundaries of extra-tropical vegetation types
344 E. O. Box and K. Fujiwara

(except in the case of permafrost). On the other hand, Greller [56] showed that the
5°C isotherm for mean minimum temperature in the coldest month does match the
northern limit of southeastern evergreen broad-leaved tree species fairly well.
Unfortunately, this relationship cannot be tested globally because this minimum
gets defined differently in different data sources.
More important may be the clear physiological limitation by ice formation
within plant cells, which can occur during even short-term exposure [129, 20, 50],
cf [134], and desiccation damage due to frozen soil (e.g. Sakai [130], [151]). For
temperate-zone evergreen broad-leaved tree species, absolute minimum temperature
appears to be a limiting factor globally [174]. Estimates of this limiting ‘cardinal
temperature’ have converged at about −15°C [20, 88, 131–133, 165]. This would
appear to limit the region of potential evergreen broad-leaved forests and wood-
lands in the Southeast to most of Florida, the southern part of Georgia, and narrow
strips along the Carolina and Gulf coasts. Harder-leaved evergreen broad-leaved
trees (i.e. Quercus virginiana and Magnolia grandiflora) survive somewhat lower
absolute minima (to about −18°C; cf [132]) but are outside their natural range
further north, where planted or escaped trees do survive near settlements and
sometimes in closed forests (numerous personal observations).
The other main climatic constraint, of course, is available moisture. Topography
and substrate aside, the geographic pattern of climatic moisture (see Fig. 1) sug-
gests forest potential all across the Southeast, but decreasing rapidly west of the
Mississippi River and less rapidly southward in Florida. There is a corresponding
general gradient in vegetation, with forests declining in stature, then opening up
into open woodlands, and finally giving way to grasslands or savannas. Within this
climatic framework, vegetation varies over much shorter distances in the
landscape-scale mosaic of wetter and drier areas dictated by topography, substrate
and recurring fire. The constraints are obvious, but understanding their effects is
complicated because, even on drier surfaces, groundwater may not be far below the
surface, at least in winter and spring. This groundwater effect is involved in the
concept of ‘flatwoods’ (well illustrated by [170], pp 194) and can be seen well at the
University of Florida preserve near Welaka, which preserves examples of ‘xeric
hammock’ (Q. virginiana) with lower groundwater; ‘mesic hammock’ (Magnolia
grandiflora-Ilex opaca) with higher groundwater; and ‘hydric hammock’ (ever-
green Quercus spp.-Sabal palmetto-Liquidambar) with groundwater near or at the
surface, at least seasonally (see initial descriptions by [84], cf [57]).
The potential water holding capacity of upland sandy soils may be influenced
more subtly by another factor, namely their humic content, which depends on the
rate at which plant litter decomposes and is made available for incorporation into the
soil. In the early 1980s, Meentemeyer [99] made a globally applicable climatic
model for potential litter decomposition rate (percent annual litter disappearance),
and Box mapped the balance between annual primary production and the amount of
litter decomposed. (This was done using a world mapping system which was lost
completely when science was shoved off mainframe computers and their high-speed
page printers, onto do-it-all-yourself desktop machines). The resulting map was thus
never published but divided the world into regions where litter could potentially
Constraints on Evergreen Broad-Leaved Forests ... 345

accumulate (be produced faster than it could decompose) or decompose completely,

with no accumulation (decomposition potentially faster than production). The
dividing line between these regions of potential litter accumulation and disappear-
ance ran through the US South, roughly along the Fall Line. Of course, not all litter is
the same: evergreen leaves with more lignin generally decompose more slowly than
do deciduous leaves [98]. Nevertheless, there is the suggestion that microclimatic
conditions may play a role in limiting the amount of humus available to developing
soils. Humus may not be incorporated rapidly into clay per se, but it can be held on
the soil surface if there is a more or less permanent layer of dead leaves and other fine
litter. If the litter decomposes too fast, as in the warmer winters of the coastal plain,
less humus would be available for eventual incorporation into the soil. Of course, a
humus layer also provides a more humid microclimate for the establishment of
species of the middle and latter stages of succession.
The particular problem in the US South is that there are four potential limitation
mechanisms for evergreen broad-leaved forests, and the critical values for all four
fall close together geographically, near or along the Fall Line boundary between the
coastal plain and the piedmont. The two non-climatic boundaries occur at the Fall
Line: 1) the major substrate discontinuity between coastal-plain sand and piedmont
clay; and 2) the more hypothetical boundary between climates of potential litter
accumulation versus disappearance. The two thermal boundaries represent infre-
quent low temperatures: 3) the 5ºC isotherm for mean minimum winter tempera-
tures [56], and 4) the −15ºC isotherm for absolute minimum temperature (although
in reality, unusual extremes occur in spots, with some outliers still further south,
e.g. −18ºC in Tallahassee; cf Table 1). These isotherms fall across the middle of the
coastal plain, near the Florida-Georgia border a bit further south (and closer to the
coast in the Carolinas and westward to Texas). This location approximates the
transition between mainly evergreen live-oak forests and semi-evergreen “southern
mixed hardwoods” forests, with their substantial deciduous components.
Numerous species range boundaries fall in or near this narrow region of multiple
hypothetical limitations, but there are also constraints on forest development in
general. The first of these is fire, which may be caused by lightning strikes or set by
man [86, 25, 78]. Florida, in particular, is one of the most lightning-prone areas in
the whole world. Fire may not always kill mature trees, but it does remove seed-
lings, saplings and young trees quite effectively. Bay forests, beech-magnolia for-
ests, floodplain forests, maritime live-oak forests, and even the transitional
oak-palm forests are usually circumscribed topographically and thus relatively well
protected from fire. The live oak-magnolia and upland-oak types, though, can be
quite vulnerable to fire and thus to retrogression to the fire-maintained (pyroclimax)
pine savannas that have historically covered large expanses on the coastal plain (cf
[27]). If the fire cycle could be broken, the broad uplands might become live oak
forest – but there is always a natural source of new fires: dry years and lightning.
The interplay of fire, topography, substrate and potential vegetation is illustrated
especially well by Wharton [170, p 178].
Another constraint on the development of evergreen broad-leaved forests
involves competition from deciduous trees, which generally grow faster (e.g. [138])
346 E. O. Box and K. Fujiwara

and have an advantage where they cannot be shaded out by more shade-tolerant
evergreens. Due to more rapid growth and response to disturbance, deciduous trees
may also have an advantage near the northern limit of the evergreen broad-leaved
trees (cf [69, 92, 116]). Sensitivity to winter cold may also limit broad-leaved
evergreens northward even before lethal cold is reached ([174], p 69). Most of the
native evergreen broad-leaved tree species seem quite able to grow on piedmont
substrates, but they may not be as able as deciduous trees to grow densely and
rapidly enough to form and dominate forests. Secondary deciduous forests may
succeed to evergreen live-oak forests only if live oaks can create sufficiently dense
canopies. Drier climatic conditions may result in less dense evergreen canopies (as
well as more fires) and thus a reduced advantage of shade tolerance. This too
suggests a greater role for deciduous trees, especially with continuing disruption
due to warming and drying.
Finally, in today’s world there is another major factor that influences forests,
namely break up of continuous woody vegetation cover into ever smaller fragments
by man’s activities. This reduction of extensive forest canopies permits more light
penetration from the sides—as well as population fragmentation and various edge
effects. Landscape fragmentation and other disturbance thus favor deciduous trees,
as depicted in Fig. 3 (from [11]). As a result, much of the currently remaining forest

Fig. 3 Results of Competition between Evergreen and Deciduous Forest Canopies in Humid
Undisturbed versus Disturbed Landscapes (from [11]). In humid climates, evergreen canopy trees
are generally more shade-tolerant than are deciduous canopy trees. Thus, in undisturbed
landscapes, where forest canopies may be relatively closed over large areas, the more
shade-tolerant evergreen trees generally shade out and replace deciduous trees. On the other
hand, in disturbed landscapes, where forests are much more fragmented and more light is available
laterally, faster-growing deciduous trees are usually able to retain their canopy dominance
Constraints on Evergreen Broad-Leaved Forests ... 347

and woodland is deciduous or at best only semi-evergreen, depending not only on

physical factors but even more on recent land-use history and the resulting local
availability or lack of live-oak propagules.

7 Conclusion

Unlike warm-temperate climatic regions elsewhere, the southeastern US is cli-

matically drier and the coastal plain provides a harsh, immature, nutrient-poor
substrate in which water and nutrient conditions vary drastically over short dis-
tances, from deep, xeric sands to permanently moist depressions, seasonally
inundated flats, or even more hydric bottomland forests and swamps. As a result,
the original landscape was a forest-scrub-wetland mosaic, and its current potential is
not for “laurel forests” over large areas but rather, where soil, topography and fire
permit, for a perhaps slow progression to domination by coriaceous-leaved, even
sclerophyll forests and woods with somewhat more open canopies and a greater role
for deciduous trees and dense understoreys, even at maturity. Where limestone
underlies shallow sand, as in northern and central Florida, the greater availability of
nutrients seems to favor deciduous trees (cf [94, 95, 109]). Even so, large trees of Q.
virginiana or Q. hemisphaerica may also occur, suggesting that these evergreen
species may have been dominant or co-dominant before the original forests were
cut. As in East Asia, secondary forests on the coastal plain are overwhelmingly
deciduous. The long periods of time required for evergreen broad-leaved trees to
return to dominance, perhaps 100 years, explain partly why the regional forest
potential has been described as mainly deciduous “Southern mixed hardwoods.”
Taken together, these environmental conditions suggest a complicated pattern of
vegetation limitation across the coastal plain and raise many questions concerning
vegetation dynamics and status. Zonal warm-temperate laurophyll forests occur on
the coastal plain only in moist depressions that compensate for drier overall con-
ditions, as suggested by the ‘law of relative habitat constancy’ ([163], pp 32–33).
The term ‘climax’ has been criticized frequently, especially in Europe, but does
have the advantage of permitting useful, widely used terminology, such as climatic
climax, edaphic climax, and fire climax [21, 28], cf [97]. These terms can reflect the
vegetation status more completely than the single term ‘potential natural vegetation’
– which still has not resolved the role of fire in vegetation. (Can a fire-climax
vegetation type be seen as potential natural vegetation, sensu Tüxen or Miyawaki?)
Pyrogenic vegetation types, such as pine woodlands/savannas, represent arrested
development of vegetation which might, in the absence of fire, succeed to zonal
vegetation on “zonal” sites or to topo-climax or edaphic-climax vegetation on
corresponding azonal sites.
Since most currently mature evergreen broad-leaved forest stands were not
observed (or identified as important) during earlier successional stages, it is difficult
to match secondary forests and mature evergreen broad-leaved forests in charac-
teristic successional sequences. Many studies document changes in species
348 E. O. Box and K. Fujiwara

composition, invoking “succession” but without identifying any mechanism except

fire suppression (e.g. [61, 85, 110, 159]). Evidence that evergreen broad-leaved
forests can replace deciduous forests must come from remaining large trees (e.g. Q.
virginiana) and from good regeneration of their seedlings and saplings in shady
understoreys.
How might these limiting factors, and vegetation, change with global warming
and other environmental changes? Isotherms for mean and mean-minimum winter
temperatures will move northward with general warming, but it is much more
difficult to anticipate absolute minima, which may be the best temperature delimiter
for temperate-zone evergreen broad-leaved tree species and for a warm-temperate
climate in general (see global systematization in [12], cf [174]). In Florida, for
example, many new absolute minima were reached during the 1980s, when winters
elsewhere in eastern North America were returning to normal after the unusually
cold winters of the latter 1970s. This may be due to weakening of the polar vortex
with overall global warming, which lets the coldest polar air masses escape
southward more easily [77]. Even if the [potential] absolute minima remain the
same, though, they may become less frequent. In addition, most global and regional
models have suggested that global warming will bring generally drier conditions in
the US South (e.g. [70, 90, 164]). Ever since their beginnings in the 1980s, global
climate models have consistently suggested that, in a warmer world, the subtropical
high-pressure belts would move farther away from the equator, in both the
northern and southern hemispheres ([148], cf [2]). The recent drier conditions and
more frequent, larger wildfires in the western USA, Australia, and some
Mediterranean areas, suggest that this is already happening, at least on continental
west sides.
Warming will mean two things: the northern climatic limit for evergreen
broad-leaved trees would move northward, onto the clay substrates of the piedmont;
and potential evapotranspiration rates would be higher. If this latter is not com-
pensated by sufficient increases in precipitation, there would be net drying. Net
drying would mean that:
– current areas of open woodland and pine savanna (e.g. central Florida) would
become even more open;
– current areas of upland forest may become too dry to support closed forest;
– some bottomland and floodplain forests, and some wetlands, could contract;
– some moist flats (e.g. pocosins) and depressions (bay forests) could also dry out,
at least partly; and
– .the overall potential for evergreen broad-leaved forests, and other forests,
would be threatened.
Some forest types are somewhat protected by topography, against drying as well as
fire, but drying may make them more susceptible to encroachments by fire from
surrounding landscapes. Upland forests, already in the driest situations, will become
threatened even more by fire than now. Although changing physical conditions will
influence species range boundaries, these will depend also on properties of the
Constraints on Evergreen Broad-Leaved Forests ... 349

species themselves: phenology and abilities to reproduce, germinate, colonize and

become established, and to compete with other species.
Warming and drying may also affect some non-climatic factors. Topography and
substrate type will not be affected by global warming, except perhaps along the
immediate coast, due to sea-level rise. Soil humic content may not be affected
quickly, but soil moisture and near-surface temperatures and humidity will be
affected (warmer and drier), with consequent effects on litter decomposition,
infiltration of water into the soil, seedling establishment and thus succession. These
consequences may be difficult to anticipate, and the high variability between wet
and dry years, as well as between warmer and colder winters, will make interpre-
tation of constraints and future potentials even more difficult. Wildfires can be
expected to become more frequent. Meanwhile, landscape fragmentation continues,
causing further drying and local warming.
The main purpose of this study was thus to identify the types of evergreen
broad-leaved forest that do occur and try to understand their constraints. From this
information it may be possible to infer forest dynamics and ultimate potential, based
on understorey composition, remaining large individual trees in similar situations,
and observed patterns of species replacement over sufficient periods of time. As a
final prognosis, though, we must unfortunately expect a reduced potential for
warm-temperate evergreen broad-leaved forests on the coastal plain, with more
open landscapes and a greater role for faster-growing deciduous species, including
subtropicals. The area in which winters permit evergreen broad-leaved trees may
well expand northward, in cities as well as countryside. Even in the absence of land
conversion, other disturbance by humans, and fires, however, the area over which
evergreen broad-leaved trees can form more or less permanent evergreen
broad-leaved forests will probably be reduced. There seem to be two main
unknowns:
1. Milder winters may not change absolute temperature minima significantly; in
some places, new record lows continue to occur, in spite of more general global
warming.
2. Can the coastal-plain canopy dominants, especially live oak, colonize and
capture areas currently covered by deciduous forest on piedmont red clay? Live
oaks can grow on clay, as in piedmont cities, but can they colonize these areas
and outcompete the existing forest?
Mangrove species are being seen further north in Florida (e.g. [22]), but otherwise
there has been very little published evidence of subtropical species moving
northward or of evergreens moving from the coastal plain onto the piedmont. This
is partly because individual observations and anecdotes are not publishable units,
which require well documented multi-year studies. One exception to this lack is the
Cinnamomum-Ardisia forest near Tallahassee documented by Clewell and Tobe
[30]. At any rate, species are probably migrating northward without being recog-
nized, but changes in forest structure will be slow to follow [92].
350 E. O. Box and K. Fujiwara

Appendix: Composition and Structure of the Main

Evergreen Broad-Leaved Forest Types

The six main types of evergreen broad-leaved forests in the southeastern USA, as
identified in the main text, are described here by means of Braun-Blanquet relevés,
which show species composition and abundance in individual forest layers and thus
the physical structure of the forests themselves. Each of the following six tables
(identified as Relevé Tables A.1–A.6, as opposed to the Tables in the main text)
includes three relevés, from different areas. Each relevé has four columns, for the
T1, T2, S and H layers of the respective forests. The numbers in these layer
columns represent species cover and sociability on the normal Braun-Blanquet scale
of 1-5, with a plus sign (+) for sporadic occurrence.

Table A.1 Laurophyll “Bay” forest examples from the Southeastern US coastal plain

All three sites were in slight depressions, some with peat, in the otherwise sandy coastal plain
topography. Lake Louise is in south-central Georgia and has a boardwalk-accessible “bay” forest
on one side. Lake Kissimmee is in central Florida, where Gordonia lasianthus becomes more
prominent. Tiger Creek is a very diverse preserve, also in central Florida. Evergreen species are
shown in boldface. Numbers under the headings represent the height and cover percentage of each
stand layer (T1, T2, S and H). The following companion species occurred only infrequently (+ or
+0.2) in a lower layer (unless epiphytic) of only one relevé.
At Lake Louise, S: Myrica cerifera, Leucothoe axillaris, Vaccinium stamineum, V. atrococcum.
At Lake Kissimmee, T2: Tillandsia fasciculata, T. setacea; S: Vaccinium arboreum, Decumaria
barbara; H: Quercus hemisphaerica, Osmunda regalis, Vaccinium “hairy”, Araceae sp., and
Rosaceae “maloid”. At Tiger Creek, S: Vaccinium sp.; H: Tillandsia recurvata.
Constraints on Evergreen Broad-Leaved Forests ... 351

Table A.2 Beech-Magnolia forest examples from the US Southeastern coastal plain

The Zemurray Estate is in eastern Louisiana, north of Lake Ponchartrain, and contains classic
examples of tall, mature beech-magnolia forest, sometimes with relatively little understorey.
Apalachicola Bluff is a TNC preserve in western Florida and contains especially species-rich
examples of beech-magnolia forest with other canopy trees and rich understoreys. Wakulla Springs
is a state park in northern Florida, near Tallahassee, and also contains a species-rich
beech-magnolia forest with other canopy trees. These last two stands could also be considered
especially species-rich examples of “Southern Mixed Hardwoods” (sensu Quarterman & Keever
1962). Evergreen species are shown in boldface. (Polypodium polypodioides is poikilohydrous and
perhaps best described as semi-evergreen.) The numbers under the headings represent the height
and cover percentage of each stand layer (T1, T2, S and H). The following companion species
occurred only infrequently (+ or +0.2) in one lower layer of only one relevé:
At Zemurray, S: Magnolia virginiana, Trichomanes petersii; H: Erythrina herbacea, Epifagus
virginiana, Vaccinium tenellum, Ligustrum sinense, Elephantopus tomentosus, Smilax
rotundifolia, Smilax glauca, Smilax sp., Vitis sp., Liliaceae sp., Panicum sp.
352 E. O. Box and K. Fujiwara

At Apalachicola Bluff, T2: Halesia caroliniana, Planera aquatica, Prunus (decid.); S: Vaccinium
arboreum, Pinus taeda, Hydrangea quercifolia, Aesculus pavia, Smilax auriculata; H:
Ampelopsis arborea, Passiflora lutea, Parthenocissus quinquefolia, Agrimonia microcarpa,
Cnidoscolus stimulosus, Croomia pauciflora, Sabaletonia, Chasmanthium latifolium, Aster sp.,
Elephantopus sp., Senecio sp., Asplenium sp., Viola sp., Panicum “narrow”, Panicum ``wide''
At Wakulla Springs, S: Aralia spinosa, Bumelia lanuginosa, Chionanthus virginicus, Crataegus
marshallii, C. viridis, Ilexambigua, Morus rubra, Prunus serotina, Sabal palmetto, Ulmus alata,
Vaccinium stamineum, Viburnum nudum, Menispermum canadense; H: Botrychium dissectum,
Desmodium galactia, Rubus flagellaris, Ruellia caroliniensis, Sabal minor, Solanum americanum,
Uniola laxa, Desmodium “big”, Sanicula sp., Carex “short”, Carex “big”.

Table A.3 Live Oak-Magnolia forest examples from the Southeastern US coastal plain.

The Naval Live Oaks is a preserve of maritime live-oak forest on sand only a few meters above the
nearby sea level, near Pensacola in western Florida. The Mississippi Delta site is on a levee (sand)
of the Mississippi river in Plaquemines Parish, southeast of New Orleans, at elevation of only a
very few meters above sea level (high groundwater). Dudley’s Hammock is an especially good
example of a mature oak-magnolia hammock, located and well preserved within Moody Air Force
Base in southeastern Georgia. Numbers under the headings represent the height and cover
percentage of each stand layer (T1, T2, S and H). Evergreen species are shown in boldface.
(Polypodium polypodioides is semi-evergreen, and Quercus nigra is “tardily deciduous”.) The
following companion species occurred only infrequently (+ or + 0.2) in only one lower layer (S or
H) of only one relevé:
At Naval Oaks, S: Smilax laurifolia, Smilax smallii; H: Ilex vomitoria, Hypericum reductum,
Solidago sp. on the Mississippi levee, S: Cornus foemina, Sambucus canadensis; H: Smilax
rotundifolia, Parthenocissus quinquefolia, Tradescantia sp.
At Dudley’s Hammock, S: Quercus michauxii, Carya tomentosa, Acer rubrum; H: Persea
palustris, Lyonia lucida, Gelsemium sempervirens, Smilax bona-nox, Pteridium aquilinum,
Uniola laxa, Scleria triglomerata.
Constraints on Evergreen Broad-Leaved Forests ... 353

Table A.4 Upland Oak forest examples from the Southeastern US coastal plain.

Kolomoki Mounds is a state park in southwestern Georgia and contained this especially good
example of an upland Quercus hemisphaerica forest that may succeed to Q. virginiana
dominance. In the 1980s, many state parks in Florida were in this stage of succession. The
Suwanee River site, already dominated by live oak, is in a state park in northern Florida, on a
humic-sand upland near the river. The Welaka site is a preserve owned by the University of
Florida, in north-central Florida (southeast from Gainesville); it contains good examples of this
“moist” live-oak hammock as well as several other forest types. Numbers under the headings
represent the height and cover percentage of each stand layer. Evergreen species are shown in
boldface. (Polypodium polypodioides is semi-evergreen.) The following companion species
occurred only infrequently (+ or +0.2) in only one lower layer (S or H) of only one relevé.
At Kolomoki Mounds, S: Fagus grandifolia, Carya tomentosa, Nyssa biflora, Diospyros
virginiana, Acer leucoderme, Euonymus americanus, Rhamnus caroliniana, Aralia spinosa,
Nandina domestica, Smilax laurifolia, Ribes sp.; H: Sassafras albidum, Yucca filamentosa,
Sabal minor, Asplenium platyneuron, Smilax herbacea.
At Suwanee River, S: Asimina triloba, Chionanthus virginicus, Ilex vomitoria, Juniperus
virginiana, Viburnum nudum, Campsis radicans; H: Smilax pumila, Cocculus carolinus,
Cnidoscolus stimulosus.
At the Welaka Preserve, S: Persea borbonia, Malus sp.; H: Smilax rotundifolia, Galium sp.
354 E. O. Box and K. Fujiwara

Table A.5 Oak-Palm forest examples from the Southeastern US coastal plain

Lake Kissimmee is in central Florida, surrounded by low flat terrain with high groundwater. Little
Corkscrew Island is in the Corkscrew Swamp sanctuary and is the farthest south of all relevés
shown; this represents the transition region from temperate to subtropical flora, as suggested by the
numerous epiphytes. Merritt Island is on the central east coast of Florida, near Cape Canaveral.
Numbers under the headings represent the height and cover percentage of the vegetation layers
(T1, T2, S and H). Evergreen species are shown in boldface. (Polypodium polypodioides is
semi-evergreen.) The following companion species occurred only infrequently (+ or + 0.2) in only
one lower layer of only one relevé:
At Lake Kissimmee, S: Cinnamomum camphora, Cephalanthus occidentalis, Myrica inodora,
Asimina obovata,
Apios americana, Eugenia sp., Desmodium sp., Polypodium “big”, Malvaceae sp.; H: Ilex opaca,
Citrus paradisi, Oplismenus setarius, Psoralea sp., Hypericum sp., Galium sp., Carex “big”,
Panicum sp., Compositae sp.
At Little Corkscrew, T2: Ficus aurea; S: Persea borbonia, Bumelia reclinata, Myrica cerifera,
Serenoa repens, Schinus terebinthifolius, Vittaria lineata; H: Baccharis halimifolia, Boehmeria
cylindrica, Rubus argutus,
Mikania scandens, Smilax sp., Viola sp., Erechtites (?) sp., Arundinella (?) sp., Rubiaceae sp.
At Merritt Island, T2: Persea palustris; S: Acer rubrum, Illicium floridanum, Psychotria sulzneri,
Convolvulaceae sp., Ligustrum sinense; H: Mikania cordifolia, Ampelopsis arborea.
Constraints on Evergreen Broad-Leaved Forests ... 355

Table A.6 Bottomland evergreen oak forests on the Southeastern US coastal plain

All relevés represent floodplain forests, with evergreen species shown in boldface. Each stand was
dominated by Quercus laurifolia, with some deciduous canopy trees, and often with evergreen
Magnolia virginiana or M. grandiflora. The Pearl River basin is in southeastern Louisiana and
shows a classic Q. laurifolia floodplain forest. Tiger Creek is a diverse preserve in central Florida;
the example shown here had a mixed canopy of Q. laurifolia and M. virginiana, plus deciduous,
typically bottomland trees. The Biedler Forest stand is in the Fall Line sandhills of central South
Carolina, is slightly more open, and is dominated by Q. laurifolia plus deciduous trees. Numbers
under the headings represent the height and cover percentage of each stand layer (T1, T2, S and
H). The following companion species occurred infrequently (+ or + 0.2, or 1.1 if not completely
identified) in a non-canopy layer (T2, S or H) of not more than two relevés:
356 E. O. Box and K. Fujiwara

At the Pearl River basin, T2: Nyssa sylvatica; S: Betula nigra, Crataegus sp., Ilex decidua; H:
Crinum americanum,
Hypericum sp., Leersia sp., Euphorbia sp., Commelina sp., Acanthaceae sp., Carex sp., Panicum
“big wide”.
At Tiger Creek, S: Baccharis halimifolia, Callicarpa americana, Cornus foemina, Psilotum
nudum; H: Phytolacca americana, Rubus argutus, Parthenocissus quinquefolia, Thelypteris sp.,
Galactia sp., Sorghum sp., Gramineae sp.
At Biedler Forest, S: Juniperus virginiana, Ilex decidua, Cornus amomum; H: Euonymus
americanus, Rubus trivialis,
Ulmus alata, Viburnum dentatum, Ampelopsis arborea, Anisostichus capreolata, Clethra alnifolia,
Lonicera japonica,
Mikania scandens, Smilax glauca, Cornus amomum, Desmodium rotundifolium, Hexastylis
arifolia,
Hypoxis micrantha, Lobelia elongata, Polygonum hydropiperoides, Rhododendron canescens,
Thelypteris kunthii,
Triadenum virginicum, Woodwardia areolata, Uniola laxa, Aster sp., Cirsium sp., Galium sp.,
Solidago “small lvs”,
Panicum “small”, Panicum “big”, Carex sp.

References

1. Alexander TR (1953) Plant succession on Key Largo, Florida, involving Pinus caribaea and
Quercus virginiana. Q J Fla Acad Sci 16(3):133–138
2. Angell JK (2006) Changes in the 300-mb north circumpolar vortex, 1963–2001. J Climate
19:2984–2994
3. Bartram W (1791) (facs. 1973). Travels through North and South Carolina, Georgia, East
and West Florida. Beehive Press, Savannah. 142 pp (facs. of London eds.)
4. Bazzaz FA (1979) The physiological ecology of plant succession. Ann Rev Ecol Syst
10:351–371
5. Blaisdell RS, Wooten J, Godfrey RK (1974) The role of magnolia and beech in forest
processes in the Tallahassee, Florida, Thomasville, Georgia, area. Procs Tall Timbers Fire
Ecol Conf 13:363–397
6. Boufford DE, Spongberg SA (1983) Eastern Asian-Eastern North American phytogeo-
graphical relationships a history from the time of linnaeus to the twentieth century. Annals
Missouri Bot Garden 70(3):423–439
7. Bourdeau PF, Oosting HJ (1959) The maritime live oak forest in North Carolina. Ecology
40:148–152
8. Box EO (1986) Some climatic relations of the vegetation of Argentina. global perspective.
Veröff. Geobot. Inst. ETH, Stiftung Rübel (Zürich), pp 181–216
9. Box EO (1995a) Climatic relations of the forests of east and Southeast Asia. In: Box EO,
et al. (eds) Vegetation Science in Forestry, Handbook of Vegetation Science, 12/1. Kluwer,
Dordrecht, pp 23–55
Constraints on Evergreen Broad-Leaved Forests ... 357

10. Box EO (1995b) Global potential natural vegetation: dynamic benchmark in the era of
disruption. In: Murai Sh (ed) Toward Global Planning of Sustainable Use of the Earth –
Development of Global Eco-engineering. Elsevier, Amsterdam, pp 77–95
11. Box EO (1997) Bioclimatic position of evergreen broad-leaved forests. In: Island and
high-mountain vegetation: biodiversity, bioclimate and conservation. Procs., annual IAVS
meeting, Tenerife, April 1993. Universidad de La Laguna, Servicio de Publicac., Tenerife
(Canary Islands), pp 17–38
12. Box EO (2016) Global bioclimatic zonation. In: Box EO (ed) Vegetation Structure and
Function at Multiple Spatial, Temporal and Conceptual Scales. Springer, pp 3–52
13. Box EO, Crumpacker DW, Hardin ED (1993) A climatic model for plant species locations in
Florida. J Biogeography 20:629–644
14. Box EO, Fujiwara K (1988) Evergreen broad-leaved forests of the southeastern United
States: preliminary description. Bull. Inst Envl Sci Techn Yokahama Nat Univ 15:71–93
15. Box EO, Fujiwara K (2013) Vegetation types and their broad-scale distribution. In: van der
Maarel E (ed) Vegetation Ecology, 2nd edn. Blackwell Scientific, Oxford, , pp 455–485
16. Box EO, Fujiwara K (2015) Warm-temperate deciduous forests around the northern
hemisphere Springer. Cham, Heidelberg, p 292
17. Bratton SP (1993) Survivorship of evergreen hardwoods after wildfire in maritime forest,
Cumberland Island National Seashore, Georgia. Castanea 58:34–44
18. Braun EL (1950) Deciduous forests of eastern North America. Blakiston, Philadelphia, p 596
19. Braun-Blanquet J (1964) (1928, 1951). Pflanzensoziologie: Grundzüge der
Vegetationskunde, 3rd edn. Springer, Berlin, p 330 (English 3rd ed. 1965: Plant
Sociology, Hafner, New York)
20. Burke MJ, Gusta LV, Quamme HA, Weiser CJ, Li PH (1976) Freezing and injury in plants.
Annual Review Plant Physiol 27:507–528
21. Cain SA (1939) The climax and its complexities. Amer Midland Naturalist 21:146–181
22. Cavanaugh K, et al (2019) Climate-driven regime shifts in a mangrove-salt marsh ecotone
over the past 250 years. Procs Natl Acad Sci 116(43):21602–21608
23. Cavender-Bares J, Apostol S, Moya I, Briantais J-M, Bazzaz FA (1999) Chilling-induced
photoinhibition in two oak species: are evergreen leaves inherently better protected than
deciduous leaves? Photosynthetica 36(4):587–596
24. Chabot BF, Hicks DJ (1982) The ecology of leaf life spans. Ann Rev Ecol Syst. 13:229–259
25. Chen E, Gerber JF (1990) Climate. In: Myers RL, Ewel JJ (eds) Ecosystems of Florida,
pp 11–34
26. Chen Y-F (1995–98) Taiwan Zhibei [Vegetation of Taiwan]. (in Chinese). Yushan,
Chenxing, vol 3. Qianwe Publishers, Taipei, pp 303 + 621 + 632
27. Christensen NL (2000) (1988) Vegetation of the Southeastern coastal plain. In: Barbour MG,
Billings WD (eds) North American Terrestrial Vegetation, 2nd edn. Cambridge University
Press, pp 398–448
28. Clements FE (1936) Nature and structure of the climax. J Ecol 24:253–284
29. Clewell, A. F. 1985. Guide to the Vascular Plants of the Florida Panhandle. 605 pp.
30. Clewell AF, Tobe JD (2011) Cinnamomum-Ardisia forest in Northern Florida. Castanea 76
(3):245–254
31. Cowles HC (1911) The causes of vegetational cycles. Ann Assoc Amer Geographers 1:3–20
32. Crumpacker DW, Box EO, Hardin ED (2001) Temperate-subtropical transition areas for
native trees and shrubs in Florida: present locations, predicted changes under climatic
warming, and implications for conservation. Nat Areas J 21(2):136–148
33. Daubenmire RF (1990) The Magnolia grandiflora - Quercus virginiana forest of Florida.
Amer Midland Naturalist 123:331–347
34. Delcourt HR, Delcourt PA (1974) Primeval magnolia-holly-beech climax in Louisiana.
Ecology 55:638–644
35. Delcourt HR, Delcourt PA (1975) The Blufflands: Pleistocene pathways into the Tunica
Hills. Amer Midland Naturalist 94:385–400
358 E. O. Box and K. Fujiwara

36. Delcourt HR, Delcourt PA (1977) Pre-settlement magnolia-beech climax of the Gulf Coastal
Plain: quantitative evidence from the Apalachicola River Bluffs, north-central Florida.
Ecology 58:1085–1093
37. Dhaila S, Singh SP, Negi GCS, Rawat YS (1995) Shoot-growth phenology of co-existing
evergreen and deciduous species in an oak forest. Ecol Res 10:151–159
38. Duncan WH, Duncan MB (1988) Trees of the Southeastern United States. Univ, Georgia
Press, Athens, p 322
39. Eyre SR (1968) Vegetation and Soils: A World Picture, 2nd edn. Arnold Co., London, p 314
40. Fenneman NM (1931) Physiography of Eastern United States. McGraw-Hill, New York
41. Fenneman NM (1946) Physical Divisions of the United States. Map at scale 1:7,000,000. US
Geol. Survey, Reston (Virginia)
42. Fernald ML (1946) Identification and re-identification of North American plants: the identity
of Quercus laurifolia. Rhodora 48:137–145
43. Fowells HA (1965) Silvics of Forest Trees of the United States. Dept. Agric. Handbook 271,
Washington, 762pp
44. Fujiwara K (1981–86) 日本の常緑広葉森林の群落体系(Nihon-no jōryoku kōyō shinrin-no
gunraku taikei) [Phytosociological investigation of the evergreen broad-leaved forests of
Japan]. Bull Inst Env Sci Tech Yokohama Nat Univ 7:67–133, 8:121–150, 9:139–160,
13:99–149 (in Japanese)
45. Fujiwara K (1987) Aims and methods of phytosociology or “vegetation science.” In:
Takeda Y (ed) Papers on Plant Ecology and Taxonomy to the Memory of Dr. Satoshi
Nakanishi, pp 607–628. Kobe Geobotan. Society
46. Fujiwara K, Box EO (1994) Evergreen broad-leaved forest region of the Southeastern United
States. In: Miyawaki A, et al. (eds) Vegetation in Eastern North America, pp 273–312.
Tokyo Univ. Press
47. Fujiwara K, Box EO (1999) Evergreen broad-leaved forests in Japan and Eastern North
America. In: Klötzli F, Walther G-R (eds) Conference on Recent Shifts in Vegetation
Boundaries of Deciduous Forests, especially Due to General Global Warming, pp 273–300.
Birkhäuser
48. Gano L (1917) A study in physiographic ecology in northern Florida. Bot Gaz 63:337–372
49. Gaussen H (1955) Les climates analogues à l’echelle du monde. Compt Rend Acad Agr,
France, p 41
50. George MF, Burke MJ, Pellett HM, Johnson AG (1974) Low-temperature exotherms and
woody-plant distribution. HortScience 9:519–522
51. Gibson DJ (1992) Vegetation-environment relationships in a Southern Mixed Hardwood
forest. Castanea 57:174–189
52. Godfrey RK (1989) Trees, Shrubs, and Woody Vines of Northern Florida and Adjacent
Georgia and Alabama. Univ. of Georgia Press, Athens
53. Graham A (ed) (1972) Floristics and Paleofloristics of Asia and Eastern North America.
Elsevier, Amsterdam, p 278
54. Gray A (1846) Analogy between the flora of Japan and that of the United States. Amer J Sci
Arts 2(2):135–136
55. Greller AM (1980) Correlation of some climatic statistics with distribution of broad-leaved
forest zones in Florida, USA. Bull Torrey Botan Club 107:189–219
56. Greller AM (1989) Correlation of warmth and temperateness with the distributional limits of
zonal forests in eastern North America. Bull Torrey Botan Club 116:145–163
57. Greller AM (2003) A review of the temperate evergreen forest zone of southeastern North
America: floristic affinities and arborescent vegetation types. Bot Rev 69(3):269–299
58. Grisebach ARH (1866) Die Vegetationsgebiete der Erde, übersichtlich zusammengestellt.
Petermanns Mitteilungen 12:45–53
59. Harper RM (1914) Geography and vegetation of northern Florida. Fla. State Geol. Survey,
6th Ann. Rept., pp 163–437
60. Hart JF (1968) Loss and abandonment of cleared farm land in the eastern United States. Ann
Assoc Amer Geographers 58(3):417–440
Constraints on Evergreen Broad-Leaved Forests ... 359

61. Hartnett DC, Krofta DM (1989) Fifty-five years of post-fire succession in a southern mixed
hardwood forest. Bull Torrey Bot Club 116:107–113
62. Hildebrand-Vogel R (2002) Structure and dynamics of southern Chilean natural forests with
special reference to the Relation of evergreen versus deciduous elements. Folia Geobotanica
37:107–128
63. Hocker HW (1956) Certain aspects of climate as related to the distribution of Loblolly Pine
(Pinus taeda). Ecology 37:824–834
64. Holdridge LR (1959) A simple method for determining potential evapotranspiration from
temperature data. Science 130:572
65. Hsieh C-F, Chao W-C, Liao C-C, Yang K-C, Hsieh T-H (1997) Floristic composition of the
evergreen broad-leaved forests of Taiwan. Nat Hist Res special issue no. 4:1–16.
66. Hübl E (1988) Lorbeerwälder und Hartlaubwälder (Ostasien, Mediterraneis und
Makaronesien). Düsseldorfer Geobot Kolloq 5:3–26
67. Hunt C (1974) Natural Regions of the United States and Canada. W. H., Freeman, San
Francisco
68. Hunt DM (1994) Morphology and Ecology of Quercus Series Laurifoliae, Marilandicae and
Nigrae. In: Miyawaki A, et al. (eds) Vegetation in Eastern North America, pp. 99–120.
Tokyo University Press
69. Isogai T, Okutomi K (1991) Dynamics of an evergreen oak forest in the Kannami virgin
forest, central Japan. Jap J Ecol 41:209–223 (in Japanese)
70. Karl TR, Melillo JM, Peterson TC (2009) Global Climate Change Impacts in the United
States. Cambridge University Press
71. Kira T (1945) 農業地理学の基礎としての東亜の新気候区分 (Nōgyō Chirigaku-no Kiso to
shite no Tō-A no Shin Kikō Kubun) [A new classification of climate in eastern Asia, as a
basis for agricultural geography]. Horticult. Inst., Kyoto University, 23 pp (in Japanese)
72. Kira T (1949) 日本の森林帯 (Nippon-no shinrin-tai)[Forest zones in Japan]. Ringyō
Kaisetsu, 17:105–141. Nippon Ringyō Gijutsu Kyōkai, Tōkyō (in Japanese)
73. Kira T (1969) 照葉樹林とは何か[What is a laurel forest?] In: 照葉樹林文化 (Shōyō jurin
bunka) [Culture in the Laurel Forest Zone] (Sh. Ueyama, ed.), pp 43–84.
Chūōkōron-Shinsha, Tokyo. (In Japanese)
74. Kira T (1977) A Climatological Interpretation of Japanese Vegetation Zones. In:
Miyawaki A (ed) Vegetation Science and Environmental Protection. Maruzen, Tokyo,
pp 21–30
75. Kira T (1991) Forest ecosystems of East and Southeast Asia in a global perspective. Ecol
Research 6:185–200
76. Köppen W (1931) Grundriss der Klimakunde. Walter de Gruyter, Berlin
77. Kolstad EW, Breiteig T, Scaife AA (2010) The association between stratospheric weak polar
vortex events and cold-air outbreaks in the Northern Hemisphere. Q. J. Royal Meteorol. Soc.
136:886–893
78. Komarek EV (1968) Lightning and lightning fires as ecological forces. In: Procs. 9th annual
tall timbers fire ecology conference, vol 9, pp 169–197
79. Komarek EV (1972) Ancient fires. In: Procs. 12th annual tall timbers fire ecology
conference, pp 219–240
80. Küchler AW (1964) Manual for map “The Potential Natural Vegetation of the Conterminous
United States.” American Geographical Society, special research publ. no. 36, New York
81. Kurz H (1944) Secondary forest succession in the Tallahassee Red Hills. Proc Fla Acad Sci
7(2–3):59–100
82. Kurz H, Godfrey RK (1982) Trees of Northern Florida. Univ, Florida Press, Gainesville,
p 311
83. Kwit C, Schwartz MW, Platt WJ, Geaghan JP (1998) The distribution of tree species in
steepheads of the Apalachicola River Bluffs. Florida J Torrey Bot Soc 125:309–318
84. Laessle AM (1942) The plant communities of the Welaka Area. Biol Sci Ser (Univ. Fla.
Presses) 4(1):1–141
360 E. O. Box and K. Fujiwara

85. Laessle AM, Monk CD (1961) Some live oak forests of northeastern Florida. Quart J Fla
Acad Sci 234:39–55
86. Lafon ChW (2010) Fire in the American South: vegetation impacts, history, and climatic
relations. Geogr Compass 4(8):919–944
87. Lakela O, Wunderlin RP (1980) Trees of Central Florida. Banyan Books, Miami, p 208
88. Larcher W (1976) (1973). Ökologie der Pflanzen. Eugen Ulmer, Stuttgart, 320 pp (2nd ed.:
322pp)
89. Lenihan JM, Neilson RP (1993) A rule-based vegetation formation model for Canada.
J Biogeogr 20:615–628
90. Li WH, Li LF, Fu R, Deng Y, Wang H (2011) Changes to the North Atlantic subtropical
high and its role in the intensification of summer rainfall variability in the southeastern
United States. J Climate 24:1499–1506
91. Lieth H (1975) Primary production of the major vegetation units of the world. In: Lieth H,
Whittaker RH (eds) Primary productivity of the biosphere, pp 203–215
92. Loehle C (1998) Height growth-rate tradeoffs determine northern and southern range limits
for trees. J Biogeography 25:735–742
93. Long RW, Lakela O (1971) Flora of Tropical Florida. Univ, Miami Press, p 962
94. Loveless AR (1961) A nutritional interpretation of sclerophylly based on differences in the
chemical composition of sclerophyllous and mesophytic leaves. Ann Bot 25:168–184
95. Loveless AR (1962) Further evidence to support a nutritional interpretation of sclerophylly.
Ann Bot 26:551–561
96. Martin WH, Boyce SG (1993) Introduction: the Southeastern Setting. In: Martin WH et al
(eds) Biodiversity of the Southeastern United States. Wiley, New York, pp 1–46
97. Meeker DO, Merkel DL (1984) Climax theories and a recommendation for vegetation
classification – a viewpoint. J Range Manag 37(5):427–430
98. Meentemeyer V (1978) Macroclimate and lignin control of litter decomposition rates.
Ecology 59:465–472
99. Meentemeyer V (1984) The geography of organic decomposition rates. Annals Assn Amer
Geogr 74(4):551–560
100. Miyawaki A (1968) 関東地方の潜在自然植生と代償植生との考察 [The current potential
natural vegetation and its substitute communities on the Kanto Plain (central Japan): First
Report]. In: Yoshioka (ed) Annual Report JIBP-CT(P) for Fiscal Year 1967, pp 89–95 (in
Japanese). Sendai
101. Miyawaki A (1984) A vegetation-ecological view of the Japanese Archipelago. Bull Inst
Env Sci Technol. Yokohama Nat. Univ. 11:85–101
102. Miyawaki A (ed) (1989) 日本植生誌 (Nihon Shokusei-Shi) [Vegetation of Japan], vol. 10:
Okinawa and Ogasawara. 676pp, + 123 tables (separate volume), and 5 maps. Shibundo,
Tokyo (in Japanese, with English summary)
103. Miyawaki A, Fujiwara K (1983) Evergreen broad-leaved forests and secondary forests in the
Boso Peninsula (SE parts of Kanto region). Bull Inst Envir Sci Technol. Yokohama Natl.
Univ. 9:63–76
104. Miyawaki A, Fujiwara K, Box EO (1987) Toward harmonious green urban environments in
Japan and other countries. Bull Inst Environm Sci Technol. Yokohama Nat. Univ. 14:67–82
105. Miyawaki A, Iwatsuki K, Grandtner MM (eds) (1994) Vegetation in Eastern North America.
Univ. of Tokyo Press, Tokyo, p 515
106. Miyawaki A, Suzuki K, Fujiwara K (1977) Human impact on forest vegetation in Japan.
Naturaliste Canad 104:97–107
107. Miyawaki A, Suzuki K (1980) Natürliche und ihre sommergrünen Sekundär-Wälder in
Mitteljapan. Phytocoenologia 7:492–506
108. Monk CD (1965) Southern mixed hardwood forest of north-central Florida. Ecol Monogr
35:335–354
109. Monk CD (1966) An ecological significance of evergreenness. Ecology 47:504–505
110. Monk CD (1968) Successional and environmental relationships of the forest vegetation of
north-central Florida. Amer Midland Naturalist 79:441–457
Constraints on Evergreen Broad-Leaved Forests ... 361

111. Monk CD (1987) Sclerophylly in Quercus virginiana Mill. Castanea 52:256–261

112. Mueller-Dombois D, Ellenberg H (1974) Aims and methods of vegetation ecology. Wiley,
New York, p 547
113. Muller CH (1961) The live oaks of the series Virentes. Amer Midland Naturalist 65:17–39
114. Myers RL (1985) Fire and dynamic relationship between Florida sandhill and sand-pine
scrub vegetation. Bull Torrey Bot Club 112:241–252
115. Myers RL, Ewel JJ (1990) Ecosystems of Florida. Univ, Central Florida Press, Orlando,
p 765
116. Nitta I, Ohsawa M (1997) Leaf dynamics and shoot phenology of eleven warm-temperate
evergreen broad-leaved trees near their northern limit in central Japan. Plant Ecol 130(1):71–
88
117. Noss RF (2018) Fire Ecology of Florida. University Presses of Florida, 336pp
118. Numata M (ed) (1974) The Flora and Vegetation of Japan. Elsevier, Kodansha (Tokyo),
p 294
119. Numata M, Miyawaki A, Itow D (1972) Natural and semi-natural vegetation in Japan.
Blumea 20:435–481 (with 26 photos, veg. maps)
120. Orme AR (2002) The Physical Geography of North America. Oxford Univ. Press, New
York, p 551
121. Ovington JD (ed) (1983) Temperate Broad-Leaved Evergreen Forests. “Ecosystems of the
World”, vol 10. Elsevier, Amsterdam, p 241
122. Penfound WT (1952) Southern Swamps and Marshes. Botan. Review 18:413–446
123. Pessin LJ (1933) Forest associations in the uplands of the lower Gulf Coastal Plain (longleaf
pine belt). Ecol. 14:1–14
124. Poorter L (2009) Leaf traits show different relationships with shade tolerance in moist versus
dry tropical forests. New Phytol 181:890–900
125. Prentice IC, Cramer W, Harrison SP, Leemans R, Monserud RA, Solomon AM (1992) A
global biome model based on plant physiology and dominance, soil properties, and climate.
J Biogeogr 19:117–134
126. Quarterman E, Keever Ch (1962) Southern mixed hardwood forest: climax in the
southeastern coastal plain. USA. Ecol Monogr 32:167–185
127. Ramankutty N, Heller E, Rhemtulla J (2010) Prevailing myths about agricultural
abandonment and forest regrowth in the United States. Annals Assoc Amer Geographers
100(3):502–512
128. Rübel EF (1930) Pflanzengesellschaften der Erde. Huber, Berlin
129. Sakai A (1968a) Frost damage on basal stems in young trees. Contrib Inst Low Temp Sci
Hokkaido Univ, series B, 15:1
130. Sakai A (1968b) Mechanism of desiccation damage of forest trees in winter. Contrib Inst
Low Temp Sci Hokkaido Univ Ser B, 15:15
131. Sakai A (1971) Freezing resistance of relicts from the Arcto-Tertiary flora. New Phytol
70:1199–1205
132. Sakai A, Weiser CJ (1973) Freezing resistance of trees in North America with reference to
tree regions. Ecology 54(1):118–126
133. Sakai A, Paton DM, Wardle P (1981) Freezing resistance of trees of the South Temperate
zone, especially subalpine species of Australasia. Ecology 62:563–570
134. Sakai A, Larcher W (1987) Frost Survival of Plants: Responses and Adaptation to Freezing
Stress. Springer, Berlin, p 321
135. Sargent ChS (1880) Catalogue of the Forest Trees of North America. US Govt. Printing Ofc,
Washington
136. Schimper AFW (1898) Pflanzengeographie auf physiologischer Grundlage.
Gustav-Fischer-Verlag, Jena (English transl. 1903, by W. R. Fisher, Oxford Press; 3rd
edition with F. C. von Faber 1935).
137. Schmithüsen J (ed) (1976) Atlas zur Biogeographie. Meyers Grosser Physischer Weltatlas,
vol 3. Bibliographisches Institut, Mannheim/Vienna/ Zürich, p 80
362 E. O. Box and K. Fujiwara

138. Schulze E-D (1982) Plant life forms and their carbon, water and nutrient relations. In:
Lange OL, et al. (eds) Encyclopedia of plant Physiology, vol. 12B, pp 616–676
139. Shantz HL, Zon R (1924) Natural Vegetation. Part 1, section E, in Atlas of American
Agriculture. US Dept. Agriculture, Washington
140. Sharitz RR, Gibbons JW (1982) The Ecology of Southeastern Shrubs Bogs (Pocosins) and
Carolina Bays: A Community Profile. Publ. no. FWS/OBS-82/04. US Fish and Wildlife
Service, Div. of Biological Services, Washington, 93 pp.
141. Smith WB, Miles PD, Vissage, JS, Pugh SA (2004) Forest Resources of the United States.
In: General Technical Report NC-241, 146. US Forest Service, North Central Research
Station, St. Paul
142. Song Y-C (1988) [Evergreen Broad-Leaved Forest Ecological Research]. Southwest Normal
University Press, China, 696 pp (in Chinese)
143. Song Y-C (1995) On the global position of the evergreen broad-leaved forests of China. In:
Box EO, et al (eds) Vegetation Science in Forestry. Handbook of Vegetation Science, 12/1,
pp. 69–84. Kluwer, Dordrecht
144. Song Y-C (2013) [Evergreen Broad-Leaved Forests in China], vol 2, 788pp (in Chinese,
with English summary) + 558 pp (tables). Science Press, Beijing
145. Song Y-C, Xu G-S (2003) A scheme of vegetation classification of Taiwan. Acta Botanica
Sinica 45(8):883–895
146. Spicer BE (1969) Characteristics of the loess deposits and soils in East and West Feliciana
Parishes, Louisiana. MS thesis, Louisiana State University, Baton Rouge, 69 pp
147. Stalter R, Dial S, Laessle A (1981) Some ecological observations of the arborescent
vegetation in Highlands Hammock State Park, Florida. Castanea 46:30–35
148. Svoma BM, et al (2013) Expansion of the northern-hemispheresubtropical high-pressure
belt: trends and linkages to precipitation and drought. Phys Geogr 34(3):174–187
149. Tagawa H (1997) World-wide distribution of evergreen lucidophyll oak-laurel forests.
Tropics (Kyōto) 6(4):295–316
150. Taneda H, Tateno M (2005) Hydraulic conductivity, photosynthesis and leaf water balance
in six evergreen woody species from fall to winter. Tree Physiol 25:299–306
151. Tharp BC (1939) The Vegetation of Texas. Anson Jones Press, Houston
152. Thyroff EC, Burney OW, Mickelbart MV, Jacobs DF (2019) Unraveling shade tolerance and
plasticity of semi-evergreen oaks: insights from maritime-forest live oak restoration.
Frontiers Plant Sci 10:1–11
153. Trelease W (1917) Naming American hybrid oaks. Procs Amer Phil Soc 56:44–52
154. Trelease W (1924) The American oaks. Mem Natl Acad Sci 20:1–255
155. US Geological Survey 1970. National Atlas of the United States of America. Washington
156. Valladares F, Niinemets Ü (2008) Shade tolerance, a key plant feature of complex nature and
consequences. Ann Rev Ecol Syst 39:237–257
157. Van Auken OW, Ford AL, Dougherty SL, Stein AG (1980) Woody vegetation of upland
plant communities in the southern Edwards Plateau. Texas J Sci 32:23–35
158. Veno PA (1976) Successional relationships in five Florida plant communities. Ecology
57:498–508
159. Walter H (1954) Klimax Und Zonale Vegetation. Angewandte Pflanzensoziologie 1:144–
150
160. Walter H (1968) Die Vegetation der Erde in öko-physiologischer Betrachtung. Vol. II: Die
gemäbigten und arktischen Zonen. VEB Gustav-Fischer-Verlag, Jena, 1002 pp
161. Walter H (1977) Vegetationszonen und Klima, 3rd edn. Eugen-Ulmer-Verlag, Stuttgart,
p 309
162. Walter H (1985) Vegetation of the Earth, 3rd edn. Springer, Berlin/New York, p 318
163. Wang H, Fu R, Kumar A, Li WH (2010) Intensification of summer rainfall variability in the
southeastern United States during recent decades. J Hydrometeorol 11(4):1007–1018.
1010.1175/2010
164. Wardle P (1971) An explanation for alpine timberline. NZ J Botany 9:371–402
Constraints on Evergreen Broad-Leaved Forests ... 363

165. Watts WA (1980) The late Quaternary vegetation history of the southeastern United States.
Ann Rev Ecol Syst 11:387–409
166. Watts WA (1983) Vegetational history of the southeastern United States. In: H. E. Wright jr.,
(ed) Late Quaternary Environments of the United States. University of Minnesota Press,
Minneapolis, pp 294–310
167. Wells BW (1928) Plant communities of the coastal plain of North Carolina and their
successional relations. Ecology 9:230–242
168. Wells BW (1942) Ecological problems of the southeastern United States coastal plain. Bot
Review 8:533–561
169. Wharton CH (1978) The Natural Environments of Georgia. Georgia Dept. Nat, Resources,
Atlanta, p 228
170. White DA (1983) Plant communities of the lower Pearl River basin. Louisiana Amer Midl
Naturalist 110:381–396
171. White DA (1987) An American beech-dominated original growth forest in southeast
Louisiana. Bull Torrey Botan Club 114:127–133
172. Wood VS (1981) Live Oaking. Northeastern University Press, Boston, Southern Timber for
Tall Ships
173. Woodward FI (1987) Climate and Plant Distribution. Cambridge University Press, 174 pp
174. Wu Z-Y, Committee (eds.) (1980, 1995) Zhongguo Zhibei [Vegetation of China]. Science
Press, Beijing, 1375 pp + 339 B/W photos (in Chinese, Latin-Chinese species lists; no index)
Role of Riparian Zones in Reducing
Pollution of Surface and Ground Water,
Increase Agricultural Production
and Nutrient Acquisition and Storage
in River Catchments

Iuliana Florentina Gheorghe, Daniela Strat, and Simona Mihailescu

Abstract Romania is a country relatively poor in water resources: almost 75 000

millions m3 from which 67 000 millions m3 surface water and 8000 millions m3
ground water. Half of the surface water (68%) of Romania is of the first quality,
while degraded water represents only about 11% of the whole volume. The average
amount of water used in Romania annually is of ca. 9.051 billions m3, of which:
industry 4.823, domestic 2.887, agriculture 1.299 and others 0.042. Irrigation in
Romania is fully controlled. Nutrients input of from agriculture into the surface
waters by percolation from the soil in river basins is high, majority of domestically
wastewater (74%) are not collected and treated [1]. Riparian zones are an important
role in nutrient acquisition and storage reducing pollution of surface water, ground
water and increase agricultural production. In river catchment a green infrastructure
with lakes and rivers, wetland, different types of forest, pastures, shrubs including
different types of crops, it represent the ideal structure to harmonize the develop-
ment and nature conservation. Plants have limited ability in uptake and storing
nutrients, and storage time is different, finally the nutrients reach the litter that is
decomposed. Decomposition and the nutrient cycles are fundamental to ecosystem
biomass production. Most natural ecosystems are nitrogen (N) limited and biomass
production is closely correlated with N turnover [2, 3]. In natural ecosystems,
external input of nutrients is very low and efficient recycling of nutrients maintains
productivity [4]. This chapter presents the effectiveness of different types of riparian
zone in nutrient acquisition and storage with a role in reducing pollution of surface
water, ground water and increase agricultural production.

I. F. Gheorghe (&)
Faculty of Natural Science and Ecology, Ecological University of Bucharest,
Blvd. VasileMilea 1G, Bucharest, Romania
D. Strat
Faculty of Geography, University of Bucharest, Blvd. N. Bălcescu 1, Sector 1,
010041 Bucharest, Romania
S. Mihailescu
Institute of Biology - Bucharest - Romanian Academy, 296 Splaiul Independentei,
060031 Bucharest, POBox 56-53, Romania

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 365
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_18
366 I. F. Gheorghe et al.

Keywords Riparian vegetation
Villages
Small towns
Water pollution

Nutrient storage

1 Introduction

Romania is relatively poor in water resources: almost 75 000 millions m3 from

which 67 000 millions m3 surface water and 8000 millions m3 ground water. The
specific water resource is 3246 m3/inhabitant/year, of which 1650 m3/inhabitant/
year come from Danube River. At national level the diffuse and point pollution
sources are uniformly distributed and polluted water increases the costs of treatment
of its. More than half (68%) of the surface water of Romania is of the first quality
class, while degraded water represents only about one tenth (11%) of the whole
volume. The average amount of water used in Romania annually is of 9.051 billion
m3, of which: the industry is using 4.823, the housing 2.887, the agriculture 1.299
and others 0.042 [1]. After 1989, the trend, in agricultural irrigation water using, is
a decreasing one due to both reductions of the irrigated area and to dismantling of
the huge industrial complexes of livestock breeding. Waste water use in irrigation is
still very limited, on one side due to the reduced facilities for treatment, but mostly
because surface water and ground water resources are enough to satisfy the needs of
the areas that are prone to irrigation. Keeping a good water quality and an increase
of agricultural production in Romania is an important need for political, econom-
ical, moral and scientific reasons. The political reason is that Romania is a member
of European Union and it must to respect European legislation, meaning the Water
Framework Directive objectives in the field of water conservation policy. The aim
of directive is to establish a framework for the protection of intern surface waters,
transitional waters, coastal waters and groundwater. The most important moral
reason is the water quality. A good water quality in sufficient quantity influences the
health condition of the people. The preservation of riparian zone (the interface
between land and a river or a stream), and aquatic ecosystems is also a very
important scientific and economical reason due to a lot of natural resources pro-
duced by these complex ecosystems, including a good water quality. Agriculture
and housing wastewater represent two type of pollution source with nutrients.
Riparian areas can help as buffer zones and reduce (retention, remove) the effects of
non-point source pollution, riparian areas play a key-role in the nutrients flux.
Unfortunately, when the buffer zone capacity has been exceeded a very large
amount of nutrients is discharged in the river water, for this reason we must to know
the buffer zones capacity to stock the nutrients. Nutrient availability in water and
soil is highly heterogeneous in space and time. Consequently, efficient growth up of
vegetation and uptake of nutrient can be strongly influenced by the ability of the
roots system that constantly it does develop, and it’s most absorptive elements in
the most favorable soil, temperature and humidity [5]. One of the major global
issues particularly at European level is the use of fertilizers in agriculture. On the
one hand the need for food requires the use of a surplus of nutrients with role in
Role of Riparian Zones ... 367

increase in agricultural production, on another part, the excess use of these nutrients
leads to soil degradation and pollution of ground water and surface water.
Agriculture, previously dominated by productivity, now has multiple objectives.
The crops low environmental impact, the quality of crop products, the low cost of
production and hence increased nutrients use efficiency, are among these objectives.
Understanding the processes that govern nutrient fluxes, particularly nutrients
uptake and distribution in crops, is of major importance with respect to both
environmental concerns and the quality of crop products [6].

2 Water Resources of Romania

Romania occupies approximately 29% of the Danube basin area (Fig. 1, Table 1)
the total length of rivers is 78,905 km, total volume - 40 billions m3, and almost
1,840 m3/inhabitant.
Almost all-surface water resources originate from the inland rivers and from the
Danube. The average multiannual volume of inland waters amount to an of 42,293
millions m3, where the largest share is held by the Siret (17%), Mures (13.8%) and
Olt (13%) rivers.

Fig. 1 Maps of the Danube River Basin District [8]

368 I. F. Gheorghe et al.

Table 1 The hydrologic network capacity in Romania

Source Total capacity Potential used capacity Capacity used Percent of total use
(millions m3) (millions m3) (millions m3) %
Interiors rivers 40.000.000 13.059.071 3.940.724 9.85
Danube river 85.000.000 20.000.000 4.737.664 5.57
Ground waters 9.600.000 6.677.150 758.628 7.9

At the inlet into the country, the Danube’s multiannual input is of 175,598
millions m3. The water resources of Danube are also available to the neighboring
countries (Yugoslavia, Bulgaria and Ukraine). The Romanian share from the
Danube waters is estimated to 85 000 millions m3 [1, 7].
In years with normal droughty, the total flow by the inland rivers is almost 29
658 millions m3 and represent just 70% of the mean multiannual volume that is 42
293 millions m3. During a year with accentuated droughts the total flow is 22 309
millions m3 and represent just 52.7% of the multiannual one. The Danube displays
a different regime. The contribution of the tributaries from the Alps region preserves
the flow along the river at a high level even during years with high droughty this
being between 84 and 72% of the multiannual flow [7].
On the surface of Romania there are 11 hydrographic basins of relatively equal
size, the largest being of the Siret river (42,890 km2) and the smallest Dobrogea -
Litoral (5,480 km2) (Table 2).
The most populated basin is the basin of Argeş - Vedea River with a relatively
small surface (21,847 Km2). It is one of the most vulnerable the river basin
regarding nutrient pollution and represents the ideal case study for following study.

3 Geography Description of River Basin

The Argeş-Vedea catchment lies on central part of the Romanian Plain. This major
landform unit is located on the left side of the Danube River where it is the border
between Romania and Bulgaria. Tectonically, the Romanian Plain has developed on
the northern part of Moesian Platform, over which overlaps a thick sedimentary
cover [9].
The most recent deposits consist of loess and loess-like deposits, dominantly
composed by silty clay [10–12]. Beneath them are gravels and sands of lower
Pleistocene ages [9] that represent a phreatic stratum which host valuable groundwater
resources. Along the river valleys, the alluvial deposits are composed by fine gravels,
sands and clay. The most distinctive landforms of the Argeş-Vedea catchment are the
tabular interfluves, which have been shaped by the rivers network. The altitudes slightly
decrease from north to south, ranging from 180 to 50 m. Slopes are nearly level and
very gentle, with a mean value of slopes of 1°. These morphologic and morphometric
features of this lowland area are key drivers for rivers meandering. Different
Role of Riparian Zones ... 369

Table 2 Human activities, population density in all 11 river basins present at national level
Name of Surface area Number of Number of city Industry Houses
catchments (km2) inhabitants and villages
1. Someş-Tisa 22,380 2 090,000 243 722 37,200
2. Crişuri 14,860 1 282,800 195 175 80,477
3. Mureş 27,890 2 190,000 238 381 60,167
4. Banat 18,320 3 640,000 349 415 69,213
5. Jiu 18,975 1 638,900 179 122 5,937
6. Olt 24,050 2 676,000 245 154 4,376
7. Argeş-Vedea 21,847 3 942,500 180 457 24,228
8. Buzău-Ialomiţa 26,205 2 604,000 209 220 22,457
9. Dobrogea-Litoral 5,480 680,000 13 - -
10. Siret 42,890 2 792,400 143 328 17,793
11. Prut 20,680 1 821,000 193 210 14,306

generations of old meanders of Argeş-Vedea River can be identified and oxbow lakes
occur in the flood plain. The wide of flood plain ranges from about 100 to 500 m.
Based on geographical location and altitudinal values, in the area of Argeş-
Vedea catchment there is a temperate continental climate, with cold winters hot
summers, huge range of extreme temperatures and quite equal distribution of
rainfall per seasons. With a mean annual sunshine duration of 2254 h, the mean
annual air temperature is 11.1 °C and the annual amount of precipitation is about
532 mm for the period 1961–2013. The mean air temperature of January, the
coldest month, is below 0 °C but the annual number of air frost days fluctuates
from 100 to 120. During the summer, the mean air temperature in July is around
23 °C; the mean maximum is 32 °C and de annual number of days with mean air
temperature over 30 °C is around [13]. The degree of continentality of the climate
in the study area is depicted by value of the Johansson Continentality Index that is
44.6, which means a continental climate [14, 15]. For the same time period, in terms
of the annual De Martonne aridity Index (25.2), the climate can be described as a
semi-humid one. Annually, potential evapo-transpiration (PET) varies between 598
and 718 mm, surpassing the yearly precipitation, but during the winter (December -
February) the PET value is 0 mm, while the highest values are calculated during the
summer months, June–August, with an average that range between 143/month and
147 mm/month, much higher than precipitation amount (Fig. 2).
As a result, during the growing season and especially during the summer months
there is water deficit. Regarding nebulosity, during the whole year, on average there
are 135 clear sky days, with those of 117 cloudy sky and the sky covered with those
of 123. During the study period (2008–2013), the average annual air temperature in
the sampled catchment area was higher (11.9 °C) that in 1961–2013 period and the
amount of precipitations was 529.6 mm. Rainfalls were distributed almost equally,
but in terms of quantity, the summer is the rainiest season (117.7 mm) with the
peak precipitations in June. Figure 3 depicts the temperature-precipitation diagram
370 I. F. Gheorghe et al.

Fig. 2 The monthly air temperature –precipitation diagram during 2008–2013 in the Argeş-Vedea
catchment

at the meteorological station that is located at the central-western part of the Argeş-
Vedea catchment. The climate diagram shows that only two months—August and
September are characterized as climatic arid according to Gaussen classification.
However, the value of the summer drought index, calculated as the ratio between
summer precipitation amount and mean maximum air temperature of the hottest
month [16], that is 5.3, emphasizes that in the study area the summers are droughty.
The annual De Martonne index was 24.1, which depicts a semi-humid climate but
compared to the multiannual value (1961–2013) it is a threshold change, from forest
steppe, previously, to steppe grassland during the study period time. The mean
distribution of the monthly De Martonne aridity index shows the maximum climate
diversity—all seven climate types (Fig. 3), according to De Martonne classification
[17, 18], from extremely humid, in January, to arid, in August. The greatest diversity
of climate types is during summer because each month has different characteristics,
humid, semi-humid and arid respectively. Overall, De Martonne aridity index value
was 21, which correspond to the Mediterranean climate.
According to climate conditions and natural vegetation, in the area are developed
fertile soils that belong tochernozem types and red brown main types respectively
with the following particular subtypes: typical red brown, luvicverticred brown,
luvicpseudogleyed red brown, gleyed red brown, gleycpodzolicchernozem-like
soils, and leached chernozem.
Role of Riparian Zones ... 371

Fig. 3 The monthly distribution of the mean of De Martonne aridity index, from 2008 to 2013 in
the Argeş-Vedea catchment

4 Land Use

About About ¾ of the Argeş-Vedea catchment (as the other river basins present in
the Romanian Plain) area is covered with agricultural land (72.5%), this is followed
by the land covered with various constructions, (13%), it represent the rural areas
and consists of villages. The forests occupy 12% of the territory, wetland 2%, rivers
and lakes 0.5% (Fig. 4). Although the forest area is not very large there is a great
diversity of temperate continental types of forest. Dominant tree species are of the
genera Quercus, Salix, Populus, Alnus, Acer, Ulmus, Fraxinus, etc. [19, 20] https://
www.ijese.org/download/volume-2-issue-5/.
22.4% of this field is present in the riparian zone and covered with these types
of vegetation: wetland with Carex sp. Lythrum sp. Scyrpus sp.; wetland with Salix
sp. Phragmites and Typha sp., Scyrpus sp.; pastures; meadows, grasslands, orch-
ards; forest (24 typologies); crops: wheat (81%), sunflower (3%), maize (11%), rape
(5%) (Table 3).
For maps of land use were used digitized maps “Corine land cover 2006”, being
processed in the ArcView GIS 3.2 a, example Fig. 5,
372 I. F. Gheorghe et al.

Fig. 4 Type of land cover at catchment level (ha)

Table 3 Types of riparian zones in rivers catchments

Structure of Dominant species plants Longer Surface % of different
riparian (km) (ha) land type present
zone in riparian zone
Wetland Carex sp., Lythrum sp., 6.3 69.3 77%
Scyrpus sp.
Wetland Typha sp., Scyrpus sp. and 1.3 7.8 8.7%
with Salix Salix sp.
sp.
individuals
Village 1.8 19.8 3.4%
Forest Robinia pseudaccacia, 6.9 350.9 65%
Quercus cerris, Q. frainetto,
Q. robur, Q. petraea,
Fracxinus excelsior,
Populus nigra,
P. alba,
P. tremuloides x canadensis,
Acer campestre, Ulmus laevis,
Carpinus betulus, Alnus
glutinosa, Tilia cordata,
Salix fragilis, S. nigra
Crops 10.4 560 17.2%
Pasture 0.85 5.9 1.12%
Rivers and 26.7 26.7 100%
lakes
Role of Riparian Zones ... 373

Fig. 5 Land use in river basin (Corine land cover map 2006) [19]

5 System of Agricultural Crops

Using the information from National Statistical Yearbook for 2007 generated by
Agency of Payments and Interventions in Agriculture and Land Register Book, has
been evaluated the crops system in Romania considering the following terms:
Agricultural area - the lands with agricultural destination, owned by natural or
legal persons, classified as follows: arable land, natural pastures and hayfields,
vineyards and vine nurseries, orchards and tree nurseries.
Arable land represents area which is ploughed each year or at several years,
cultivated with annual or perennial plants.
Pastures represent lands covered with herbal vegetation, grown in natural way,
or regenerated by sowing, for animal pasturing.
Hayfields represent lands covered with herbal vegetation, grown in natural way,
or regenerated by sowing, for hay harvesting.
Vineyards and nurseries represent areas with vineyards, vine nurseries and land
prepared for vineyards.
Orchards and tree nurseries represent areas with tree plantations, fruit
younglings, tree nurseries and land prepared for orchards [7].
Cultivated area represents the sown area in agricultural year (October 1 -
September 30) and includes:
– sown area in the previous autumn excluding re-sown area in spring with other crops;
– sown area in spring of current year.
374 I. F. Gheorghe et al.

Agricultural holding represents economic unit of agricultural production carry-

ing out its activity under a current unique management and includes all the animals
owned and all the land area used partly or completely, to perform an agricultural
production, no matter of ownership type, legal type or size.
Crop agricultural production represents gross harvested production, minus losses
at harvest and includes:
– production in own field;
– production of combined crops;
– production of successive crops;
– production obtained in kitchen gardens (only for vegetables, fruit, grapes) [7].
The number of types of crops at the national level is 36 of which 25(69.4%) are
found in the river basins, which represents a great diversity.
The largest area is occupied by cereals for grains—wheat and corn—(62.7%),
most crops are annual, and for the most of it, the sowing period, is the spring. The
area occupied by industrial crops has declined lately from 24.92% to 16 0.41%, the
rapeseed culture was introduced, replacing the sunflower (Fig. 6).

6 Nutrient Input

The point sources of pollution in surface water, at national level are uniformly
distributed.
In Romania the diffuse and point sources of pollution are uniform distributed
except the hilly and mountain areas where the agriculture activities are reduced and
the human population is very low. The most vulnerable zones in pollution by
nitrates are: North and North-East of Moldova, and South-Eastern Wallachia, the
largest areas are in proximity of Bucharest.
In basin of most rivers are two sources of diffuse pollution: untreated sewage and
fertilizers used in agriculture. Both sources of diffuse pollution create an additional
intake of nutrients. At national level the population connected to sewerage is
between 6.36% and 67.64% with mean value 34.9%.
In the whole area of the country, the annually agriculture input with quantities of
nutrients in excess varies in different basins in range 0.39–8.7 kg/ha P; 6.91–
23.6 kg/ha N; with the most large quantity in the area (Argeş-Vedea’s Basin)
23.6 kg/ha N. The quantities of nitrogen and phosphorus in excess coming from
agriculture lands and accumulated in the period 1998–2000 was between 60 and
87 kg/ha/year P; 12–91 kg/ha/year N, with the large quantity in Argeş-Vedea’s
Basin (91 kg/ha/year N).
 Role of Riparian Zones ...

Fig. 6 Type of crops, surface (%)

375
376 I. F. Gheorghe et al.

6.1 Input of Nutrient by Fertilizers

In Argeş-Vedea basin not all agricultural land is fertilized the fertilizers are used
only for three main crops (maize, wheat and sunflower). 3% of the cultivated area of
grain maize is fertilized with mineral NPK (75 kg N/t, 130 ka P/t and 175 kg K/t)
and 10% with organic fertilizer (manure). Surface of the wheat crop fertilized with
NPK is about 20% and 5% with the manure is. For fertilization of sunflower crops
are used only NPK and only 30% of cultivated area being fertilized. Manure comes
from raising cows and has a nitrogen content of 5 kg N/t and 0.49 kg P/t (mean
values in dry substance). For fertilization with NPK the farmers use between 250
and 300 kg/ha, and amount of manure used as fertilizer is between 10 and 15 t/ha.
Table 4 [19, 20] https://www.ijese.org/download/volume-2-issue-5/.

6.2 Input of Nutrient by Untreated Sewage Water

In most polluted Argeş-Vedea basin, are 6 villages (Negrişoara, Glavacioc, Şelaru,

Cătunu, Buteşti, Purani) and a small town (Ştefan cel Mare) situated along the river
course. The population is supplied with water from the river Arges and groundwater
aquifers. The N content in both sources have an average of 15 mg/L and the P
content is 2.5 mg/L. After water use in the house hold, N and P contents increase at
values of 25 mg/lN and 3.5 mg/lP (Table 5) [19, 20] https://www.ijese.org/
download/volume-2-issue-5/.
Comparing the two sources of input of nutrient it can be noted that most is owed
by the fertilization of crops. Total nutrients input introduced by fertilization of crops
was 15 483 ka N and 13 206 kg P. Water used by the population and untreated
made an annual intake of 8577 kg N and 1299 kg P, the contribution of population
is 3169 kg N and 632 kg P. Although a small area of agricultural crops is fertilized
and the amount of fertilizers used per area is relatively small these fertilizers pro-
duce a significant increase in soil nutrients. Intake of nutrients coming from

Table 4 Quantity of nutrients input in river catchment by fertilizers

Crop type Fertilizer Fertilized Quantity Quantity Quantity Total Total Total
type surface (ha) of fertilizer of TN of PT quantity quantity quantity
(t/ha) (kg/ha) (kg/ha) of of TN of TP
fertilizer (kg) (kg)
(t)
Maize NPK 31.29 0.275 20.62 35.75 8.6 645 1119
Manure 104.3 10 50 4.9 1043 5215 511
Wheat NPK 166 0.300 22.5 39 49.8 3735 6474
Manure 41.6 15 75 7.35 624 3120 306
Sunflower NPK 123 0.300 22.5 39 36.9 2768 4797
Total 466.19 25.875 190.62 126 1762.3 15,483 13,206
Role of Riparian Zones ... 377

Table 5 Quantity of nutrients input in river catchment by untreated sewage water

Locality No. of Water volume Water total Content of nutrients Total input
inhabitants used/ volume in wastewater and (kg/year) of nutrients,
inhabitant used difference between increase of nutrients added
(mean value)/ /year the nutrient content of to water using
month (m3) sewage and water
(m3) supply
(mg/L) mean values
mg/l N mg/l P Kg/year N Kg/year P
Negrişoara 796 1.7 16,238.4 24 9 3.05 1.2 390 146 50 19
Glavacioc 814 2.1 20,512.8 26 11 3.35 1.5 533 226 69 31
Şelaru 2140 3.1 79,608 28 13 3.65 1.8 2229 1035 291 143
Cătunu 927 2.1 23,360.4 24 9 3.45 1.6 561 210 81 37
Buteşti 885 2 21,240 21 6 3.15 1.3 446 127 67 28
Purani 1685 2.8 56,616 25 10 4.15 2.3 1415 566 235 130
Ştefan cel 3405 3.5 143,010 21 6 3.55 1.7 3003 858 508 243
Mare
Total 10,652 17.3 272,638.8 8577 3169 1299 632

fertilizers compared with the intake of nutrient from untreated domestic water is 5
times higher in case of N and 20 for P. Therefore the policy of protection of surface
water should be focusing on the sources of diffuse pollution from agriculture and
not on the wastewater. In Romania’s strategy to reduce pollution of surface water
that is focusing on the requirements of European Water Framework Directive the
diffuse sources in agriculture are not taken into account [19, 20] https://www.ijese.
org/download/volume-2-issue-5/.

7 Vegetation Description of River Basin

Taking into account physico-geographical conditions, the catchment area is located

in forest-steppe biome but because of the agricultural practice the natural vegetation
is replaced by agricultural lands. The remained patches of forest are dominantly
composed by species of the genus Quercus, in particular thermophilic species like
Quercus cerris, Q. frainetto. In in composition of this forest are present also
Fraxinus ornus Acer campestre, Ulmus laevis, Carpinus betulus, Fraxinus excel-
sior, and Tilia cordata.
Additionally, the floodplain forests are composed by Populus nigra, P.alba,
Alnus glutinosa, Salix fragilis, S. alba. and S. nigra.
378 I. F. Gheorghe et al.

The vegetation is part of subunits: U22 - Southern European forests of willow,

black and white poplar (Salix alba, Populus alba, P. Nigra);
U14 - Panonic - Carpathian Meadow Forests (Quercus robur, Fraxinus
angustifolia) in complex with poplar plume and willow (Populus alba, P.nigra,
Salix alba); and U15 - Danube-Pontic forests of meadow (Fraxinus angustifolia, F.
pallisae, Quercus robur, Q. pedunculiflora)
In subunit U14 and U15 the herbaceous layer consists of hydrophilic species
such as Rubus caesius, Glechoma hederacea, Lysimachia nummularia, Galium
mollugo, mesophilous species such as Asparagus tenuifolius, Veronica chamaedrys,
Potentilla reptans, Geum urbanum, ruderal species such as Amaranthus retroflexus,
Arctium lappa, Bromus sterilis, Capsella bursa pastoris, Cirsium vulgare, Conyza
canadensis, Daucus carota, Setaria pumila, Sonchus asper, Xanthium strumarium,
Cichorium intybus, Cirsium arvense, Cirsium canum, Sonchus arvensis, the edi-
fying species encountered in this area are characteristic of both vegetation subunits.
The arboricol layer is represented by populations belonging to the species
Quercus robur, Q. cerris, Q. pedunculiflora, Fraxinus excelsior, F. ornus, Populus
nigra, Acer campestre, A. tataricum, Ulmus laevis (Quercus robur, Fraxinus
excelsior, F. ornus, Populus nigra, Acer campestre), with the exception of species
of the genus Fraxinus, the rest are typical of U14, U15 vegetation subunits.
The shrubs layer consist in Cornus sanguinea, C. mas, Crataegus monogyna,
Evonymus europaeus, Ligustrum vulgare, Rosa canina, Rubus caesius, Salix
caprea, Salix triandra, Viburnum opulus, Cornus mas, Corylus avellana, Prunus
spinosa, and contains species of shrubs characteristic of both U14 and U15
subunits.
In subunit 22 the grassy layer consists of hydrophilic species such as: Bidens
tripartitus, Polygonum hydropiper, P. mite, Galium palustre, Stachys palustris,
Methha aquatica, Lycopus europaeus, Scutellaria hastifolia, Iris pseudacorus,
Lythrum salicaria, Solanum dulcamara; besides these edifying species, there are
also species such as: Rorippa silvestris, Gratiola officinalis, Lysimachia nummu-
laria, Eleocharis palustris, Juncus effusus, Veronica anagallis-aquatica, and due to
the anthropic activities that occur in this area (grazing, grubbing, introduced into the
vegetal and ruderal communities such as Amaranthus retroflexus, Arctium lappa,
Atriplex patula, Bromus sterilis, Capsella bursa-pastoris, Carduus nutans,
Chenopodium album, Conyza canadensis, Datura stramonium, Daucus carota,
Galium aparine, Lamium purpureum, Setaria pumila, Solanum nigrum, Sonchus
asper, S. oleraceus, Verbascum blattaria, Xanthium strumarium, Aristolochia
clematitis, Ballota nigra, Cichorium intybus, Cirsium arvense, Elymus repens, etc.
The trees layer consists of three species: Salix alba, Populus alba, P. nigra, and
the grassy layer is dominated by Rubus caesius.
Arges- Vedea basin belongs to the boundary between subunits U14, U15 and U22.
Seven types of plant communities have been identified: Molinio-Arrhenatheretea,
Fraxino oxycarpe-Ulmetum, Fraxino-pallisae-angustifoliae-quercetum roboris,
Populeto-Salicetum, Salicetum albae-fragilis, Phragmition, Populion albae.
Role of Riparian Zones ... 379

7.1 Types and Structure of Vegetation

Riparian vegetation is important for the health of waterways, contributing to the

balance of oxygen, nutrients and sediment, and providing habitat and food for
fauna. It grows along banks of a waterway extending to the edge of the floodplain
(also known as fringing vegetation). This includes the emergent aquatic plants
growing at the edge of the waterway channel and the ground cover plants, shrubs
and trees within the riparian zone [19, 20] https://www.ijese.org/download/volume-
2-issue-5/. Depending on the degree of representation, the most important are five
riparian vegetation types: wetland (W) with Carex sp. Lythrum sp., Scyrpus sp.,
pasture (P), mixed forest (F1), forest with Quercus species (Querceta) (F2) and
agriculture land (A) (wheat, sunflower, corn crops, etc.). For each zone were
estimated: structure of vegetation, dominant species, biomass, primary productivity,
C, N stocks and C, N uptake.
The vegetation of wetland zone (W) is homogenous (the SD of the cover degree
and height not vary more than ±5% and ±7 cm), the species richness value is low
(13 taxa). In plain forested area, at catchment river level, are present 24 typologies
of forest with nine types. Because the largest surface of forested area is covered with
two forest types (mixed and querceta), the study of biomass and storage capacity
was carried out in these two forest types. The mixed forest (F1) is a natural forest,
plurien, with a great vitality, a complex structure (16 trees species) and an average
productivity. The querceta forest is (F2) a young and natural forest, with a great
vitality, with a structure a medium complexity (12 trees species) and a high pro-
ductivity. Concerning the structure of pasture vegetation, the species richness value
is highest compared with other herbaceous layer present in the other four vegetation
type (32 taxa), the dominate specie is Elymus repens. Except for the dominant
species individuals of other species are equitable distributed. The herbaceous layer
has a low heterogeneity (cover degree and hight of vegetation varies slightly around
the average and SD has following values of: ±5% and respectively ±5 cm)
Table 6 [19, 20] https://www.ijese.org/download/volume-2-issue-5/. The most low
values of species richness is in agricultural land, here, compared with the herba-
ceous layer present in the other four vegetation types (7 taxa), the dominate species
is Triticum aestivum. Except cultivated plants, ruderal plants are extremely
underdeveloped. The herbaceous layer is very homogeneous (cover degree in all
quadrates was 100% and height of vegetation varies slightly around the average, SD
of is ±5 cm). In terms of plants specific composition, the species present in
common, in all type of vegetation represents less than 17%. Although mixed forest
(F1) and querceta forest (F2) belong to the same vegetation unit, specific compo-
sition is completely different (less 20% taxa is in common). Vegetation on agri-
cultural land (A) has a very low number of taxa and is similar with the pasture (P).
The wetland vegetation (W) has no species in common with the other areas (A, F1,
F2 and P) (Fig. 7).
380 I. F. Gheorghe et al.

A
P
F2
F1
W
-0.2 0 0.2 0.4 0.6 0.8 1
Jaccard's Coefficient
Fig. 7 Jaccard similarity between the species composition in agriculture land (A), pasture (P),
querceta forest (F2), mixed forest (F1) and wetland (W) [20] https://www.ijese.org/download/
volume-2-issue-5/

7.2 Biomass–Primary Production and Productivity C, N

Stocks and Nutrient Uptake

The largest quantities of biomass are produce by mixed forest following by querceta
forest, the layer of trees has significant contribution. The herbaceous layer present
in mixed forest is least productive followed by the one from the querceta forest;
because there it is increased competition for space, light and nutrients.
Nitrogen content in soil of riparian zone varies inversely proportional with
altitude of land. In agriculture land where the altitude is high the content is low and
in wetland where the altitude is low the content is very high. The low content of
nitrogen from agricultural land, pasture and forest with low slope is due to takeover
by plants, here the oxygenation and humidity conditions of the soil not favor the
removed of nitrogen by denitrification. In wetland where the content of nitrogen is
high the denitrification conditions are favorable because there is enough substrate
for the mineralization processes. In wetland and forest with low slope the fertil-
ization is inefficiently because there not produce an increase of content of nitrogen
in soil like, in crops land, pasture and forest with high slope. In all 5 zones the
nitrogen content in belowground biomass is higher compared to aboveground
biomass. Except wetland, the growth trend of nitrogen content in the other zones is
similar. In both type of the forest the nitrogen content in aboveground biomass, in
July, is high because the plants present in herbaceous layer are at maturity, at end of
biological cycle. The plants present in herbaceous layer in the forest he ends his
biological cycle early to avoid competition for light and nutrients with trees leaves.
Nutrient availability in soil is highly heterogeneous in space and time (Fig. 8). Soil
resources are unevenly distributed in space and time. Water availability can dra-
matically affect soil nutrient availability, root physiology, and plant nutrient
acquisition. If respiration does not decrease at a rate similar to those of water or
Role of Riparian Zones ... 381

nutrient uptake, then the roots become more costly to the plant. Moreover, dry
surface soil is often associated with high soil temperature, potentially further
increasing root costs [21]. Typically, if soil resources such as nitrogen or phos-
phorus are limiting plant growth, plants increase the amount of root biomass
allocation, and thus help maintain a “functional equilibrium” between shoot
acquisition of C and root acquisition of mineral nutrients [22]. In an extensive
review of plant is specified the responses are non-uniform supplies of nutrients,
[23]. Production of root hairs or extrametrical mycorrhizal hyphae can be a very
efficient way by which a plant can increase absorptive surface for the same biomass
allocation. Higher efficiency does not always lead to higher plant fitness. In a
competitive environment where resources are available only for short periods, rapid
resource acquisition, rather than high efficiency, may be a key to plant success [24].
High expenditures for rapid resource acquisition may be an ecologically effective
strategy if fitness of neighbors is diminished to a greater extent than is the fitness of
the individual exhibiting rapid growth. Plants may also overproduce tissues as a
means of coping with herbivore, or as insurance against extreme events. The notion
that plants might not be efficient in resource use was underscored in a review by
Thomas and Sadras (2001) [25].They argue that there may be many instances where
plants may support large numbers of “unproductive” tissues that may provide
secondary benefits for N storage, as a buffer against herbivore, and as a way of
offloading excess C and other nutrients. For example, Thomas and Sadras (2001)
[21, 25] speculate that plant species in fertile environments may exhibit high rates
of leaf and root turnover, not in response to a reduced need for nutrient conservation
[21, 26, 27] but rather because of a greater need to offload excess resources
associated with overproduction of carbohydrates. Consistent with the excess tissue
hypothesis of Thomas and Sadras (2001) [21, 25] are the arguments that plants may
use the alternative respiratory path as an “energy overflow” pathway [21, 28], and
the evidence from the elevated CO2 literature that shows an average of 42%
enhanced soil respiration (root plus microbial respiration) in response to elevated
CO2 with-out an increase in shoot growth [29]. Our view is that while there may be
times when plants appear “wasteful”, especially over short time spans, maintaining
redundant absorptive tissues to offset the risks of herbivory or extreme weather
events still follows general concepts of optimization in an uncertain environment.
Consequently, broad economic analogies [30] of plant resource acquisition and
allocation may be useful tools as a first approximation for interpreting plant
responses to multiple resource limitations and strategies for tissue deployment. One
of the major global issues particularly at European level is the use of fertilizers in
agriculture. On the one hand the need for food requires the use of a surplus of
nutrients with a role in increasing agricultural production, on another part of the
excess use of these nutrients leads to soil degradation and pollution of groundwater
and surface water. Nitrogen uptake and accumulation in crops represent two major
components of the N cycle in the agro-system. Nitrate ions that not taken up by a
crop, in most of them are infiltrate in underground water. Modeling N uptake
together with soil water transfers is, therefore, key in quantifying and preventing
nitrate leaching [31]. 20% of the phosphorus that coming from fertilizations and
382 I. F. Gheorghe et al.

Comparison between the biomass, C, N stocks and uptake in vegetation

120000

100000

80000
W
kg/ha
F1
60000
F2
P
40000
A

20000

0
Dry bimass C stock N stock C uptake N uptake

Fig. 8 The biomasses, C, N stocks and uptake in vegetation [39] https://www-pub.iaea.org/

MTCD/Publications/PDF/TE_1784_web.pdf

50% of applied nitrogen to land reach receiving waters [3]. So far studies have been
conducted only in the field of crop plants, but the wild plant communities that
uptake part in excess of N are not been studied. Also is not been studied the role of
wild plant communities, that function as buffer zones leading to reduction of
nutrient pollution of ground and surface water. Plants have limited ability in
retention and storing nutrients, and storage time is different, finally they reach the
litter, litter that is decomposed.

7.3 Decomposition and Nutrient Cycling

The litter represents the dead plant parts that are in decay, which will decompose.
The amount of litter is strictly related f biomass that is transformed into necro-mass.
In ecosystems where no are trees and shrubs, the litter is represented by the parties
dead of grass. In forests and shrubs areas the litter is represented largely by leafs.
Decomposition and nutrient cycling are fundamental to ecosystem biomass pro-
duction. The productivity of most natural ecosystems, are limited by nitrogen
(N) input and biomass production is closely correlated with N turnover [2, 3].
Typically external input of nutrients is very low and efficient recycling of nutrients
maintains productivity [4]. Decomposition of plant litter ensure for the majority of
nutrients recycled through ecosystems. Rates of plant litter decomposition are
highly dependent on litter quality,high concentration of phenolic compounds and
Role of Riparian Zones ... 383

Table 6. Species composition and dominance in riparian vegetation types

Frequency % Braun- Dominance
Species Blanquet
index
Plant species composition and dominance in wetland vegetation (W)
Typha latifolia 2 +
Stachys palustris 5 1
Scirpus lacustris 100 5 First dominant
S. sylvaticus 15 2
Lycopus europaeus 0.5 R
Phragmites australis 2 +
Lythrum salicaria 69 4 Second dominant
Ranunculus acris 1 +
Galium palustre 1 +
Epilobium hirsutum 35 3
Juncus glomeratus 2 +
Carex pseudocyperus 56 4 Second dominant
Acorus calamus 0.75 R
Plant species composition and dominance in mixed forest(F1)
Trees Shrubs Herbaceous
Fraxinus excelsior Cornus mas Bromus sterilis 5 +
Fraxinus ornus Cornus sanguinea Buglossoides purpurocaerulea 28 3
Fraxinus pennsylvanica Corylus avellana Galium schultesii 34 3
Acer campestre Crataegus monogyna Glechoma hederacea 67 4 Second dominant
Prunus cerasifera Prunus spinosa Lolium perenne 2 +
Pyrus pyraster Rosa canina Lysimachia nummularia 14 2
Acer tataricum Rubus caesius Plantago major 7 1
Malus sylvestris Viburnum opulus Plantago media 9 1
Quercus cerris Ligustrum vulgare Ranunculus acris 38 3
Q. pedunculiflora Evonymus europaeus Taraxacum officinale 11 2
Q. frainetto Salix triandra Erigeron canadensis 78 5 First dominant
Q. robur Geranium phaeum 4 +
Ulmus laevis Asperula glauca 6 1
Robinia pseudaccacia Alliaria officinalis 23 2
P. tremuloides x P. Stellaria aquatic 4 +
canadensis
Populus nigra Mercurialis perennis 0.75 r

Plant species composition and dominance in Querceta forest (F2)

Trees Shrubs Herbaceous
Fraxinus excelsior Cornus mas Anemone nemorosa 17 2
Fraxinus ornus Cornus sanguinea Buglossoides purpurocaerulea 6 1
Fraxinus pennsylvanica Corylus avellana Asparagus tenuifolius 3 +
Acer campestre Crataegus monogyna Corydalis cava 21 2
Populus alba Rosa canina Corydalis solida 24 2
Prunus cerasifera Rubus caesius Circaea lutetiana 0.5 R
Pyrus pyraster Viburnum opulus Galium mollugo 11 2
Quercus cerris Ligustrum vulgare Galium schultesii 9 1
Q. pedunculiflora Geranium phaeum 7 1
Q. frainetto Geranium pretense 4 +
Q. robur Geum urbanum 29 3 Second dominant
Tilia cordata Glechoma hederacea 58 4 First co-dominant
Heracleum sphondylium 1 +
Lamium album 2 +
Lysimachia nummularia 14 2
Plantago major 9 1
Polygonatum latifolium 3 +
Potentilla reptans 2 +
Ranunculus acris 28 3
Ficaria verna 65 4 First co-dominant
Rumex crispus 0.7 R
Salvia nemorosa 2 +
Scilla bifolia 18 2
Silene vulgaris 12 2
384 I. F. Gheorghe et al.

Veronica chamaedrys 8 1
Viola odorata 46 3 Second dominant
Plant species composition and dominance in pasture (P)
Amaranthus retroflexus 4 +
Arctium lappa 2 +
Bromus sterilis 3 +
Capsella bursa-pastoris 3 +
Daucus carota 1 +
Echinochloa crus-galli 3 +
Erigeron canadensis 11 2
Erodium cicutarium 0.5 R
Setaria pumila 0.7 R
Achillea millefolium 1 +
Cichorium intybus 2 +
Galega officinalis 0.5 R
Hypericum perforatum 1 +
Lolium perenne 23 2
Lotus corniculatus 9 1
Mentha longifolia 3 +
Plantago major 4 +
Potentilla reptans 1 +
Prunella vulgaris 16 2
Ranunculus acris 2 +
Senecio jacobaea 3 +
Taraxacum officinale 5 +
Trifolium hybridum 7 1
Trifolium pratense 8 1
Trifolium repens 11 2
Dactilis glomerata 4 +
Elymus repens 76 5 Dominant
Vicia cracca 11 1
Cirsium vulgare 0.4 R
Inula britanica 1 +
Ranunculus sardous 1 +
Sonchus asper 0.7 R
Plant species composition and dominance in agriculture land
Triticum aestivum or (Helianthus annuus) 100 5 Dominant
Cirsium vulgare 1 +
Setaria pumila 0.5 R
Sonchus asper 1 +
Bromus sterilis 2 +
Viola tricolor 0.5 R
Vicia cracca 6 1

especially the lignin, in plant litter has a retarding effect on litter decomposition [32,
33]. Globally, rates of decomposition are mediated by litter quality and climate
[37]. Ecosystems dominated by plants with low-lignin concentration often have
rapid rates of decomposition and fast nutrient cycles [38].
The decomposition of the litter is a process that follow an exponential law and
whose rate can be appreciated by calculating the constant K.
Values of decomposition rate constant (k) are comparables with values present in
literature. The highest value of the constant decomposition rate (k) was recorded in
forest with low slope—F1—mixed forest; here the decomposition process is most
intense. In forest with high slope—F2—querceta forest the k value is close to that
of F1. High speed of decomposition in F1 is due a saturated soil in water (here is
sufficient water like humidity necessary for bacterial exo-enzymes activity) and the
nature of litter (the quantity of lignin and cellulose in trees and shrubs leafs is low to
compare the wheat stems and Scyrpus sp. Typha sp.) (Table 7).
Role of Riparian Zones ... 385

Table 7 Comparison between average values of k with average values present in literature
Zone K (days−1) mean values K (days−1) mean values present in literature
–2
W 2.281  10 2.464  10–2 Gessner et al. [34]
F1 5.327  10–2 2.354  10–2 Nelson et al. [35]
P 4.283  10–2 1.044  10–2 Nelson et al. [35]
A 2.578  10–2 0.332  10–3 Salamanca et al. [36]
F2 5.061  10–2 0.367  10–3 Aerts [27]

The lower value of k was calculated in wetland; plant species present here are
adapted to high soil moisture conditions; in tissues structure of these plant the cell
wall are impregnated with silica salts that is difficult to break down in small
fragments. The wheat straws have also in the structure the tissues impregnated with
silica salts; therefore the value of k in agriculture land is low and similarly with
wetland. In both agricultural land and in wetland the decomposition process takes
place slowly. More complex compounds of C are decomposed more slowly and
may take many years to completely breakdown.

8 Conclusions

Riparian zones are the transitional areas between land and water, including the
margins of streams, rivers, lakes, and wetlands. They are rich in biodiversity and
play an important role in protecting water quality and stream ecosystem health.
Riparian vegetation functions as a large sponge that reduces surface flow and
absorbs the nutrients (C, N, P) in excess and pollutants from storm water runoff
During the development process, riparian areas are degraded when vegetation is
removed, the terrain is cultivated or plowed, are installed the utilities, are built the
different structures and river borders are regularized. These changes to the land-
scape and subsequent human activity in the riparian zone have consequences on
ecosystem health from impact of nutrients in excess that came from fertilizers,
wastes, atmospheric pollutants generated by cars the roads, and soils degradation
[40]. The temperate forests play a very important role in terms of the amount of C
and N stored, storage period of them and fertility of the soil.
In trees layer the amount of nutrients (C, N) stored in the wood, is 10 times
greater than that stored in the leaves. The nutrients stock that accumulate in wood as
productivity, grow from year to year and only the leaves supplies the litter, which
decomposes. In the two forest types (F1 and F2), 4/5 the amount of litter is
decomposed, and 1/5 accumulates at the soil surface and supplies the horizon 0 of
the soil with organic matter. This organic matter plays a fundamental role in soil
processes; this is an energy source of microorganisms and precursor in soil humic
acids [19, 20, 39] https://www-pub.iaea.org/MTCD/Publications/PDF/TE_1784_
web.pdf. Although part of N is lost in the process of denitrification,one of the final
386 I. F. Gheorghe et al.

stage of process of decomposition; a large amount of N returns to soil as nutrients

from which is taken by plants. Making a unilateral analysis regarding takeover
efficiency and nutrient use by plants in crops, imply the risk of ignore the role of
other types of ecosystem in nutrients cycle.
A holistic approach, those make a simultaneous analysis for all function of
ecosystems (reducing pollution, creating local microclimates, etc.) outside the
production of them can give an overview and help making the best decisions
concerning the use of different types of land. Natural and semi-natural ecosystems
are the main sources in the production of resources and energy generation and play
an important role in reducing of pollution. In modern society the required of
resources and energy to developed is greater, also the human pressures exerted on
ecosystems and biodiversity is in increase, which implies the need for preservation
of these. Keeping an ecosystem mosaic structure is an ideal solution to harmonize
the development of society with nature conservation. A green infrastructure with
lakes and rivers, wetland, different types of forest, pastures, shrubs including dif-
ferent types of crops, it represent the ideal structure to meet both goals [19, 20, 39].

Acknowledgements Many thanks to Lee Heng officer of the funded program by International
Atomic Energy Agency - Soil and Water Management & Crop Nutrition Section, Joint FAO/IAEA
Division of Nuclear Techniques in Food and Agriculture, in the research program called
Investigation of Nitrogen pathways in typical entropic-modified riparian zones of the Arges river
catchment’s area, Romania, Contract no. 15352.

References

1. National institute of statistics, National Statistical Yearbook for 2007 [Internet] (2007).
https://www.insse.ro/cms/files/Anuar%20arhive/serii%20de%20date/2007/ASR_2007.pdf
2. Vitousek PM, Howarth RW (1991) Nitrogen limitation on land and in the sea: how can it
occur? Biogeochemistry 13:87–115
3. Reich PB, Grigal DF, Aber JD, Gower ST (1997) Nitrogen mineralization and productivity in
50 hardwood and conifer stands on diverse soils. Ecology 78:335–347
4. Likens GE, Bormann FH, Johnson NM, Fisher DW, Pierce RS (1970) Effects of forest cutting
and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-ecosystem. Ecol
Monogr 40:23–47
5. Eissenstat DM (2005) Nutrient acquisition by plants: an ecological perspective. In: The
Efficiency of Nutrient Acquisition over the Life of a Root, pp. 185–220. https://doi.org/10.
1007/3-540-27675-0_8
6. Tosens T, Laanisto L (2018) Mesophyll conductance and accurate photosynthetic carbon gain
calculations. J Experimental Botany 69(22):5315–5318. https://academic.oup.com/jxb
7. Institutul Naţional de Statistic (2015) România în cifre – ISNN 2066–4079. https://www.
insse.ro/cms/files/publicatii/Romania_in_cifre_2015final.pdf
8. Maps of the Danube River Basin District Management Plan 2015, [Internet] (2015). https://
www.icpdr.org/flowpaper/viewer/default/files/nodes/documents/drbmp-update2015-maps.pdf
9. Ionesi L (1994) Geologia unităţilor de platformă şi a orogenului Nord-Dobrogean. Editura
Tehnică, Bucureşti, p 278
10. Conea A, Ghiţulescu N, Vasilescu P (1963) Consideraţii asupra depozitelor de suprafaţă din
Câmpia Română de Est,- Studii tehnice şi economice. Seria C 11:61–85
Role of Riparian Zones ... 387

11. Conea A (1970) Head deposits in the Romanian Danube Plain. Studii Tehnice Şi Economice
Seria C 18:251–268
12. Jipa DC (2014) The loess-Like deposits in the Lower Danube Basin. Genetic significance,
Geo-Eco-Marina, vol 20, pp 7–18
13. Călinescu G (2008) Temperaturi extreme caracteristice. In: Clima R, Ion S, Pescaru VI,
Poiană II, Geicu A, Cândea I, Tâştea D (eds) Bucureşti, Editura Academiei Române, pp 184–
196
14. Gorczynski W (1920) Sur le calcul du degré de continentalisme et son application dans la
climatologie. Geographie Annaler 2:324–331
15. Deniz A, Toros H, Incecik S (2011) Spatial variations of climate indices in Turkey. Int J
Climatol 31:394–403
16. Lebourgeois F (2010) Cours de bioclimatologie a l’usage des forestiers. Nancy:
Departement SIAFEE. UFR Forets, Abres et Milieux Naturels
17. Baltas EA (2010) Surface representation of climate variables and indices in Greece using GIS
Methods. Int J Meteorol 35(348):123–136
18. Croitoru A-E, Piticar A, Imbroane AM, Burada DC (2013) Spatiotemporal distribution of
aridity indices based on temperature and precipitation in the extra-Carpathian regions of
Romania. Theoret. Appl. Climatol. 112(3–4):597–607
19. Gheorghe IF (2014) Comparative efficiency in nutrient retention between the vegetation of the
natural ecosystems and of the agro-systems. Int J Innov Technol Exploring Eng (IJITEE) 3
(11):98–103. ISSN 2278–3075. https://www.ijitee.org/download/volume-3-issue-11/
20. Gheorghe IF, Barbu I (2014) Green infrastructure - an important factor in the preservation and
use of biodiversity to reduce diffuse pollution and production of biomass. Int J Emerg Sci Eng
(IJESE) 2(5):88–94. ISSN 2319–6378. https://www.ijese.org/download/volume-2-issue-5/
21. Eissenstat DM, Volder A (2005) The Efficiency of Nutrient Acquisition over the Life of a
Root, Chapter from book Nutrient Acquisition by Plants: An Ecological Perspective
Ecological Studies. H. Bassiri Rad (Ed.) Nutrient Acquisition by Plants an Ecological
Perspective. Springer, Heidelberg, vol 181 pp 185–220
22. Thornley JHM (1972) A balanced quantitative model for root: shoot ratios in vegetative
plants. Ann Botany 36(2): 431–441
23. Robinson SL, Rousseau DM (1994) Violating the psychological contract: not the exception
but the norm. J Organ Behav 15(3):245–259
24. Jackson RB, Caldwell MM (1989) The timing and degree of root proliferation in fertile-soil
microsites for three cold-desert perennials. Oecologia 81(2):149–153
25. Thomas H, Sadras VO (2001) The capture and gratuitous disposal of resources by plants.
Funct Ecol 15:3–12
26. Berendse F, Aerts R (1987) Nitrogen-use-efficiency: a biologically meaningful definition?
Funct Ecol 1(3):293–296
27. Aerts R (1997) Climate, leaf litter chemistry and leaf litter decomposition in terrestrial
ecosystem: a triangular relationship. Oikos 79:439–449
28. Lambers H (1987) Growth, respiration, exudation, and symbiotic associations: the fate of
carbon trans-located to roots. In: Gregory PJ, Lake JV, Rose DA (eds) Root Development and
Function. Cambridge Univ. Press, Cambridge
29. Zak DR, Pregitzer KS, Curtis PS, Vogel CS, Holmes WE, Lussenhop J (2000) Atmospheric
CO2, soil-N availability, and allocation of biomass and nitrogen by Populus tremuloides. Ecol
Appl 10:34–46
30. Bloom A, Chapin III SF, Mooney HA (1985) Resource limitation in plants – an economic
analogy. Ann Rev Ecol Syst 16:363–392
31. Gastal F, Lemaire G (2002) N uptake and distribution in crops: an agronomical and
ecophysiological perspective. J Exp Bot 53(370):789–799. https://doi.org/10.1093/jexbot/53.
370.789
32. Melillo JM, Aber JD, Muratore JF (1982) Nitrogen and lignin control of hardwood leaf litter
decomposition dynamics. Ecology 63:621–626
388 I. F. Gheorghe et al.

33. Hättenschwiler S, Vitousek PM (2000) The role of polyphenols in terrestrial ecosystem

nutrient cycling. Trends Ecol Evol 15:238–243
34. Gessner MO, Meyer E, Schwoerbel J (1991) Rapid processing of fresh leaf litter in an upland
stream. Verh. Int. Verein. Limnol. 24:1846–1850
35. Nelson WJ, Kadlec JA, Murkin HR (1990) Seasonal comparison of weight for two types of
Typha glauca (Godr.) leaf litter. Aquat. Bot. 37:299–314
36. Salamanca FE, Nobuhiro K, Shigeo K (1998) Effect of leaf litter mixture on the
decomposition of Quercus serrata and Pinus densiflora using field and laboratory microcosm
methods. Ecol Eng 10:53–73
37. Meentemeyer V (1978) Macroclimate and lignin control of litter decomposition rates.
Ecology 59:465–472
38. Chapin FS, Stuart III (1980) The mineral nutrition of wild plants. Ann Rev Ecol Syst 11:233–
260. https://doi.org/10.1146/annurev.es.11.110180.001313.JSTOR2096908
39. Gheorghe IF (2016) Investigation of nitrogen pathways in typical entropic-modified riparian
zones of the Arges river catchment’s area, Romania, IAEA TECDOC – 1784:135–149.
https://www-pub.iaea.org/MTCD/Publications/PDF/TE_1784_web.pdf
40. Harpet Conservancy (2018) Riparian Zones: What are they and why protect them? https://
www.harpethconservancy.org/programs/restoration/riparian
Conservation
Vegetation and Flora of Sacred
Natural Sites in North-Western
Morocco—Landscape Context
and Conservation Value

Ulrich Deil, Birgit Frosch, Holger Jäckle, and Allal Mhamdi

Abstract Sacred sites came into the focus of UNESCO, IUCN and conserva-
tionists recently. In many societies the surroundings of such sites shelter remnants
of vegetation. They are called sacred natural sites (= SNS) and are often considered
as primeval forests. While numerous geobotanical studies are available from
Southeast Asia and sub-Saharan Africa, such biocultural elements were underex-
plored until recent years in Muslim societies, especially in the Maghreb countries
(NW Africa).
In the last decade (2007 to 2017) we studied and analysed the vegetation around
the tombs of local saints (“Marabout”) and on graveyards on the Tangier Peninsula
(NW Morocco). Total number and size of the SNS in the study area were analysed
from topographic maps and verified by field studies in 8 landscape Sects. (36 km2
plots), representing the mayor ecoregions on a bioclimatic gradient from thermo-
mediterranean subhumid coastal sandy terraces to humid mountain ridges of mid
altitude with sandstone and schist as bedrock material. On 86 SNS, the vegetation
mosaic was recorded: Frequency, spatial extension and naturalness of the habitats
were analysed. The habitats were classified, based on physiognomic characters of
the plant cover and dominating woody species, and grouped according to natu-
ralness. 203 phytosociological relevés have been sampled in the forested parts of
the SNS and were clustered according to floristic similarity. The resulting forest
associations are discussed in a broader geographical context (SW Mediterranean)
and compared with data from non-sacred forests. Vascular plants were classified
according to pre-defined risk factors like stenochory or grazing sensitivity.
Naturalness index and refugial value were calculated for each site.
SNS are common phenomenon of the Moroccan cultural landscape. On the
Tangier Peninsula, 100 km2 plots hold on average 29 sacred sites. All over

U. Deil (&)
B. Frosch
H. Jäckle

Faculty of Biology, Department of Geobotany, University of Freiburg, Schänzlestraße 1,
79104 Freiburg, Germany
e-mail: [email protected]
A. Mhamdi
Département d’ Écologie Végétale, Institut Agronomique et Vétérinaire Hassan II, BP 6202,
Rabat, Morocco

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 391
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_19
392 U. Deil et al.

Morocco, more than 10 000 SNS can be estimated. The mean spatial extent is less
than 1 ha. Only 19% of the SNS area was closed forest. These forests are domi-
nated by Quercus coccifera, Q. suber or Olea europaea var. sylvestris, depending
on the prevailing substrate and bioclimatic conditions. Only a fraction of these
forest patches are near-natural and steno-endemics in these forests are rare.
Numerous sacred sites and graveyards are grazed. Under grazing impact, stenoe-
cious perennial forest floor species are replaced by euryoecious annuals. Hence the
processes of therophytization and ruderalisation also occur on such sites. Constant
overgrazing results in an over-aging of the tree layer (“sub-fossilisation”) and
ultimately results—in spite of the usage taboo of the tree layer—in the disap-
pearance of the sacred grove, due to missing regeneration of woody plants.
Nevertheless, very well preserved holy forests such as the Rusco hypophylli-
Quercetum cocciferae and the Teucrio baetici-Quercetum suberis still persist. They
have a high conservation value, because a) they are rare examples of intact
Mediterranean sclerophyllous forests, b) represent close-to-climax vegetation, c)
can serve as stepping stones in a biotope network and d) are aesthetic elements in
otherwise intensively used cultural landscapes.

Keywords Cultural ecology Sacred grove Holy forest Marabout






Phytosociology Quercetea ilicis Landscape approach Conservation biology

Naturalness

1 Introduction

In many countries and societies, sacred places are an integral part of the religious
self-conception. Most of these sites are designated for cultural and religious rea-
sons, providing a specific space for ceremonies like burials, adoration of local
saints, veneration of ancestors, initiation rites etc. [40, 41]. Generally, the idea of
nature conservation did not play a role in their foundation. However, taboos like the
interdiction of hunting, collecting of plants, deforestation, grazing etc. are indirectly
conservation-related and can have positive effects: these taboos can ensure the
persistence of near-natural habitats like forests in landscapes otherwise strongly
modified by man, and can make sacred sites refuges for threatened species (see for
example [8, 22, 62, 70, 72, 118, 132]. They are thus a traditional mode of con-
serving biodiversity [41, 140].
In the last two decades, conservationists all over the world have become more
and more interested in forest patches considered by local societies as “sacred”.
A symposium [113], organized by UNESCO, IUCN and other international
organisations was dedicated to this topic and resulted in the Yamato-declaration,
formulating guidelines for the conservation management of sacred places. Such
places (including their surrounding vegetation) are called sacred natural sites (=
SNS).
Vegetation and Flora of Sacred Natural Sites … 393

A review of the vast body of conservation-related literature about SNS

world-wide reveals that the vegetation of many sites is more or less strongly
influenced by man. Indicators are, e.g., opening of the tree layer by logging, pas-
turing or burning, the existence of a mosaic of different degradation stages of the
climax vegetation and the invasion or planting of non-native tree species (e.g.
[5, 25, 48, 56, 63, 82, 83, 86, 102, 126]). The statements on rare species, habitats or
vegetation types on SNS diverge in the scientific literature: There are studies
concluding that sacred sites hold a high number of rare and endemic plant species
[59, 77, 118]. Other studies report that larger non-sacred forests contain more
endemic species than small sacred sites [56]. It can be concluded that “more
information about the conservation value of sacred lands is needed”, as Dudley
et al. [41: 568] pointed out.
However, profound studies about the plant cover, biodiversity and conservation
status of SNS are unevenly distributed around the world. The synoptic volume of
Ramakrishnan et al. [101] (see first of all the contributions by Hughes & Chandran
and Hay-Edie & Hadley therein) documents that holy forests have been intensively
analysed in India (e.g. [92, 93]), and, to a lesser extent, in China (e.g. [55]), Korea
[28], Indonesia and Japan. A synopsis of SNS studies in Africa and Asia is pre-
sented by Deil et al. [34] and Dudley et al. [42]. In Europe, the role of SNS for the
conservation of organisms or habitats recently gained some interest (e.g. [6, 7, 49,
50, 78, 108]). Concerning SNS studies in different cultures and religions, the
Muslim world has been more or less ignored: In the IUCN-volume, edited by
Schaaf and Lee [113], only one contribution is from an Islamic country. Recently,
the role of Muslim graveyards in Turkey for protecting orchids from tuber harvest
(for salep production) was studied by Molnár et al. [84].
In Africa, sacred site studies concentrate on West and East Africa (see [42, 68,
115, 116]). Inventories of the flora and an analysis of the vegetation and stand
structure are for example presented by Guinko [60] and Neumann and
Müller-Haude [88] from Burkina Faso, by Tchamié [122], Kokou and Kokutse [71]
and Kokou et al. [72] from Togo, by Campbell [25] from Ghana, by Mgumia and
Oba [81] from Tanzania, by Virtanen [134] from Mozambique, and by a number of
studies in the afro-montane zone of Ethiopia [5, 137, 136, 21, 138, 139].
In contrast to sub-Saharan Africa, the African countries north of the Sahara have
been poorly investigated, despite their numerous sacred sites and their biotic rele-
vance. Morocco is part of the Mediterranean Biodiversity Hotspot [87]. About 22%
of the known 4200 Moroccan phanerogams are considered endemic [103, 131].
Morocco is faced with the challenge of taking conservation measures for an esti-
mated total number of 1700 endangered plant species (including mushrooms, algae,
lichens etc.), among them many narrow endemics (SEEE [114]). The main threats to
the survival of endangered species in the southern Mediterranean include over-
grazing, deforestation, development (tourism), agricultural intensification [100],
urbanisation [129], and, although direct effects are difficult to detect, climate change,
above all in mountainous regions [124]. Similar threats affect near-natural habitats
like dense sclerophyllous forests. Overgrazing, timber and firewood exploitation,
replacement by Eucalyptus plantations, wild fires, and intentional burning with
394 U. Deil et al.

subsequent cultivation mainly with Kif (Cannabis sativa) result in an ongoing

degradation and decline of low and mid altitude thermomediterranean dense scle-
rophyllous forests in the country [10, 27, 85, 105].
The contribution of sacred sites in Morocco for the conservation of near-natural
habitats and the protection of rare and endemic plants has not been extensively and
quantitatively studied up to now. Preliminary observations about the flora and the
naturalness of the vegetation of SNS in NW Morocco were presented by Deil et al.
[32–34]; preliminary data about number, size and locations of SNS were given by
Demdan et al. [37] and Taïqui et al. [120].
In Morocco and other Northern African countries, the appreciation of the spir-
itual authority of patron saints is a common and vivid phenomenon [74], and is
expressed by hundreds of collective pilgrimages (“moussem”) to the saints’ tomb
(“koubba” in Arab) every year ([18]; Fig. 1 in [32]). Initially, faithful people
assembled around the living saints, called “Marabout”, forming a religious broth-
erhood. In most cases, SNS in Morocco are burial places of Muslim saints. Their
foundation lies in the religious belief that the surroundings of the burial place are
blessed through the saint’s presence, thus rendered sacred themselves [74]. Public
rural graveyards lacking a saint’s tomb are often equally protected from overex-
ploitation or habitat destruction. Therefore, they have been included into the con-
cept of SNS in this study.
In northern Morocco, some rare thermomediterranean forest types and very
well-preserved stands of formerly wide-spread forest types are conserved almost
exclusively at sacred sites [12, 52, 96, 98, 110]. However, the exemplary study of
two sacred sites in northern Morocco by Deil et al. [33] revealed a mosaic character
of the vegetation due to pronounced human impact. As a result of overgrazing and
nitrification, sacred groves seem to be often degraded, overaged or reduced to tree
groups or a single tree, and processes like the replacement of woody underbrush
and perennial forest herbs by ruderals and annual species (therophytization, rud-
eralisation) take place [51, 52].
Concerning the conservation of rare and endemic plants, Deil et al. [33] spec-
ulate that sacred sites in Morocco may provide protection to only few endemic
(stenochorous) or stenoecious vascular plant species as sacred sites are generally
found where ‘normal’ site conditions prevail (climatic and edaphic climax). The
majority of the known rare and endemic plant species in Morocco however is
associated with extreme soil conditions (rocks, dunes, saline habitats) or occurs in
high mountain areas [47, 80].
In the current study we hypothesise that sacred sites function to some extent as
informally protected ‘nature reserves’ and can contribute to the protection of
endangered plant species and habitats in Morocco. The following main questions
are addressed:
1) How many sacred sites exist? What is their cultural context (saints’ tomb,
cemetery, both functions, destination of pilgrimage)? Where are they located? Is
the number and density of SNS different according to ecoregion or land use
intensities surrounding them? What is the size of the protected area? Are
Vegetation and Flora of Sacred Natural Sites … 395

topographic maps a probable data source for studying abundance and spatial
patterns of SNS? Are SNS restricted to certain relief features or does the variety
of their relief positions allow the existence of a broad spectrum of climax
communities?
2) Do the sacred sites shelter forests? What percentage of the sites does the forested
area have? Are these groves virgin forests or are they subjected to pasturing and
other human impacts (beyond burial activities!)? Are they better preserved than
non-sacred forests? Does the floristic composition of the groves reflect the
abiotic differentiation of the study area and are the well-preserved sacred groves
benchmarks of the potential natural vegetation? Can we arrange the holy forests
of NW Morocco into associations which are known and described from
non-sacred plots?
3) Which other habitats occur on SNS? How natural is the overall vegetation
mosaic of SNS in NW Morocco? ‘Naturalness’ in this context is defined by
species combinations, vegetation structure and life-forms. For the site conditions
of the study area, a closed sclerophyllous forest is presumed as most natural,
while bare soil without plant cover is considered least natural.
4) Is the present land use sustainable and does it guarantee the long-term persis-
tence of the sacred forest?
5) Are the sacred sites refugial areas for rare, vulnerable or threatened plants,
especially stenoecious and stenochorous species? Are ‘naturalness’ of the plant
cover and ‘refugial value’ for plants correlated attributes? Is the extension of the
sacred site correlated with the refugial value?
A first starting hypothesis is that the vegetation of sacred sites is closer to the
climax vegetation than the surrounding landscape and that the floristic composition
of the SNS does mirror the abiotic patterns of the study area. Concerning natu-
ralness and phytodiversity, the hypothesis is that disturbance by men and his
livestock can create a vegetation mosaic on the SNS and will hereby increase the
species and the habitat diversity (medium disturbance hypothesis). In accordance
with the “therophytization” model of the degradation of Mediterranean forests,
proposed by Barbero et al. [10], we expect a shift from perennial herbaceous
species to annuals in the forest floor vegetation and an increase of ruderals,
over-compensating the reduction of stenoecious forest taxa. The opening of the
forest canopy and the nitrification of the forest soil should result in a floristic
trivialization of the forest floor vegetation.
The study was carried out on the Tangier Peninsula in NW Morocco, an area
known for its Marabutism [16]. Finally, the situation on SNS in north-western
Morocco will be compared with other parts of the world.
396 U. Deil et al.

2 Study Area

The study area is located on the Tangier Peninsula in north-western Morocco

(Fig. 1). It is delimitated to the west by the Atlantic coast and to the east by the
Outer Rif Ranges or Pre-Rif Mountains. The northern limit is 35º 30′ N, the
southern limit 34º 45′ N. It features a surface of ca. 7300 km2 and an altitudinal
range from sea level to about 1700 m a.s.l. The climate is of Mediterranean type
with winter rains and summer drought. The location close to the Atlantic ocean
results in an Atlantic variant of the Mediterranean climate with a relatively high
precipitation (mean annual precipitation 600 mm near the coast and more than
1000 mm on the highest peaks) and mild winters (extension of the thermo-
mediterranean belt up to 900 m a.s.l.; thermomediterranean sensu [94], mean of the
minimum temperature of the coldest month [m] > 3 °C). Figure 1 documents the
differentiation of the hygro-climate from sub-humid in the lower parts to per-humid
conditions on the Mountain ridges [12].
Geologically the western part of the study area near to the coast is dominated by
Pliocene and Quaternary sands, the hilly area in the centre by marls. The Outer Rif
Ranges mainly consist of a series of parallel Mountain ridges of Oligocene

Fig. 1 Study area: delimitation, bedrock strata, climatic differentiation and location of the selected
landscape sections (from [66])
Vegetation and Flora of Sacred Natural Sites … 397

sandstone. In between these ranges and further to the east, Cretaceous flysch or
schist prevail [58, 119].
The study area falls into the phytochorological units “Tangier Area” and
“Loukkos” [131]. The potential natural vegetation throughout the thermomediter-
ranean belt in the study area is Mediterranean evergreen forest with dominating
Cork Oak (Quercus suber), Holm Oak (Q. rotundifolia), Kermes Oak (Q. coccifera)
or Wild Olive (Olea europaea var. sylvestris) [12, 44, 131]. Details of the biocli-
matic niches of the dominating tree species are presented by Frosch and Deil ([52],
Fig. 3). The actual vegetation in the lower parts is mainly composed of cropland
and pastures. Remaining forests and maquis are intensely used as wood pasture, as
well as for firewood collection, timber exploitation and cork extraction. On the
sandstone ridges in mid altitude, Cork Oak forests and their first degradation stages
(maquis and heathland) are still widespread. Clearing for Kif (Cannabis sativa)
cultivation, intentional burning and overgrazing are ongoing degradation processes
[27, 85, 105].

3 Methods

Stratification and Random Selection of Subplots: The study was performed at 4

spatial scales and with different intensities of field data sampling (Fig. 2). In order
to obtain representative data for the Tangier Peninsula, a pre-stratified random
sampling procedure was applied. Urban areas were excluded. The study area was
divided into 4 eco-regions (= strata), which are homogenous in relief type and
dominating bedrock material: A) Littoral sandy plains and hills, B) Prerifean marly
hills, C) sandstone ridges, D) flysch and schist mountains (Fig. 2).

Fig. 2 Schematic topographical and geological profile across the study area, illustrating the 4
spatial scales of data sampling (from [34])
398 U. Deil et al.

Within each ecoregion, two corresponding landscape sections of 36 km2 size were
selected randomly. They had to fulfil the following additional criteria: 1) a mini-
mum of five SNS according to records on topographic maps, 2) a minimal distance
of 20 km to avoid spatial auto-correlation, 3) the two corresponding sections must
be situated in the same bioclimate according to Benabid [12].
The sections span an altitudinal range from 40 m a.s.l. near the coast to about
900 m a.s.l. in the Pre-Rif Mountains. Each section has homogeneous geological
and climatic conditions. The sampling locations are exclusively situated within the
thermomediterranean belt. The hygroclimatic conditions of the studied sacred sites
range from subhumid to humid according to the classification in Benabid [11],
modified from [111]. Spatial geological information was taken from Suter [119] and
generalised to four main strata: littoral sands (sections A), marls (B), sandstone (C),
and schist/flysch (D + E). In these eight landscape sections all sacred sites were
mapped, apart from landscape section D, where a random sub-sample was chosen
due to the unusual high number of sacred sites. All in all, 86 sacred sites with a
minimal spatial extension 100 m2 were investigated. Localities and size (bound-
aries) of the sacred sites were determined after consultation with local authorities.
Figures 3 and 4 give an impression of two landscape sections and they illustrate
the spatial relations of sacred forest patches to the settlements and the surrounding
land use pattern.

Fig. 3 Sacred grove Sidi el Mokhfi near the village Er Rhorba. The landscape structure is typical
for landscape section D, situated in the tribal area of Beni-Ider in the north-western Rif Mountains.
The sacred grove is a mixed stand with Quercus coccifera and Olea europaea var. sylvestris
Vegetation and Flora of Sacred Natural Sites … 399

Fig. 4 Two sacred groves near Mokrisset (Province of Chaouen) in the western Rif Mountains.
The landscape structure is typical for landscape section E, situated in the tribal area of Beni-Chaïb
with valleys deeply incised into schistose mountain ridges. The holy forests are dominated by
Quercus coccifera, Q. suber and Olea europaea var. sylvestris

Density and Spatial Pattern of SNS: In a first step, the number and topographic
position of SNS (= all the notations “saints tomb” and “cemetery”) were extracted
from topographic maps of different scale (see [65] for exploited sheets and scales).
To verify the information on the maps, a ground check was carried out in all the
eight landscape sections. The total number of all existing SNS was identified
through field surveys. The area covered by the ground check corresponds to around
4.5% of the surface of the study area. The total number of SNS in the study area
was estimated by a linear regression, regressing the number of SNS proved by
ground check against the number of sites on the topographic maps 1:50 000. Spatial
pattern, relief position and possible explanatory variables (e.g. slope, elevation or
dominating land cover in the sections) were analysed by statistical tests described
by Jäckle et al. [65] in detail.
Vegetation Analysis Using the Phytosociological Approach: On each sacred
site, one to several vegetation relevés were made. A partial species list for each
sacred site and a total species list for all investigated sacred sites were derived from
203 relevés and additionally recorded notable plant species. Sampling took place in
spring and early summer 2008 and 2009. Nomenclature and taxonomic concept of
400 U. Deil et al.

plant species follow Valdés et al. [131]. Collected specimens are stored in
Herbarium Freiburg (FB). Plant communities are named according to Fennane [45]
and Rivas-Martínez et al. [106]. Accompanying site parameters like bedrock, pH,
and content of CaCO3 were noted or measured. Taking soil samples is a delicate
task on sacred sites. We therefore sampled only one per cemetery and only from the
upper soil layer (1–10 cm).
On 68 SNS, phytosociological relevés in plots with a tree layer could be
sampled, in total 140 relevés. Site names, spatial extension, geographical coordi-
nates of the sites and the complete phytosociological table are given in Frosch and
Deil [52]. Data sampling (cover-abundance scale) and data analysis (classification
according to floristic similarity) follow the method of Braun-Blanquet [23]. The
relevés were stored in a TURBOVEG database and exported into the JUICE pro-
gram for sorting. Floristic composition as well as vertical structure of the stands was
taken into account for classification. A flow chart in Frosch and Deil ([52], Fig. 5)
gives an overview of the methodological steps of sampling and analysis of the
phytosociological data.
In order to discuss the forest communities on sacred sites in NW Morocco in a
broader geographical context and to evaluate their naturalness and their role as
documents of potential natural vegetation (PNV), the best preserved holy forests
were compared with reference data of similar forest communities, recorded mainly
on non-sacred sites in southern Spain and Portugal, and northern Morocco and
Algeria. Multivariate analysis was applied to determine the floristic affinity within
the data set [123]. The synoptic tables and the ordination diagrams are documented
in Frosch and Deil [52].
Physiognomic Approach and Classification of Naturalness: The vegetation of
the 86 SNS was mapped based on vertical structure of the vegetation, the combi-
nation of growth forms and the dominant species in each stratum. The application
of physiognomic and structural parameters follows Küchler [73]. Cover of growth
forms was estimated in seven categories and seven height classes were distin-
guished (Tab. 2 and 3 in [66]). The minimum size for recording of a habitat was
100 m2.
In a first approach, focussing on the highest vegetation layer, the mapped veg-
etation patches were grouped to five major habitat types (forest, shrubland,
herbaceous formations, open and sparse formations, formations dominated by
planted non-native tree or shrub species) (for details see Tab. 1 in [53]). To separate
forest from shrubland, a 5 m height threshold (often used for the classification of
Mediterranean maquis) was applied (e.g. [46]). To each habitat type, a level of
naturalness was assigned on an ordinal scale of 1–8 (Tab. 1 in [53]). Closed forest
was presumed to represent the most natural habitat type (highest level 8). Bare soil
and formations with more than 50% cover of introduced species represent the least
natural habitat types (lowest level 1). Naturalness combined with the respective
surface of the habitats on a given locality was used to calculate an average level of
Vegetation and Flora of Sacred Natural Sites … 401

naturalness (see formula in [53]) and this value was used to analyse the influence of
the naturalness on the refugial value of SNS (see below).
In a more detailed second approach a system of 20 habitat types was developed
using the vertical structure of the vegetation, the cover value of the strata, the
compositions of growth forms and the dominant species as parameters for the
classification (for details see Fig. 2 in [66]). The naturalness values attributed to the
habitat types follow the naturalness classes proposed by Ellenberg [43].
Conservation Value of the SNS and Vulnerable Plant Species: 460 auto-
chthonous vascular plant species were recorded on the 86 investigated sites. They
were ranked concerning their vulnerability. Ranking was based on five pre-defined
risk factors concerning limited geographic range, habitat restriction and sensitivity to
grazing (A-E, for definitions see Tab. 2 in [53]). Information on distribution, rarity,
and habitat restriction was mainly derived from Fennane and Ibn Tattou [47],
complemented by information from Quézel and Santa [97], Castroviejo [26], Valdés
et al. [130] and African Plants Database [1]. Risk factors A and B are limited
geographic range (narrow endemics respectively distribution restricted to the SW
Mediterranean). Risk factor C encompasses rare Mediterranean species which are
stenoecious throughout their distribution area like Davallia canariensis, as well as
species at the margins of their geographic range, and thus stenoecious, like
Cephalanthera longifolia. Risk factor D (habitat restriction: mesophilous) comprises
species more or less restricted to thermomediterranean forests (e.g. Ruscus
hypophyllum) and shade-tolerant mesophilous herbaceous taxa not restricted to
forests (e.g. Oenanthe pimpinelloides). Namely diagnostic species of Mediterranean
sclerophyllous forests and western Mediterranean thermomediterranean evergreen
forests (Querco rotundifoliae-Oleion sylvestris Barbero et al.), according to
Rivas-Martínez et al. [107], Barbero et al. [9] and Benabid and Fennane [14] were
included in this category. Risk factor E comprises potential grazing sensitive plants.
As more than one of the five risk factors could be attributed to one species, 10
exclusive vulnerability categories were defined through the combination of risk
factors. Vulnerability categories were ranked as to their assumed degree of vul-
nerability from 5 = highest vulnerability to 1 = lowest vulnerability (for definitions
and ranking see Tab. 2 in [53]). For each sacred site, vulnerability ranks of the
vulnerable species present were added up to obtain a sacred site-based value
concerning the vulnerable species representation. We call this value ‘refugial
value’.
To evaluate the possible influence of different explanatory variables on the
refugial value, a multiple linear regression model was calculated with the size of the
sacred sites (log-transformed), their calculated naturalness value, and the number of
habitat subtypes of sacred sites as independent variables. Since the refugial value
showed a non-normal distribution, square-root-transformation was applied to
reduce the influence of the few sites with a high refugial value. Statistical analyses
were carried out using the statistical environment R (version 2.15.3, [99]).
Quantitative analyses were performed using a vector GIS (ArcGIS 9.3, ESRI).
402 U. Deil et al.

4 Results

4.1 Number, Density and Size of SNS

1450 SNS were documented on the topographic maps 1:50 000 (= TM50) within
the study area (Fig. 5). The ground check within the eight landscape sections
showed that out of 120 existing SNS only 90 were mapped on TM50. The number
of existing SNS varied from 4 to 25 between the landscape sections. The estimation
of the total number of SNS using the calculated linear regression model [65]
resulted in a total number of 1548 existing SNS and a density of 29.4 ± 6.9 SNS/
100km2. Similar densities were recorded for other regions of the Tangier Peninsula
by Demdan et al. [37] and Taïqui et al. [121]. Projection of the average density of
29.4 SNS/100 km2 on the whole study area resulted in a total number of 2127 SNS.
Thus, the 1.450 SNS mapped on the TM50 represent around 67% of the real

Fig. 5 Number and locations of SNS in the study area, analysed from topographic maps (from
[65])
Vegetation and Flora of Sacred Natural Sites … 403

existing SNS. There is a generalization effect, depending from the map scale (67%
shown on TP50, 20% on TM100; 2% on TM250; 0% on TM500). For the whole
country we can estimate more than 10 000 SNS, distributed over all ecoregions
[36]. Because of their abundance and distribution pattern they could serve as ele-
ments of a biotope network in degraded landscapes of Morocco.
The 86 investigated sites ranged in size from 85 to 48,950 m2. The average size
was 0.59 ha (median). Only 5 sacred sites were larger than 2 ha (see Fig. 2 in [53]
for the frequency of size classes). The total area of the 86 sacred sites amounted to
0.68 km2. Within the seven completely inventoried landscape sections of 36 km2
each, the sacred site area varied between 0.09% and 0.36% of the total area. The
point pattern analysis shows a non-random distribution (Fig. 5). Sacred sites occurr
mainly at elevated topographic positions (hilltops, mountain ridges), but can be
found in almost all relief positions, thus offering the potential of supporting dif-
ferent types of climatic climax vegetation. However, depressions and river channels
are unlikely positions as well as very steep slopes (<15°) with shallow soils. Such
relief positions—representing a special pedoclimax (narrow groundwater level
respectively very shallow soils)—are not suitable for graveyards.
By their function as cemeteries, the SNS occur with preference close to the
villages [65]. The spatial pattern of settlements in northern Morocco strongly
depends on the occurrence of springs and surface water. Especially the borderline
between sandstone and underlying marls and flysch is a spring horizon and thus a
preferred location for settlements—and in consequence—of cemeteries. Most of the
sacred sites consist of a saint’s tomb, surrounded by a cemetery. People want to be
buried near a sacred tomb to profit from the “Baraka”, the spiritual blessing of all
objects in spatial context to the Marabout [74].

4.2 Vegetation Cover and Degree of Naturalness of the SNS

Almost half of the investigated sacred site area was covered by forests (42%). The
subtype closed forest, attributed with the highest level of naturalness, occurred on
31 sacred sites, covering 19% of the total area. Another 40% of the sacred site
surface was shrubland, 30% with and 10% without trees. Herbaceous vegetation
with (9%) and without (7%) trees or shrubs occupied less space. Open herbaceous
formations and sparse vegetation made up only 2% of the total sacred site area.
Habitats dominated by non-native woody plants were found on 9 of the 86 sacred
sites, covering less than 0.5% of the total sacred site surface. Per sacred site, one to
five habitat subtypes were distinguishable, on average 2.6. At most of the localities
(60%), a vegetation mosaic occurred with two or three habitat subtypes. 20% of the
sites had a rather uniform vegetation cover with only one habitat subtype. Another
20% had a fine-grained vegetation mosaic with four or five different habitat sub-
types. The average degree of naturalness per sacred site was 5.9 (range 2.0–8.0)
(see naturalness classes in [53], Tab. 1). Only two sites were completely covered
404 U. Deil et al.

with closed forest. Most of the SNS in the study area underlie a moderate to strong
anthropogenic influence.
The classification process applied by Jäckle [66], including all vegetation strata,
the combination of growth forms and the dominant species as criteria resulted in 20
habitat types (for details see [66], Tab. 5 and Appendices 2 and 3). These 20 habitat
types represent a degradation series from near-natural forest with a more or less
closed canopy and typical forest species in the shrub- and herb-layer to sparse
vegetation with less than 30% total vegetation cover or patches dominated by
planted or sub-spontaneous non-native tree or shrub species. The most natural
categories “primeval” and “natural” were not found on the investigated SNS as well
as the least natural category “artificial”.
The cover value (in ha) of the 20 habitat types on the 86 mapped SNS is
illustrated in Jäckle ([66], Fig. 3), the habitat types, their naturalness value and their
composition of layer-types is documented in Jäckle ([66], Tab. 5), the vegetation
mosaic present on each SNS is shown in Jäckle ([66], Tab 7).
The overall naturalness of the vegetation on the SNS in the different landscape
sections is shown in Fig. 6 here, differences in the composition of habitat types
among the landscape sections are given in Jäckle ([66], Tab 6). Forest patches occur
mainly in the eastern part of the study area in the landscape sections C1, C2, D and
E. Very well conserved forest patches were found exclusively in the landscape
sections D and E. In sections B1 and B2 forests are rare and, if present, they are
overaged and have a nitrified herb layer. In sections A1 and A2 the few forests are
also overaged and without tree regeneration in most cases. Maquis with
Pistacio-Rhamnetalia shrubs occur in all landscape sections. Vegetation dominated
by Asphodelus ramosus is restricted to SNS in the marly lowland.

Fig. 6 Pattern of naturalness of the SNS vegetation in the landscape sections A to E (from [66])
Vegetation and Flora of Sacred Natural Sites … 405

4.3 Forest Plant Communities on SNS in NW Morocco

The preliminary typology of sacred forests on the Tangier Peninsula presented by

Deil et al. [34] was based on 66 relevés, sampled in holy forests. A bigger data set
(140 relevés from 68 SNS) was used by Frosch [51] to compare the vegetation of
trees stands on SNS with well-preserved sacred and non-sacred forests in
north-western Morocco. The same data set was used to determine the forest com-
munities on the forested parts of SNS and to compare them with well-preserved
non-sacred forests in the West-Mediterranean countries. Details and a complete
documentation with constancy tables and with tables of the single relevés are given
in Frosch and Deil ([52], Tab. 1 and 2, Appendix 2).
The forested vegetation of sacred sites shows a great variety. Abiotic factors,
such as substrate and bioclimate, and human interventions (grazing, fire, etc.) were
found to be important differentiating factors. Besides well-conserved forests, a
broad spectrum of degradation stages was documented. Most stands still shelter the
original tree species combination, but vertical structure and floristic composition of
the herb layer are strongly modified. Processes related to anthropo-zoogenic pres-
sure, like therophytization, ruderalisation and over-aging of the tree layer occur
despite religious taboos.
The sacred sites’ forest stands were classified into three clusters, the dominant
tree species being either Quercus suber, Q. coccifera or Olea europaea. Overall, 17
forest types could be distinguished. A number of well-preserved stands belong to
the associations Teucrio baetici-Quercetum suberis Rivas-Martínez and Rusco
hypophylli-Quercetum cocciferae [12] and their subassociations and variants. In
comparison to forests on non-sacred sites, these holy forests are very well pre-
served, as is indicated by the presence of strict forest species of the Quercetalia ilicis
and the Querco-Oleion sylvestris, by transgressives of the Quercetalia pubescentis,
and by a high constancy and abundance of ombro- and mesophilous taxa.
Cork Oak Forests: Well preserved stands of the Teucrio-Quercetum suberis
occur in the thermomediterranean humid bioclimate, on sandstone and sandstone
screes, with a subvariant of Davallia canariensis on north facing slopes in section
C1 and a subvariant with Quercus coccifera on shallow sandstone screes in sections
D and E. Open and stronger grazed Cork Oak forests are marked by a monospecific
tree layer of Quercus suber and heliophilous species in the shrub and herb layer.
Figure 7 illustrates a variant with Cistus crispus on the forest floor, a result of
repetitive ground fires.
In the broader SW-Mediterranean context, the well preserved sacred Cork Oak
groves on the Tangier Peninsula take an intermediate position between subhumid
thermomediterranean and humid thermo- to mesomediterranean Cork Oak forests
([52], Tab. 2, Fig. 11). In general, Teucrio-Quercetum suberis stands from SNS
give the impression of very well preserved forests in comparison to non-sacred
Quercus suber forests. Heliophilous species of the classes Cisto-Lavanduletea and
406 U. Deil et al.

Fig. 7 Sacred grove with overaged Cork Oak (Quercus suber) on the cemetery ‘Rija Ibled’ in
landscape section C2. Cistus crispus dwarf shrub vegetation dominates the left foreground,
ruderals and annuals to the right, inconspicuous tombs (scattered stones) in the background

Calluno-Ulicetea, indicating fire and opening of the tree canopy by overexploita-

tion, are less frequent and less abundant in their shrub layer. On the other hand, the
forests on SNS shelter more ombrophilous and mesophyllous species such as
Davallia canariensis, Polypodium cambricum, Ruscus hypophyllum, Asplenium
onopteris etc., indicating a more humid micro-climate inside the well protected
forests.
Kermes Oak Forests: The Rusco-Quercetum cocciferae also occurs in the ther-
momediterranean humid bioclimatic belt, but grows on schists, marls and shallow
sandstone screes. The forests are marked by a species-rich stand structure of several
tree- and shrub-layers and a distinct liana layer (Fig. 8). Quercus coccifera and
Phillyrea latifolia are co-dominant in the middle tree-layer (5–15 m); Pistacia
lentiscus forms a third tree-layer of up to five meters height. Cork Oak contributes
to the upper tree stratum (subassociation brachypodietosum sylvatici) on sandstone
screes in section D. On calcareous soils, Holm Oak (Quercus rotundifolia) and
Wild Olive (Olea europaea var. sylvestris) occur in the middle tree layer (subas-
sociation quercetosum rotundifoliae and variant with Olea europaea).
Vegetation and Flora of Sacred Natural Sites … 407

Fig. 8 Near natural Rusco hypophylli-Quercetum cocciferae around Marabout ‘Sidi Arabi’ in the
Western Rif Mountains (landscape section D). The site is dominated by arborescent individuals of
Quercus coccifera and Phillyrea latifolia. Due to opening of the canopy around the qubba, grasses
dominate the foreground

The ground floor of the sacred Quercus coccifera forests of the Tangier
Peninsula is composed of a number of ombrophilous species, e.g. Ruscus
hypophyllum, Galium scabrum, Brachypodium sylvaticum and Luzula forsteri.
These species are restricted to non-grazed sites or to very humid sites with moderate
grazing. Remarkable is the high constancy of Cephalanthera longifolia in sacred
groves of the Rusco-Quercetum brachypodietosum. This orchid is at its southern-
most limit of distribution in NW Morocco and is linked to intact meridional brown
forest soil with a thick humus layer.
Wild Olive Forests: The Wild Olive groves occur in the thermomediterranean
sub-humid bioclimate, on Quaternary and Pleistocene coastal sandy terraces (sec-
tions A1, A2 and B1) as well as on sandstone and schists in lower altitudes (up to
490 m a.s.l.). None of the Olea europaea groves could be assigned to a syntaxon at
association level. The stands must be treated as rankless communities. These units
obviously comprise sacred groves which are submitted to strong anthropo-zoogenic
impact (grazing, fire, burial activities, and exploitation of wood). Facies with
Hordeum murinum ssp. leporinum or Brachypodium distachyon indicate the strong
therophytization and nitrification of the ground layer. Figure 9 illustrates such a
stand: Under a still quite closed tree canopy with over-aged Olea europaea var.
408 U. Deil et al.

Fig. 9 Sacred site ‘Sidi Amour Al Hadi’ in Douar Swahal with Olea europaea var. sylvestris. The
forest floor under the overaged Wild Olive trees is ruderalised by heavy grazing, by defaecation of
resting animals and by seasonal pilgrimage activities of humans (grove used as campsite)

sylvestris, the forest floor vegetation is strongly ruderalised by heavy grazing,

defaecation of resting animals and seasonal pilgrimage activities of humans. In a
long run, these sacred groves will disappear despite people respecting the taboo of
logging and clearing, because regeneration of the tree layer is missing since dec-
ades. We must consider these holy groves as “sub-fossile”.
In conclusion, sacred groves often seem to represent the potential natural veg-
etation. Exceptions are sacred groves with Wild Olive, which occur in an ecoregion
(sections A1 and A2 on coastal sandy terraces) where a thermomediterranean
Quercus suber forest is considered to be the climax (see discussion in [52]).

4.4 Vulnerable Plant Species and Refugial Value of SNS

Of the 460 autochthonous species growing on the investigated sites, 103 (22%)
featured one or several of the predefined risk factors indicating vulnerability. Only
four species (Davallia canariensis, Echinops spinosus ssp. fontqueri, Scilla
monophyllos, Silene lagrangei) were narrow endemics with at least one additional
risk factor and thus ranked in the highest vulnerability category. The most common
Vegetation and Flora of Sacred Natural Sites … 409

risk factors were “Grazing sensitivity” and “Habitat restriction to forest”. 10.9%
and 7.8% respectively of all plants on the sacred sites were categorised to these risk
factors. Less species (5.2%) were endemic with a southwest Mediterranean distri-
bution. 2% of the 460 species were stenoecious only and 2.6% were more narrow
endemics. A complete list of the 103 vulnerable vascular plant species, their risk
factors and their frequency on the studied 86 sites is documented in Frosch et al.
([53], Tab. 3, Appendix A).
The number of vulnerable plant species per sacred site varied between 0 and 25.
The refugial value, calculated by summarising the vulnerability ranks of the
co-occurring species, ranged from 0 to 51. The value shows a non-normal distri-
bution with many sacred sites with a low and only few sacred sites with a high
refugial value. A multiple linear regression revealed that the spatial extent of the
sacred sites was not significantly associated with their refugial value. However, a
higher degree of naturalness and a higher number of habitat subtypes resulted in a
higher refugial value (see Fig. 4 and Tab. 4 in [53]). The total number of vulnerable
species and the number of species with risk factor “Habitat restriction to forests”
was highest on sacred sites with patches of forest and shrubland.

5 Discussion

5.1 Number, Density and Size of SNS in Morocco

in Comparison to Other Countries

SNS in the study area in NW-Morocco are very numerous (about 29.4 sites per 100
km2) and in reality more frequent than the locations mapped as “saint tombs” or
“cemetery” on TM50. We can expect more than 10 000 SNS all over the country. In
most cases they consist of a venerated tomb, surrounded by a cemetery, and they
are small in size (on average ca. 0.7 ha). Despite the high number, they cover less
than 1% of the landscape. High density and small size of SNS in Morocco is also
recorded by other authors [37, 74, 121]. High numbers and small size of SNS are
also mentioned from other parts of the world. The number of holy forests all over
the Indian subcontinent is estimated to be several hundred thousand. Campbell [25]
calculated at least 1900 for Ghana, Mwihomeke et al. [86] counted several hundred
in a small district of Tanzania. The groves there are mostly only a few hectares in
size.
Often the human impact results in a fine-grained mosaic of different habitats,
where forests constitute only one element among others. The mean size of the forest
patches in NW Morocco is therefore still smaller (0.5 ha on average per site) than
the total mean size of the SNS. We can state a tendency to less conserved sacred
sites in lowland areas, where they are embedded into an intensely used agricultural
landscape.
410 U. Deil et al.

5.2 Naturalness of the Vegetation on Sacred Sites

in Northern Morocco

As climax and close-to-climax forest communities have vanished in vast areas of

thermomediterranean Morocco, and the remaining forests on non-sacred sites are
more or less degraded [14], sacred groves can be of high conservation value. Our
detailed studies in NW Morocco confirm the presumptions of Emberger [44]
Sauvage [110] and Benabid [12] that certain sclerophyllous lowland forests types
such as Rusco hypophylli-Quercetum cocciferae and Teucrio baetici-Quercetum
suberis nowadays exist almost exclusively around Marabout tombs and on ceme-
teries. On the other hand, Frosch and Deil [52] described for the first time several
stages of degradation of these forests. Until recently the percentage of little dis-
turbed forests on sacred sites in Morocco was unknown.
In the studies realised by us on a randomly chosen sample of 86 representative
sacred sites on the Tangier Peninsula, a pronounced anthropo-zoogenic impact on
many sacred sites is documented: Only about one fifth of the total investigated
sacred site area was actually closed forest with a tree canopy cover of  75%. On
most of the sites, a vegetation mosaic consisting of several habitat types including
degradation stages of forests and maquis was found. This specific vegetation pattern
results from different human activities, especially burning, pasturing, and burial
activities [33], but also clearing, cork harvesting or logging. This is congruent with
the findings of Taïqui et al. [121] who stated that less than 50% of the sites in
Larache Province are forests and with the observations of Frosch [51], focussing on
the forested part of SNS: Even in some holy forests with a more or less closed tree
canopy (especially in groves with Wild Olive and with Cork Oak), a reduced cover
of the litter layer and a dominance of annual plants in the ground floor indicate
intense grazing and browsing.
The habitats “shrubland” and “herbaceous formations” cover a considerable area
on the sacred sites. They represent main stages of a degradation series model for
Mediterranean forests, with matorralisation and therophytization as the dominating
processes [10, 13, 95, 127]. The term matorralisation refers to the increase in cover
and diversity of shrub species in the undergrowth of Mediterranean forests.
Therophytization describes the replacement of woody underbrush and perennial
herbs through annual species. The conversion of forests to shrublands or herbaceous
formations is a gradual process, caused by lack of tree regeneration, often due to
intense browsing, and the reduction of the tree cover through logging or natural
mortality.
In a global context, the focus of many studies has been on at least partly forested
sacred sites, the so-called ‘sacred groves’. This was also the case in most research in
Morocco in the period 1960 to 2005. Accordingly, results in these earlier studies
may be biased in favour of the best preserved sacred sites, and these selective
studies might have favoured the un-reflected impression of sacred groves as “virgin
forests”. Nevertheless, in many non-exhaustive studies from other parts of the
world, a human influence on vegetation is at least mentioned (e.g. [25, 67, 76, 82,
Vegetation and Flora of Sacred Natural Sites … 411

83, 102]), and sometimes considered to be the most important factor shaping
vegetation [70]. Furthermore, quantitative vegetation studies suggest that human
influence/degradation may be an important factor on sacred sites and sacred groves
worldwide [5, 54, 63, 86, 126]. In a few studies, less and more disturbed patches,
classified according to canopy cover level or dominant life form were described
from the same sacred site [25, 83, 126], indicating vegetation mosaics similar to
those often found in northern Morocco.
Reasons for the degradation of SNS can be manifold. Weakening of the religious
system (secularisation processes) and growing resource pressure by increasing
population have been observed by many authors from sub-Saharan Africa (see [25]
and contributions in [116] and are here confirmed by us for Morocco. In a few
cases, the tree layer of a sacred grove was cut down to finance the renovation of the
Mosque or the Qubba (the tomb of the Saint) through the purchase of the timber.
On the other hand, the high veneration a certain Saint is held in can also increase the
anthropogenic pressure on the vegetation via frequent visits and/or regularly held
pilgrimages which may temporarily transform a SNS into a campsite. From an
ecocentric viewpoint, the usage of the term “degradation” for this process may be
justified. From an ethnocentric viewpoint, taking into account the spiritual value as
an ecosystem service of the sacred sites, this process cannot be evaluated as
degradation, even if it leads in the long run to the extinction of some organisms [48,
76]. Degradation in the spiritual sense is the transformation of a pilgrimage into a
mass tourism event by an urban society with tendencies of secularisation, as it
occurs in some Marabout places along the Moroccan coast [17], at Moulay
Abdessalam in the western Rif Mountains [141] and in the Western Ghats in India
for example [133].

5.3 Sacred Sites as Refuges for Rare or Threatened Plant

Species

It is often argued that sacred sites serve as refuges for rare and endemic species [69,
112]. The results presented here only partially confirm this assumption for northern
Morocco, as stenochorous and stenoecious plants occur only on a few of the studied
sites. An exception are stenoecious forest floor herbs and epiphytes (e.g. Davallia
canariensis) which need a forest micro-climate and an intact soil profile for their
existence. The ‘service’ of the SNS in north-western Morocco as refuges for
threatened plants seems to be limited and our study confirms the statement of
Fournier [48] for Burkina Faso: “Although they are religions sanctuaries, wooded
shrines are not necessarily biological sanctuaries in the sense of strictly protected
areas sheltering exceptional species”.
This result is not too surprising, given that many endemics and plants with
narrow niches are restricted to mountainous regions or a specific pedoclimax (cliffs,
rocks, screes) or saline conditions (coastal cliffs and dunes). The investigated sacred
412 U. Deil et al.

sites were usually not situated in extreme locations concerning inclination, altitude,
soil depth, and water availability [65]. In most cases, SNS occur in a climatic
climax situation. Nonetheless, places with special environmental features or objects
of nature like mountains, sources, ponds, coastal dunes, or caves can be sacred in
Morocco [15, 38, 57]. They are often well-known, highly frequented pilgrimage
places (e.g. [104, 141]). These examples are, however, the exception rather than the
rule (see [35] for two sites in coastal environments).
The bulk of potentially threatened plant species found on the sacred sites con-
sisted of mesophilous forest species and perennial potential grazing decreasers.
Overgrazing has been identified as a main threat to terrestrial ecosystems in the
Mediterranean [79, 100]. Mesophilous forest floor species in the southern and
southwestern Mediterranean biome might become locally endangered due to the
ongoing decline and degradation of relatively intact close-canopy forests (see e.g.
[2, 105]). The survival of viable populations of mesophilous forest species and
potential grazing decreasers on sacred sites probably contributes to the conservation
of genetic diversity in the respective species, especially in heavily exploited
agrarian landscapes.
As suggested by Bhagwat and Rutte [19] for India and for regions with fertile
soils and high population densities, a potential nature conservation role of sacred
sites in north-western Morocco could be the protection of ‘normal’ habitats. Such
‘normal’ sacred sites with a relatively low anthropogenic impact could fill a con-
siderable void in the existing conservation network. In Morocco, as elsewhere in the
Mediterranean Basin, densely populated lowland and mid altitude regions are
underrepresented in the network of protected areas [39, 128].
It is a general rule in phytogeography that species, distributed over a broad
macroclimatic gradient, become more and more stenoecious at the margin of their
range and exist there in fragmented and small populations. Such locally rare plants
could possibly be protected on ‘normal’ sacred sites. The Tangier Peninsula is the
southernmost range limit of a number of Euro-Siberian plants, some of them
endangered in NW Morocco due to increasing land use pressure and spreading
urbanisation [79]. The long-term protection of locally rare taxa is important because
peripheral populations of widespread species can feature a genetic distinctiveness,
potentially providing flexibility in times of global change [75, 125].
Surprisingly, the size of the sacred site had no significant effect on the refugial
value. This outcome is probably related to the relatively small size variance. Most
of the 86 sites were smaller than 1 ha. In India, small moist evergreen sacred forest
remnants had less endemic tree species in comparison to adjacent large state forest
reserves [56, 91]. Wassie [135] and Abiyu et al. [3], on the contrary, found that size
had no significant effect on species composition of sacred dry evergreen forests in
Ethiopia and concluded that even small forest patches could be useful tools for
biodiversity conservation and serve as source populations for native tree species
and recolonization of deforested areas. To investigate this hypothesis for sacred
sites in Morocco however, more data on the dispersal strategies of the species and
the effects of fragmentation on extinction risk and in-breeding depression are
needed.
Vegetation and Flora of Sacred Natural Sites … 413

6 Conclusions

The study of SNS in northern Morocco in the last decade provided new insights into
the role SNS can play for the conservation of near-natural habitats and vulnerable
plant species. The importance of sacred sites for the protection of near-natural
forests in northern Morocco has already been known and studied since the 1930s
(e.g. [12, 44, 98]) and could be confirmed by us [52, 53]. It seemed promising that
almost half of the total investigated sacred site area was composed of forest, but
large parts of the forests were degraded and/or subjected to intense browsing,
preventing successful regeneration [53, 66]. Figure 10 illustrates such a sacred
forest on the mountain top of Jbal Lhabib, characterized by an old-growth Quercus
suber forest with senescent Cork Oaks and missing tree regeneration since decades.
A similar stand structure has been reported by Ajbilou et al. [4] for sacred forests in
NW Morocco, an over-aged tree layer in Tibetian SNS by Salick et al. [109].
Frascaroli et al. [50] report a clustering of giant trees and patches of old-growth
forests in the vicinity of SNS in Central Italy, with higher total basal area than in
non-sacred forests nearby. Gao et al. [55] stated a higher tree layer and larger DBH
in temple and cemetery forests in SE China in comparison to no-sacred forests.

Fig. 10 Over-aged Cork Oak forest of Sidi Habib (Western Rif ranges). Grazing, trampling and
camping during the annual pilgrimage period prevented any regeneration of Quercus suber in the
last few decades
414 U. Deil et al.

Fig. 11 Sacred grove with Dwarf Palm (Chamaerops humilis) in arborescent form on a cemetery
near Basra, Province Ouezzane, NW Morocco

The ecological role of small forest patches has recently been reviewed for
Central Europe. A number of ecosystem services listed by Decocq et al. [30] for
such forest fragments (lifeboat habitat for forest species, buffering of macroclimatic
changes by a closed tree canopy, aesthetic value, construction of a local identity
etc.) can be assigned to vegetation patches surrounding Marabout tombs in
Morocco. Their main conservation potential can be seen in their role as refuges for
ombrophilous forest-floor specialists and grazing sensitive plants. Furthermore,
their ecosystem values “aesthetic landscape elements” and “contribution to the
tribal identity” (called “materialization of the noosphere in the biosphere” by [31])
are obvious.
The founders of the sacred sites did not intend to conserve organisms or habitats.
This was just a side effect in the protection of a place due to spiritual reasons (see
[117] and [68] for West Africa, [133] for India). Site selection did not follow any
spatial conservation prioritisation strategies. Many sacred sites in Morocco are
situated near settlements in non-extreme topographic locations and have zonal soils.
This circumstance may be seen as a disadvantage, concerning the protection of
stenoecious and endemic species, but it may also be seen as an advantage: for the
required protection of ‘normal’ habitats and their biocoenoses (e.g. the avifauna, see
[22]) in the light of agricultural intensification processes and ongoing degradation
Vegetation and Flora of Sacred Natural Sites … 415

of forests in the southern Mediterranean. Due to their high number in the rural
landscape, sacred sites may serve as stepping stones in a habitat network [140].
Concrete measures should be taken to support local communities and caretakers
in the protection and management of their sacred sites. The exclusion of livestock—
at least for a number of years or on parts of the site—would be a suitable man-
agement to guarantee the regeneration of woody species and to conserve the forest
patches in the long term. A second strategy, to be applied to SNS with semi-natural
vegetation, is moderate browsing of the shrubland and seasonal grazing of the
herbaceous patches. This will keep these secondary habitats open, ensure the per-
sistence of a fine-grained vegetation mosaic with a high beta-diversity and support,
nevertheless, grazing-sensitive non-forest species. To reduce the considerable
impact on some sacred sites caused by the defecation of livestock resting under tree
shade (see Fig. 9), the planting of tree groups nearby the sacred sites could be
promoted.
Highest priority of conservation (no grazing, fire or logging) however should be
given to the SNS with near-natural forests sensu Jäckle [66]: These lasts remnants
of close-to-climax forests are the only sites to study ecosystem qualities such as
maximum standing crop, maximum stand density and height, regeneration beha-
viour of the matrix species, nutrient cycles, intact soils profiles and natural
regeneration of tree species. The potential growth rates of woody plants, existing
outside of SNS only in degraded shrubby form (e.g. Quercus coccifera, Phillyrea
latifolia, Pistacia lentiscus, Arbutus unedo, Chamaerops humilis), can be studied
here. An example of such ecosystem and forest regeneration studies in Indian
sacred groves is summarized by Pandey [93]. These characters of sacred groves
might become even more important in the future, because degradation of forests and
reduction of the forested area is still going on in the southern Mediterranean
countries [20, 29, 89, 105].
At the moment, only very few sacred sites in Morocco are nature reserves in a
legal sense like Sidi Bourhaba near Rabat. The proposed classification system of
naturalness, the risk factors and the vulnerability categories for plant species
developed here may serve as a starting point for a screening of Moroccan SNS with
regard to their conservation value. Due to the large number of sacred sites a pri-
oritisation according to attributes like size, level of naturalness, occurrence of
vulnerable species and location in intensely used agricultural landscapes seems
inevitable. Local societies and caretakers should be supported in the protection and
management of their sacred sites by the nature conservancy administration and/or
by NGOs. However, imposing the value system of an urban Moroccan society or a
western secular NGO on a traditional rural group is not without problems [24, 61,
64, 90, 115]. It should always be kept in mind that such a top-down approach of
conservation policy may conflict with the spiritual value. Imposing restrictions and
regulations from outside could undermine the intrinsic motivations of the local
societies to protect “their” sacred site.

Acknowledgements Sincere thanks are given to the people living in the study area for their great
hospitality and the local authorities for allowing us to enter their sacred sites and cemeteries. We
416 U. Deil et al.

are indebted to the Moroccan administration for research permissions. We thank our colleagues
and co-workers Achmed Achhal El Kadmiri (deceased), Khadija Fateh, Lahcen Jahouer and Vlad
Tataru for the fruitful cooperation. Financial support by the German Research Foundation
(DFG) and by the German Federal Ministry of Economic Cooperation and Development (BMZ) is
greatly acknowledged (Az. DE 402/6-1, 2).

References

1. African Plants Database (version 3.3.4) (2011) Conservatoire et jardin botaniques de la ville
de Genève and South African National Biodiversity Institute Pretoria. https://www.ville-ge.
ch/musinfo/bd/cjb/africa/. Accessed June 2011
2. Aafi A (2007) Étude de la diversité floristique de l'écosystème de chêne-liège de la forêt de la
Mamora. Dissertation, Institut Agronomique et Vétérinaire Hassan II, Rabat
3. Abiyu A, Tekelay D, Glatzel G, Gratzer G (2016) Seed production, seed dispersal and
seedling establishment of two Afromontane tree species in and around a church forests:
implications for practice. For Ecosyst. https://doi.org/10.1186/s40663-016-0076-5
4. Ajbilou R, Maraňon T, Arroyo J (2006) Ecological and biogeographical analyses of
Mediterranean forests of northern Morocco. Acta Oecol 29:104–113
5. Aerts RK, Van Overtveld K, Haile M, Hermy M, Deckers J, Muys B (2006) Species
composition and diversity of small afromontane forest fragments in northern Ethiopia. Plant
Ecol 187:127–142
6. Agnoletti M, Rotherham ID (2015) Landscape and biocultural diversity: an introduction.
Biodivers Conserv 24:3155–3165
7. Agnoletti M, Emanueli F (eds) (2016) Biocultural diversity in Europe. Environmental
history, vol 5
8. Anderson DM, Salick J, Moseley RK, Ou X (2005) Conserving the sacred medicine
mountains: a vegetation analysis of Tibetan sacred sites in Northwest Yunnan. Biodivers
Conserv 14:3065–3091
9. Barbero M, Quézel P, Rivas-Martínez S (1981) Contribution à l’étude des groupements
forestiers et préforestiers du Maroc. Phytocoenologia 9:311–412
10. Barbero M, Bonin G, Loisel R, Quézel P (1990) Changes and disturbances of forest
ecosystems caused by human activities in the western part of the Mediterranean basin.
Vegetatio 87:151–173
11. Benabid A (1983) Études biogéographique et dynamique des peuplements forestiers du Rif
(Maroc). Annales De La Recherche Forestière Au Maroc 23:49–129
12. Benabid A (1984) Étude phytoécologique des peuplements forestiers et préforestiers du Rif
centro-occidental (Maroc). Travaux de l’Institut Scientifique, Série Botanique 34:1–64
13. Benabid A (1995) Les problèmes de préservation des écosystèmes forestiers marocains en
rapport avec le développement socio-économique. In: Bencherifa A, Swearingen WD
(eds) L'Afrique du nord face aux menaces écologiques. Université Mohammed V., Rabat.
Publications de la Faculté des Lettres et Sciences Humaines 50:109–124
14. Benabid A, Fennane M (1994) Connaissances sur la végétation du Maroc: Phytogéographie,
phytosociologie et séries de végétation. Lazaroa 14:21–97
15. Ben-Ami I (1998) Saint veneration among the Jews of Morocco, Jerusalem
16. Berriane M (1989) Tourisme national et migrations de loisir au Maroc. Étude Géographique.
Thesis, Tours
17. Berriane M (1990) Tourisme national et migrations de loisir au Maroc: acculturation ou
évolution interne? In: Bencherifa A, Popp H (eds) Le Maroc—espace et société, Passau.
Passauer Mittelmeerstudien 1:195–214
18. Berriane M (1992) Tourisme national et migrations de loisir au Maroc (Étude géographique).
Publ. Fac. Lettres Université Mohammed V, Sér. Thèses et Mém. 16, Rabat
Vegetation and Flora of Sacred Natural Sites … 417

19. Bhagwat SA, Rutte C (2006) Sacred groves: potential for biodiversity management. Front
Ecol Environ 4:519–524
20. Bijaber N, Ahlafi Z (2005) Projet FORMA: Etudes détaillées pour la cartographie des
changements dans le cadre de l’Inventaire Forestier National. Géo Observateur 14:13–21
21. Bongers F, Wassie A, Sterck FJ, Bekde T, Tekelay D (2006) Ecological restoration and
church forests in northern Ethiopia. J Dryland 1:35–45
22. Brandt JS, Wood EM, Pidgeon AM, Han L, Fang Z, Radeloff VC (2013) Sacred forests are
keystone structures for forest bird conservation in southwest China’s Himalayan Mountains.
Biol Conserv 166:34–42
23. Braun-Blanquet J (1964) Pflanzensoziologie. Grundzüge der Vegetationskunde, 3rd edn.
Wien, New York, 865 p
24. Brown K (2003) Three challenges for a real people-centered conservation. Glob Ecol
Biogeogr 12:89–92
25. Campbell MON (2005) Sacred groves for forest conservation in Ghana’s coastal savannas:
assessing ecological and social dimensions. Singap J Trop Geogr 26:151–169
26. Castroviejo S (Coord.) (1986–2010) Flora Ibérica. Plantas vasculares de la Península Ibérica
e Islas Baleares, vol 21. Real Jardin Bot. CSIC, Madrid
27. Chabbi A (1994) Conditions écologiques et dynamiques de la végétation des basses
montagnes du Rif occidental au Maroc. Dissertation Université Michel de Montaigne,
Bordeaux
28. Choi JK, Kim DY (2014) Characteristics and management of old and sacred Dangsan forests
in Korea. In: Hong SK, Bogaert J, Min Ch (eds) Biocultural landscapes—diversity, functions
and values, pp 129–142
29. Dahmani-Megrerouche M (2002) The holm oak (Quercus rotundifolia) in Algeria:
climate-vegetation relations. Phytocoenologia 32:129–141
30. Decocq G, Andrieu E, Brunet J, Chabrerie O, De Frenne P, Wulf M et al (2016) Ecosystem
services from small forest patches in agricultural landscapes. Curr For Rep 2:30–44
31. Deil U (2003) Holy forests in Northern Morocco. a materialization of the noosphere in the
biosphere. Bocconea 16:879–904
32. Deil U, Culmsee H, Berriane M (2005) Sacred groves in Morocco—a society’s conservation
of nature for spiritual reasons. Silva Carelica 49:185–201
33. Deil U, Culmsee H, Berriane M (2008) Sacred groves in Morocco: vegetation mosaics &
biological values. In: Sheridan MJ, Nyamweru C (eds) African sacred groves. Ecological
dynamics and social change, Oxford, pp 87–102
34. Deil U, Frosch B, Jäckle H, Mhamdi A, Achhal El Kadmiri A (2009) A geobotanical
analysis of forest patches on sacred sites in Northern Morocco. Ber. d. Reinh.-Tüxen-Ges.
21:174–195
35. Deil U, Mhamdi A (2015) Les sites maraboutiques—un patrimoine naturel et culturel du
Maroc. In: El Fasskaoui B, Kagermeier A (eds) Patrimoine et tourisme culturel au Maroc.
Actes du 9ème colloque maroco-allemand de Meknes 2014 = Publ. Fac. Lettres et Sc.
Humaines, Universitè Moulay Ismail, Sér. Actes de Colloques 43:171–178
36. Deil U, Mhamdi A (in Press) Flore et végétation des sites sacrés au Maroc—cas de
marabouts et cimetières, 185 p
37. Demdam H, Taiqui L, Seva E (2008) Vers une base de données spatiales des sites sacrés de
la province de Tétouan (N du Maroc). Apport de la cartographie officielle. Mediterranea 19
(2):9–68
38. Dermenghem E (1954) Le culte des saints dans l’Islam maghrébin, Paris
39. Derneği D (2010) Ecosystem profile: mediterranean basin biodiversity hotspot. Final report
for submission to the CEPF Donor Council, 27 July 2010
40. Dudley N, Higgins-Zogib L, Mansourian S (2006) Beyond belief: linking faith and protected
areas for biodiversity conservation. WWF, Gland, 143 p
41. Dudley N, Higgins-Zogib L, Mansourian S (2009) The links between protected areas, faiths,
and sacred natural sites. Conserv Biol 23:568–577
418 U. Deil et al.

42. Dudley N, Bhagwat S, Higgins-Zogib L, Lassen B, Verschuuren B, Wild R (2010)

Conservation or biodiversity in sacred natural sites in Asia and Africa: a review of the
scientific literature. In: Verschuuren B, Wild R, McNeely J, Oviedo G (eds) Sacred natural
sites, pp 19–32
43. Ellenberg H (1963) Vegetation Mitteleuropas mit den Alpen in kausaler, dynamischer und
historischer Sicht, Stuttgart, 943 p
44. Emberger L (1939) Aperçu général sur la végétation du Maroc. Commentaire de la carte
phytogéographique du Maroc 1:500.000. In: Rübel E, Lüdi W (eds) Ergebnisse der
Internationalen Pflanzengeographischen Exkursion durch Marokko und Westalgerien, Bern,
pp 40–158
45. Fennane M (2003) Inventaire des communautés végétales à l’aide du phytosociologie, au
Maroc. Ecologia Mediterr 29:87–106
46. Fennane M (2006) Esquisse des bases d'une typologie des habitats au Maroc. Bull de
l’Institut Scientifique, Séction Sci de la Vie 28:1–5
47. Fennane M, Ibn Tattou M (1998) Catalogue des plantes vasculaires rares, menacées ou
endémiques du Maroc, vol 8. Bocconea, pp 1–243
48. Fournier A (2011) Consequences of wooded shrine rituals on vegetation conservation in
West Africa: a case study from the Bwaba cultural area (West Burkina Faso). Biodivers
Conserv 20:1895–1910
49. Frascaroli F (2013) Sacred sites for the conservation of biodiversity. Dissert. University of
Zurich, 195 p
50. Frascaroli F, Bhagwat S, Guarino R, Chiarucci A, Schmid B (2016) Shrines in Central Italy
conserve plant diversity and large trees. Ambio 45:468–479
51. Frosch B (2010) Characteristics of the vegetation of tree stands on sacred sites in comparison
to well-preserved forests in northwestern Morocco. Ecologia Mediterr 36:83–95
52. Frosch B, Deil U (2011) Forest vegetation on sacred sites of the Tangier Peninsula (NW
Morocco) - discussed in a SW-Mediterranean context. Phytocoenologia 41:153–181
53. Frosch B, Jäckle H, Mhamdi A, El Kadmiri A, Rudner M, Deil U (2016) Sacred sites in
north-western Morocco—naturalness of their vegetation and conservation value for
vulnerable plant species. Feddes Repertorium 127:83–103
54. Gadgil M, Vartak VD (1976) The sacred groves of Western Ghats in India. Econ Bot
30:152–160
55. Gao H, Ouyang Z, Chen S, van Koppen CSA (2013) Role of culturally protected forests in
biodiversity conservation in Southeast China. Biodivers Conserv 22:31–544
56. Garcia CA, Pascal JP, Kushalappa CG (2006) Les forêts sacrées du Kodagu en Inde:
écologie et religion. Bois et Forêts des Tropiques 288:5–13
57. Gellner E (1969) Saints of the Atlas, Chicago
58. Gharbaoui A (1981) La terre et l'homme dans la péninsule Tingitane. Trav. Institut Sci.,
Série Géologie et Géographie Physique, vol 15, pp 1–439
59. Githitho AN (2003) The sacred Mijikenda Kaya forests of coastal Kenya and biodiversity
conservation. In: Lee C, Schaaf T (eds) The importance of sacred natural sites for
biodiversity conservation. Proceedings of the international workshop held in Kunming and
Xishuangbanna biosphere reserve, P. R. China, 17–20 February 2003. UNESCO, Paris,
pp 27–35
60. Guinko S (1985) Contribution à l’étude de la végétation et de la flore du Burkina Faso. 1.
Les reliques boisées ou bois sacrés. Rev. Bois Forêts Trop 208:29–36
61. Hay-Edie T, Hadley M (1998) Natural sacred sites - a comparative approach to their cultural
and biological significance. In: Ramakrishnan PS, Saxena KG, Chandrashekara UM
(eds) Conserving the sacred for biodiversity management, Oxford, New Delhi, pp 47–67
62. Hong SK, Bogaert J, Min Ch (eds) (2014) Biocultural landscapes—diversity, functions and
values, 218 p
Vegetation and Flora of Sacred Natural Sites … 419

63. Huabin H (2003) Sacred natural sites in Xishuangbanna, south-western China. In: Lee C,
Schaaf T (eds) The importance of sacred natural sites for biodiversity conservation.
Proceedings of the international workshop held in Kunming and Xishuangbanna biosphere
reserve, P. R. China, 17–20 February 2003. UNESCO. Paris, pp 127–133
64. Hughes JD, Chandran MDS (1998) Sacred groves around the earth: an overview. In:
Ramakrishnan PS, Saxena KG, Chandrashekara UM (eds) Conserving the sacred for
biodiversity management, Oxford, New Delhi, pp 71–86
65. Jäckle H, Rudner M, Deil U (2013) Density, spatial pattern and relief features of sacred sites
in northern Morocco—an analysis based upon topographic maps. Landscape Online 32:1–16
66. Jäckle H (2017) Conservation status of the vegetation of sacred natural sites in north-western
Morocco. Ecologia Mediterr 43:65–83
67. Juhé-Beaulaton D (2008) Sacred forests and the global challenge of biodiversity
conservation: The case of Benin and Togo. J Study Relig Nat Cult 2:351–372
68. Juhé-Beaulathon D (ed) (2010) Forêts sacrées et sanctuaires boisés—des creations
culturelles et biologiques (Burkina Faso, Togo, Bénin), Paris
69. Khumbongmayum AD, Khan ML, Tripathi RS (2005) Sacred groves of Manipur, northeast
India: biodiversity value, status and strategies for their conservation. Biodivers Conserv
14:1541–1582
70. Kibet S (2011) Plant communities, species diversity, richness, and regeneration of a
traditionally managed coastal forest, Kenya. For Ecol Manag 261:949–957
71. Kokou K, Kokutse AD (2006) Rôle de la régénération naturelle dans la dynamique actuelle
des forêts sacrées littorales du Togo. Phytocoenologia 36:403–419
72. Kokou K, Adjossou K, Kokutse AD (2008) Conserving sacred and riverside forests and
criteria and indicators of forest management in low wood producing communities: the case
of Togo. Ecol Ind 8:58–169
73. Küchler AW (1988) A physiognomic and structural analysis of vegetation. In: Küchler AW,
Zonneveld IS (eds) Vegetation mapping. Handbook of vegetation science, Dordrecht, vol 10,
pp 37–50
74. Lang H (1992) Der Heiligenkult in Marokko. Formen und Funktionen der Wallfahrten,
Passau
75. Lesica P, Allendorf FW (1995) When are peripheral populations valuable for conservation?
Conserv Biol 9:753–760
76. Liberski-Bagnoud D, Fournier A, Nignan S (2010) Les “bois sacrés”, faits et illusions.
A propos des sanctuaires boisées des Kasena (Burkina Faso). In: Juhé-Beaulaton D
(ed) Forêts sacrées et sanctuaires boisées: Des creations culturelles et biologiques (Burkina
Faso, Togo, Benin), Paris, pp 59–90
77. Malhotra KC, Gokhle Y, Chatterjee S, Srivastava S (2001) Cultural and ecological
dimensions of sacred groves in India. INSA, New Delhi
78. Mallarach JM, Papayannis T, Vaisanen R (eds) (2012) The diversity of sacred lands in
Europe. Proceedings of the third workshop of the Delos initiative—Inari/Ananaar 2010.
IUCN, Gland, 292 p
79. Mathez J, Quézel P, Raynaud C (1985) The Maghreb countries. In: Gomez-Campo C
(ed) Plant conservation in the Mediterranean area, The Hague, pp 141–157
80. Médail F, Quézel P (1997) Hot-spots analysis for conservation of plant biodiversity in the
Mediterranean Basin. Ann Mo Bot Gard 84:112–127
81. Mgumia FH, Oba G (2003) Potential role of sacred groves in biodiversity conservation in
Tanzania. Environ Conserv 30(3):59–265
82. Miehe G, Miehe S, Will M, Opgenoorth L, Duo L, Dorgeh T, Liu J (2008) An inventory of
forest relicts in the pastures of Southern Tibet (Xizang A.R., China). Plant Ecol 194:157–177
83. Mishra BP, Tripathi OP, Tripathi RS, Pandey HN (2004) Effects of anthropogenic
disturbance on plant diversity and community structure of a sacred grove in Meghalaya,
northeast India. Biodivers Conserv 13:421–436
84. Molnár AV, Nagy T, Löki V, Tökölyi J, et al (2017) Turkish graveyards as refuges for
orchids against tuber harvest. Ecol Evol https://doi.org/10.1002/ece3.3562
420 U. Deil et al.

85. Moore HM, Fox HR, Harrouni MC, El Alami A (1998) Environmental challenges in the Rif
mountains, northern Morocco. Environ Conserv 25:354–365
86. Mwihomeke ST, Msangi TH, Mabula K, Ylhäisi J, Mndeme KCH (1998) Traditionally
protected forests and nature conservation in the North Pare Mountains and Handeni District,
Tanzania. J East Afr Nat Hist 87:279–290
87. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity
hotspots for conservation priorities. Nature 403:853–858
88. Neumann K, Müller-Haude P (1999) Forêts sèches au sud-ouest du Burkina Faso: végétation
- sols - action de l’homme. Phytocoenologia 29:53–85
89. Nsibi R, Souayah N, Khouja LM, Khaldi A, Bouzid S (2006) Impacts des facteurs biotiques
et abiotiques sur la dégradation de subéraie tunisienne. Geo-Eco-Trop 30:25–34
90. Ormsby AA (2011) The impacts of global and national policy on the management and
conservation of sacred groves of India. Hum Ecol 39:783–793
91. Page NV, Qureshi Q, Rawat GS, Kushalappa CG (2010) Plant diversity in sacred forest
fragments of Western Ghats: a comparative study of four life forms. Plant Ecol 206:237–250
92. Pala NA, Negri AK, Gokhale Y, Todaria NP (2013) Three regeneration status of sacred and
protected landscapes in Garhwal Himalaya, India. J Sustain For 32:39–246
93. Pandey HN (2010) Sacred forests: their ecology and diversity, New Delhi, 197 p
94. Quézel P (1974) Les forêts du pourtour méditerranéen. UNESCO, Notes Techniques du
MAB, vol 2, pp 1–53
95. Quézel P, Barbero M (1982) Definition and characterization of Mediterranean-type
ecosystems. Ecologia Mediterr 8:15–29
96. Quézel P, Barbero M (1990) Les forêts méditerranéennes - problèmes posés par leur
signification historique, écologique et leur conservation. Acta Bot Malacitana 15:145–178
97. Quézel P, Santa S (1962–63) Nouvelle Flore de l’Algérie et des régions désertiques
méridionales, Paris, vol 2
98. Quézel P, Médail F (2003) Écologie et biogéographie des forêts du bassin méditerranéen,
Paris
99. R Core Team (2013) R: a language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna. https://www.R-project.org/
100. Radford EA, Catullo G, de Montmollin B (eds) (2011) Important plant areas of the south and
east Mediterranean region: priority sites for conservation. IUCN, Malaga
101. Ramakrishnan PS, Saxena KG, Chandrashekara UM (eds) (1998) Conserving the sacred for
biodiversity management, Oxford, New Delhi, 480 p
102. Ramanujam MP, Kadamban D (2001) Plant biodiversity of two tropical dry evergreen
forests in the Pondicherry region of South India and the role of belief systems in their
conservation. Biodivers Conserv 10:1203–1217
103. Rankou H, Culham A, Jury SL, Christenhusz MJM (2013) The endemic flora of Morocco.
Phytotaxa 78:1–69
104. Regato P, Salman R (2008) Mediterranean mountains in a changing world: Guidelines for
developing action plans. IUCN Centre for Mediterranean Cooperation, Malaga
105. Rejdali M (2005) Forest cover changes in the Maghreb countries with special reference to
Morocco. In: Mazzoleni S, di Pasquale D, Mulligan M, di Martino P, Rego F (eds) Recent
dynamics of the Mediterranean vegetation and landscape, Chichester, pp 23–31
106. Rivas-Martínez S, Fernández-González F, Loidi J, Lousã M, Penas A (2001)
Syntaxonomical checklist of vascular plant communities of Spain and Portugal to
association level. Itinera Geobotanica, vol 14, pp 5–341
107. Rivas-Martínez S, Díaz TE, Fernández-González F, Izco J, Loidi J, Lousã M, Penas A
(2002) Vascular plant communities of Spain and Portugal. Addenda to the syntaxonomical
checklist of 2001. Itinera Geobotanica, vol 15, pp 5–922
108. Rotherham ID (2015) Bio-cultural heritage and biodiversity: emerging paradigms in
conservation and planning. Biodivers Conserv 24:3405–3429
Vegetation and Flora of Sacred Natural Sites … 421

109. Salick J, Amend A, Anderson D, Hoffmeister K, Gunn B, Fang Z (2007) Tibetan sacred sites
conserve old growth trees and cover in the eastern Himalayas. Biodivers Conserv 16(3):693–
706
110. Sauvage C (1961) Recherches géobotaniques sur les subéraies marocaines. Trav. Institut Sci.
Chérifien, Série Botanique, Rabat, vol 21, pp 1–462
111. Sauvage C (1963) Étages bioclimatiques. Notices explicatives [in Atlas du Maroc]. Comité
National de Géographie du Maroc, Rabat
112. Schaaf T (2006) Linking cultural and biological diversity: the UNESCO-MAB approach. In:
Schaaf T, Lee C (eds) Conserving cultural and biological diversity: the role of sacred natural
sites and cultural landscapes. Proceedings of the international workshop held in Tokyo
(Japan), 30 May–2 June 2005, Paris, pp 12–14
113. Schaaf T, Lee C (eds) (2006) Conserving cultural and biological diversity: the role of sacred
natural sites and cultural landscapes. UNESCO, Paris, 341 p
114. SEEE (Secrétariat d’Etat chargé de l’Eau et de l’Environnement - Département de
l’Environnement) (2014) Cinquième rapport national sur la biodiversité. Rabat : Secrétariat
d’Etat auprès du Ministère de l’Energie, des Mines, de l’Eau et de l’Environnement, chargé
de l’Eau et de l’Environnement, Dép. Environnement, Rabat. https://ma.chm-cbd.net/
implementation/rap_nat/ma-nr-05-fr.pdf
115. Sheridan MJ (2008) The dynamics of African sacred groves. Ecological, social and symbolic
processes. In: Sheridan MJ, Nyamweru C (eds) African sacred groves, Oxford, pp 9–41
116. Sheridan MJ, Nyamweru C (eds) (2008) African sacred groves. Ecological dynamics and
social change, Oxford
117. Siebert U (2008) Are sacred forests in Northern Benin “traditional conservation areas”? In:
Sheridan MJ, Nyamweru C (eds) African sacred groves, Oxford, pp 164–177
118. Sukumaran S, Jeeva S, Raj ADS, Kannan D (2008) Floristic diversity, conservation status
and economic value of miniature sacred groves in Kanyakumari District, Tamil Nadu,
Southern Peninsular India. Turk J Bot 32:185–199
119. Suter G (1980) Carte géologique de la chaîne rifaine 1:500.000. Notes et mémoires du
service géologique du Maroc 245a, Rabat
120. Taïqui L, Seva E, Roman JL, R’ha A (2005) Los bosquetes de los khaloa (morabitos) del
Rif, Atlas Medio y región de Sus de Marruecos. Ecosistemas 14:31–41
121. Taïqui L, Bensalah H, Seva E (2009) La conservation des sites naturels sacrés an Maroc:
est-elle incompatible avec le développement socio-économique? Mediterr Serie de Estudios
Biológicos 20:106–142
122. Tchamié TTK (2000) Évolution de la flore et de la végétation des bois sacrés des Massifs
Kabyè et des régions environnantes (Togo). Lejeunia 164:1–36
123. Ter Braak CJF, Šmilauer P (2002) CANOCO reference manual and CanoDraw for Windows
user’s guide: software for canonical community ordination (version 4.5). Microcomputer
Power, Ithaca, 500 p
124. Thuiller W, Lavorel S, Araújo MB, Sykes MT, Prentice IC (2005) Climate change threats to
plant diversity in Europe. Proc Natl Acad Sci USA 102:8245–8250
125. Thuiller W, Albert C, Araújo M, Berry PM, Cabeza M, Guisan A, Hickler T, Midgley GF,
Patterson J, Schurr FM, Sykes MT, Zimmerman N (2008) Predicting global change impacts
on plant species’ distributions: future challenges. Perspect Plant Ecol Evol Syst 9:137–152
126. Tiwari BK, Barik SK, Tripathi RS (1998) Biodiversity value, status, and strategies for
conservation of sacred groves of Meghalaya, India. Ecosyst Health 4:20–32
127. Tomaselli R (1976) La dégradation du maquis méditerranéen. In: Forêts et maquis
méditerranéens: écologie, conservation et aménagement. Notes techniques du MAB, vol 2,
pp 5–76
128. Underwood EC (2009) Expanding the global network of protected areas to save the
imperiled Mediterranean biome. Conserv Biol 23:43–52
129. Underwood EC, Viers JH, Klausmeyer KR, Cox RL, Shaw MR (2009) Threats and
biodiversity in the Mediterranean biome. Divers Distrib 15:188–197
422 U. Deil et al.

130. Valdés B, Talavera S, Fernández-Galiano E (eds) (1987) Flora vascular de Andalucía

occidental, Barcelona, vol 3
131. Valdés B, Rejdali M, Achhal El Kadmiri A, Jury SL, Montserrat JM (eds) (2002) Catalogue
des plantes vasculaires du Nord du Maroc, incluant des clés d’identification, Madrid, vol 2
132. Verschuuren B, Wild R, McNeely J, Oviedo G (eds) (2010) Sacred natural sites. Conserving
nature and culture, London, 300 p
133. Vipat A, Bharucha E (2014) Sacred groves: the consequences of traditional management.
J. Anthropol https://doi.org/10.1155/2014/595314
134. Virtanen P (2002) The role of customary institutions in the conservation of biodiversity:
sacred forests in Mozambique. Environ Values 11:227–241
135. Wassie A (2007) Ethiopian church forests. Opportunities and challenges for restoration.
Dissertation, University of Wageningen
136. Wassie A, Tekelay D (2006) Soil seed banks in church forests of northern Ethiopia:
implications for the conservation of woody plants. Flora 201:32–43
137. Wassie A, Tekelay D, Powell N (2005) Church forest in northern Gondar. Administrative
zone, northern Ethiopia. For Trees Livelihoods 15:329–374
138. Wassie A, Sterck FJ, Tekelay D, Bongers F (2009) Tree regeneration in church forests of
Ethiopia: effects of microsites and management. Biotropica 41:110–119
139. Wassie A, Sterck FJ, Bongers F (2010) Species and structural diversity of church forest in a
fragmented Ethiopian Highland landscape. J. Veget. Sci. 21:938–948
140. Wild R, McLeod C (eds) (2008) Sacred natural sites: guidelines for protected area managers.
Best practice protected area guidelines series, no 16. IUCN, Gland
141. Zouanat Z (2009) ‘Abd al-Salâm ibn Mashîsh and Jabal La‘lâm site, Northern Morocco. In:
Papayannis T, Mallarach JM (eds) The sacred dimension of protected areas. Proceedings of
the second workshop of the Delos initiative ouranoupolis, Gland, Athens, vol 2007, pp 66–
75
The Kaga Coast (Japan): A Natural
Ecosystem and Cultural Landscape
Ensuring Biodiversity

Kazue Fujiwara, Elgene O. Box, Jin Ōsaka, Hideki Sakai,

and Youji Kitaguchi

Abstract The Kaga Coast, along the Sea of Japan, has highly diverse ecosystems
involving natural and anthropogenic landscapes. The most natural area is old
Kashima island, with a Shintō shrine; it is connected to the main island and is a
National Natural Monument of Japan. Persea thunbergii forest remains under
protection and is the habitat of a terrestrial crab (akate-gani: Chiromantes haema-
tocheir). The crabs live in the forest, eat evergreen leaves, and hibernate in burrows.
The scene of akate-gani egg-laying is a good example of a food chain, and the
Persea forest seems like a terrestrial mangrove forest holding the habitat of crabs.
The long (16.4 km), high (46 m) sand dune along the Kaga Coast has been man-
aged since 1766, by plantation of black pine (Pinus thunbergii) on the back dune
and Vitex rotundifolia on the foredune. Nowadays the planted Pinus thunbergii
forests and sand-dune vegetation protect human lives and also restore and support
biodiversity. Even though people planted Pinus thunbergii, it still continues
nowadays and creates a unique landscape. First, as a planting for defense against
mobile sand, the local people combined traditional Japanese technology with
European (e.g. French) technology to make the planting succeed. Actually, the
method of making artificial foredunes and planting coastal plants was invented
independently in Edo-era Japan and in Europe. In Japan, they used dune herbs like
Carex kobomugi, Calystegia soldanella etc. and shrubs like Vitex rotundifolia. On
the back dune they planted mainly black-pine trees to protect farms and villages
from being buried by sand blown by winter winds. They gained partial, temporary
results, but could not get a final solution. On the Kaga Coast, it is confirmed that

K. Fujiwara (&)
Yokohama City University, Yokaohama, Japan
e-mail: [email protected]
E. O. Box
Geography Department, University of Georgia, Athens, GA 30602, USA
J. Ōsaka
Daishoji Honcho 56, Kaga City, Ishikawa Prefecture 922-0041, Japan
H. Sakai
Y. Kitaguchi
Culture Promotion Division, Kaga City Hall, Daishoji Minamicho 41, Kaga City, Ishikawa
Prefecture 922-0041, Japan

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 423
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_20
424 K. Fujiwara et al.

there are 17 kinds of animals and over 290 kinds of birds (including Nipponia
nippon). There are 18 kinds of “important species,” as designated in the Red Data
Book of Japan and of Ishikawa prefecture. Insects utilize both shrubs and herbs,
which also represented technological progress in the planting project. This tech-
nology and experience was highly significant for sand-defense forests in Japan. The
Kaga Coast can thus be described as an area where the human landscape and
biodiversity are in harmony.

Keywords Dune vegetation
Harmony of nature and civil engineering

Man-made coastal landscape
Shintō shrine forest
Terrestrial mangrove forest
1 Introduction

The “Kaga Coast” is the coastline of Kaga city, which extends for 16.4 km along
the Sea of Japan, from Shioya to Ikiri, in Ishikawa Prefecture. The Kaga coast
consists of three distinct areas: sandy beach from Shioya to Katano; reef-bound
beach from Katano to Cape Amagozen-misaki (this coastal area is designated as a
special protected area of Echizen-Kaga Kaigan Quasi-National Park); and gravel
beach from Amagozen to Ikiri [6, 13]. The Daishōji dune (4 km long and 1 km

Fig. 1 During the Heian Era (794-1185/1192) there was forest along the Kaga coast (photo by
Kaga City)
The Kaga Coast (Japan) … 425

wide, from Shioya to Katano) is in the sandy-beach portion of the Kaga Coast and
is a single line of dunes [2]. This dune is not only an important element that
signifies the natural landscape and ecology of the Kaga Coast; it also has had a great
historical influence on industry, social activities, and culture (Fig. 1) [1].
Sand blown by seasonal (winter) winds and actively moving sand dunes were
great threats to people’s lives and occupations. It was not rare that both their farms
and houses were threatened with burial and that people were forced to leave. In one
instance, winter winds buried the mouth of the Daishōji-gawa (river) under sand,
and the people had to remove the sand. The domain of the Daishōji clan set up the
position of the Matsu-bugyō (Pine Magistrate) and made sand-defense and wind-
break forests in cooperation with people living around the dunes.
From Katano-hama to the Uwagi-hama coast, there was Japanese cypress
(Chamaecyparis obtusa) forest, with a pond called Ō-ike (large pond) that was
made by people for agricultural use. Ō-ike was buried by blowing sand and sep-
arated into five ponds including Katano Ō-ike (the present Katano-kamoike, or
‘duck pond’). The outlet of Katano ō-ike was also buried by sand, and the volume
of water became larger year after year. Because the water of Katano ō-ike was about
to flood, the Daishōji domain made a tunnel in 1678 to let water flow to the Kanjo-
dani valley at Katano village [7, 17]. Today, the previous Ō-ike (Katano-kamoike)
is the smallest Ramsar-listed wetland in the world (see Fig. 2).
Halfway between Katano and Shioya there was, from about 1645 to 1844, a
village called Nakahama. The residents of Nakahama had come from Sanri-hama
village (Echizen province) and made a living by fishing or by making salt. They cut
many cypress and other trees as fuel for salt manufacture, so more sand was blown

Fig. 2 Elements of the natural and cultural landscape around Kaga City
426 K. Fujiwara et al.

Fig. 3 Kashima Island, the most natural area along the Kaga coast, with evergreen broad-leaved
forest dominated by Machilus (Persea) thunbergii

and the sand dunes became larger. Finally, in the Tempō period (1830 to 1844),
Nakahama was abandoned. In Uwagi village, blowing sand became violent and
buried about ten thousand square meters of farmland in 1744; in 1774 it buried
more farms and 21 houses. Especially in 1774, sand was deposited along about 450
m, 24 m high, and some residents were forced to move, creating a new Uwagi-
demura village [12].
Nowadays these foredunes and the black-pine plantation behind are stabilized,
and more natural species have come back. The most natural area is Kashima island
(Fig. 3), which has a Shintō shrine. The natural forest around the shrine, the dune
vegetation, some parts of the recovered pine forest (plantation), and the wetland
have high biodiversity. In 2004–2006, Kaga City made more plantations to restore
natural forests on the coastal hills, using potted seedlings raised by students.

2 Historical Planting Activities

Planting on the Kaga Coast was tried beginning in the Empō period (1673 to 1680)
but started in earnest in the late Edo era. In 1766 the local branch office of the
Daishōji sent hundreds of laborers to Uwagi-hama and made a sand-defense fence
about 1600 m long; they then planted black pines (Pinus thunbergii) and mimosa
(Albizia julibrissin) along the sand defense fence. The office continued planting
black pines, mimosa, willows (Salix), and miscanthus grass (Miscanthus sinensis),
while also building fences along the coastline in Katano, Uwagi, Segoe, Shioya,
The Kaga Coast (Japan) … 427

Table 1 Sand-defense plantings by the domain of Daishōji

Period Area Kind of trees Extension
1751 to 1763 Katano Black pine, Celtis, etc. About 360 m
From 1766 Uwagi Black pine, mimosa, Celtis, etc. About 1620 m
In 1786 Shioya Black pine, Celtis, etc.
In 1790 Uwagi Black pine, willow, Celtis, etc.
From 1833 Shioya Black pine, mimosa, Celtis, etc. About 324 m
From 1833 Segoe Black pine, mimosa, Celtis, etc. About 864 m
From 1836 Shinohara Black pine, mimosa, Celtis, etc. About 648 m
From 1851 Uwagi Black pine, mimosa, Celtis, etc. About 324 m

Shinohara and Ikiri. Daishōji’s retainer, Tojuro Koduka, planted hundreds of

thousands of black-pine trees in the sand-dune area. (According to Sozan Iko, he
planted about 20 million trees and other plants.) Also, the local leader at Oshiotsuji
village, Koshiro Kano 9th (generation), cooperated with Tojuro, inventing a way of
planting pines using straw bags. This sand-defense planting made the basis of
today’s coastal forest in the National Park of the Kaga Coast (Table 1) [7, 10, 14].
After the Meiji Restoration, the sand-defense plantings lost the support of the
feudal clan and were neglected for a long time, because of political and social
changes. In consequence, from 1876 to 1883, sand was deposited along about 130 m
of coastline, 110 m high, and buried large areas of farmland. In that situation, Chōhei
Nishino, son of the previous local leader Chobe, expended his own funds on planting
trees. After that, an important national project of sand-defense planting was started in
1911 and continued until 1925, hiring young laborers at Uwagi village (cf Fig. 4).
The workers made two lines of fencing in the sand-dune area, 90 m inland from the
sea. Then they planted over 4 million black-pine trees, along with alder (Alnus
japonica), robinia (Robinia pseudoacacia), and mimosa [10].
During World War II, people cut pine branches for fuel, as well as whole pine
trees to create farmland. They also damaged the trunks of pine trees in order to
gather turpentine, which the Japanese Navy used as a substitute for gasoline. As a
result, the forest went wild again. Still, people such as Nihei Nishino, son of
Chōhei, continued maintaining the sand-defense plantings by putting straw in the
fences and planting shrubs, which fertilize the land [5, 11]. The sand-defense forest
was designated “the National Park of Echizen-Kaga Coast” in 1968, and the
Forestry Agency designated the forest as a “natural recreation forest.”
The coast from Shioya to Katano is a typical cultural landscape of the Kaga
Coast, and has special significance. For sand-defense plantings, the people com-
bined traditional Japanese technology with European (especially French) technol-
ogy to make the planting succeed. At the Daishōji domain they planted mainly
black-pine trees to protect farms and villages from being buried by sand blown by
winter winds. As a result, they gained partial, temporary success, but could not gain
a radical solution.
After the Meiji Restoration, based on the demands of the people, interested
parties and Ishikawa prefecture, the government began a big project called “the
428 K. Fujiwara et al.

Fig. 4 Kaga pine plantation during the Meiji Era (photo by Kaga City)

coastal sand-defense project of beach national forest” and drastically solved the
sand problem. This project took 13 years, from 1911 to 1924, and was carried out
by the Kanazawa Forestry Office (the present Ishikawa Forest Management Office)
(cf Fig. 4). Excellent forest engineers working there made the project progress,
investigating the environment (e.g. the seasonal winds) and studying previous
experiences, including foreign cases, systematically and in detail.
One important method of planting involved utilizing wind power to make 10
meter-high sand dunes called mae-kyū (foredunes) on the seaward side of planting
areas; this protects the trees from blowing sand. On the foredunes the project
planted coastal plants such as Vitex rotundifolia, Elaeagnus pungens (shrub), sea
bells (Calystegia soldanella), Ixeris repens, Salsola komarovii, Carex kobomugi,
Wedelia prostrata, Linaria japonica, Artemisia capillaris, Elymus mollis, Imperata
cylindrica var. koenigii, and Arabis stelleri var. japonica. To prevent these plants
from being buried by sand, they made bamboo-mat fences. They also made ko-kyū
dunes (back dunes) behind the planting area [11].
Actually the method of making artificial sand dunes and planting coastal plants
was invented independently in Edo-era Japan and in Europe. In Japan people used
shrubs like Vitex rotundifolia, while in Europe people planted herbaceous plants
such as sea bells. On the Kaga Coast, people utilized both shrubs and herbs, which
represented a technological advance in the planting project. This technology and
experience had great significance for the creation of sand-defense forests in Japan.
The Kaga Coast (Japan) … 429

3 Methodology

The vegetation of Kaga City was surveyed in 1999–2001 by phytosociological

methods [3, 15] and was reported in 2002 (Kaga City 2002) [9]. The vegetation of
the 320 relevés collected was classified into 117 associations and communities.
Types of forest and dune vegetation on the Kaga Coast have been selected from this
report and are compared here for planted trees and herbs. Analysis of the data
identified those species that remained and their vegetation types, and identified
newcomers as endangered species, such as Viola spp. Species of natural forests that
could come back in planted pine forests were also identified.
Community Types
1. Natural Vegetation
a) Dune Grasslands (Table 2)
The species planted since 1911, on the 10 m sand dune, remain and have adapted to
each habitat. Sand-dune herbs planted at that time include Calystegia soldanella,
Ixeris repens, Salsola komarovii, Carex kobomugi, Wedelia prostrata, Linaria
japonica, Artemisia capillaris, Elymus mollis, Imperata cylindrica var. koenigii,
and Arabis stelleri var. japonica. These remain as the phytosociological units
Calystegio soldanellae-Salsoletum komarovi, Caricetum kobomugi association
group, Ichaemetum anthephoroidis association group, and an Elymus mollis
community.

1) Salsola komarovi occurs along the front line of the sand-dune vegetation. It was
planted as a dune-fixing herb (in 1911–1924) and remains on the front dunes,
growing mainly at about 2 m above the sea level. S. komarovi is nitrophilic and,
after plantation of the herbs on the dunes, has remained and created its habitat on
sites created by accumulation of seaweed and other drifting organic materials. S.
komarovi vegetation is classified, based on its characteristic species S. komarovi,
in the Calystegio soldanellae-Salsoletum komarovi, belonging to the higher
units Salsonion komarovii, Salsoletalia komarovii and Salsoletea komarovii.
2) The other main dune-fixing herbs (from 1911–1924) are Carex kobomugi,
Wedelia prostrata, Ixeris repens, Salsola komarovii, Linaria japonica, and
Artemisia capillaris; these continue to form the main dune vegetation. Carex
kobomugi forms a community with Calystegia soldanella, Ixeris repens, and
Glehnia littoralis on unstable dunes where wind-blown sand is still accumulating.
A Caricetum kobomugi association group occurs on the Kaga coast. One member
is the Wedelio-Caricetum kobomugi, which occurs in warm-temperate coastal
areas of Japan. Another association is the Elymo-Caricetum kobomugi, which
occurs in typical southern-temperate coastal areas. The Kaga coast has both of
these along its transitional warm-temperate to typical-temperate coast. This can
also be said for the Ichaemetum anthephoroides association group. The Caricetum
kobomugi association group is summarized as the Wedelio prostratae-Caricetum
430 K. Fujiwara et al.

Table 2 Salsoletea komarovii and Glehnietea littoralis

kobomugi, with Wedelia prostrata as character species, and the Elymo-Caricetum

kobomugi, with Elymus mollis and Linaria japonica as differential species. Both
of these species occur on northern warm-temperate and typical-temperate
coastlines. The Caricetum kobomugi association group and Ischaemetum
anthephoroides association group (see next sub-section) belong to the higher units
Caricion kobomugi, Glehnetalia littoralis, and Glehnietea littoralis. Calystegia
soldanella, Ixeris repens, and Glehnia littoralis occur as characteristic species of
higher units. Glehnia littoralis is a near-threatened species in Ishikawa Prefecture
(for more information see https://www.pref.ishikawa.lg.jp/sizen/reddata/rdb_
2010/data/documents/hamabouhuu521.pdf) [20].
The Kaga Coast (Japan) … 431

3) The Ischaemetum anthephoroides makes back-dune vegetation, where sand

movement is reduced. It occurs in two sub-types within the so-called Ischaemetum
anthephoroides association group. One association is the Wedelio-Ischaemetum
anthephoroides, with Wedelia prostrata as differential species. It occurs on
warm-temperate coasts. The other sub-type is the Linario japonicae-Ischaemetum
anthephoroides, with Linaria japonica and Artemisia capillaris as differential
species. This occurs on temperate coasts. Within this Linario
japonicae-Ischaemetum anthephoroides there occur several clumps of Viola
senamiensis (=V. grayi) (see https://www.pref.ishikawa.lg.jp/sizen/reddata/rdb_
2010/data/documents/isosumire306.pdf) [21], which is a species of endangerment
class II. This Viola senamiensis (see Fig. 5) is a perennial plant, flowering in
April-May. Its rhizomes are very long and branch upward so that the plant can
recover from sand burial. When the winter monsoon is strong, old leaves and
stalks stop and gather the moving sand, under which Viola senamiensis spends the
winter. The Kaga coast, especially the Shioya coast, is famous for clumps of Viola
senamiensis and V. mandshurica var. crassa also occurs in the Ischaemetum
anthephoroides association group on the Kaga coast. Here it is a coastal phenotype
of V. mandshurica, with thick, shiny, curved leaves. These Viola species are
newcomers after the plantation of dune herbs.
4) Elymus mollis is dominant on the Kaga coast in plantations for sand stabiliza-
tion. It also occurs naturally in the Elymo-Caricetum kobomugi. E. mollis was
one of the species planted in 1911 and has remained on the Kaga coast, albeit in
smaller populations. Recently it is still planted for controlling moving sand, but
it does not occur as a sand-dune herb otherwise.

Fig. 5 Viola senamiensis, part of the largest such community in the world (photo by I. Honda, on
23 April 2008)
432 K. Fujiwara et al.

5) The other communities and associations are the Ixerido-Fimbristyldetum ser-

iceae, the Wedelio prostratae-Zoysietum macrostachyae, and an Imperata
cylindrica var. koenigii community. The Ixerido-Fimbristyldetum sericeae occurs
in slightly concave areas of unstable dunes. The Wedelio prostratae-Zoysietum
macrostachyae occurs on small terraces of coastal cliffs of accumulated sand.
Grassy areas of Imperata cylindrica var. koenigii occur on back dunes. Usually, I.
cylindrica var. koenigii occurs in Miscanthus sinensis grasslands. It came onto the
Kaga coast after the sand became stable, next to the areas of pine plantation.
b) Dune Shrub Communities (Table 3)
Dune shrub species were planted with herb species in 1911, including mainly Vitex
rotundifolia and Elaeagnus pungens. V. rotundifolia adapted especially well to the
sand dunes and made back-dune scrub vegetation. This is the so-called
Linario-Viticetum rotundifoliae, with V. rotundifolia as characteristic species, and
Linaria japonica and Artemisia capillaris as differential species. In this association
Viola species make sub-associations. Violetosum crassae is differentiated by V.
mandshurica var. crassa, Cuscuta pentagona, C. japonica and C. chinensis. It
occurs less (10–50% dominant) than the area of Vitex rotundifolia and has been
disturbed less by humans. Other subassociations are typicum, with no differential
species; and violetosum senamiensis, with V. senamiensis, Arabis stelleri var.
japonica, and Cocculus orbiculatus as differential species. Typicum has dominant
Vitex rotundifolia with more than 60% cover on relatively stable dunes. Violetosum
senamiensis occurs on stable back dunes, where Vitex rotundifolia is not dominant
(only 5–20% cover). V. senamiensis (=V. grayi) makes relatively large colonies on
the tops of the main dunes and on back dunes. Vitex rotundifolia sites on the Kaga
coast can hold characteristic Viola colonies. Arabis stelleri var. japonica also occurs
here.
c) Evergreen Broad-Leaved Forests (Tables 4, 5)
Evergreen broad-leaved forests (Camellietea japonicae) were mostly destroyed and
remained only around Shintō shrines. Forests of Machilus (Persea) thunbergii on the
Kaga coast are especially remarkable around the Kashima Shintō shrine and Daishōji
temple (e.g. Fig. 6). These forests are summarized as the Polysticho-Perseetum
thunbergii, with truly laurophyllous P. thunbergii dominant in the forest canopy and
with several understoreys of evergreen tree and herb species. The Kashima shrine
forest is especially famous for its akate-gani (Chiromantes haematocheir) crab
population and habitat (Fig. 7). The canopy of the Kashima forest is dominated by
M. thunbergii, with a few trees of Zelkova serrata and some Castanopsis cuspidata
var. sieboldii. Forest height is about 25 m. Of the evergreen forests in Kaga City,
only the Kashima forest has Cinnamomum japonicum and Dryopteris erythrosora
var. dilatata. The Kashima forest was an island (elevation 30 m, perimeter 600 m) in
ancient times, but nowadays it is connected to the mainland. Kashima forest is
constituted mostly of evergreen broad-leaved species. Sub-canopy trees include
Camellia japonica, Aucuba japonica var. borealis, and Neolitsea sericea in the tree
and shrub layers; Dryopteris erythrosora, Kadsura japonica and Trachelospermum
asiaticum f. intermedium occur in the herb layer. Vine species such as
The Kaga Coast (Japan) … 433

Table 3 Viticetea rotundifoliae

434 K. Fujiwara et al.

Table 4 Summarized table of Ardisio-Castanopsietum sieboldii and Polysticho-Persetum thun-

bergii
The Kaga Coast (Japan) … 435

Table 4 (continued)

Daishōji Daishōji

Trachelospermum and Kadsura climb to the tree layers. Kashima forest is protected
as a National Natural Monument, especially since it is the important habitat for the
akate-gani crab (see below).
The evergreen broad-leaved forests in Kaga City include tall forests dominated
by Machilus thunbergii (Polysticho-Perseetum thunbergii) but also forests domi-
nated by Castanopsis cuspidata var. sieboldii (Ardisio-Castanopsietum sieboldii).
These two kinds of tall evergreen forest do not have obvious differential species in
Kaga City. Polysticho-Perseetum thunbergii occurs on humid sites such as around the
Kashima shrine, alluvial areas, and lower hill slopes. The Ardisio-Castanopsietum
occurs on slightly drier sites, such as ridges or upper hill slopes. Evergreen,
wind-exposed shrub thickets also occur on the Kaga coast, especially the
Euonymo-Pittosporetum tobira. These thickets, usually 5–10 m tall, have
Pittosporum tobira, Camellia japonica and Ilex integra as dominant species.
Sometimes Pinus thunbergii also occurs and the height becomes 10 m. The com-
bined cover of tree and shrub layers is 80–100%, so these thickets work well as
436 K. Fujiwara et al.

Table 5 Euonymo-Pittosporetum tobira

Mean altitude(m): 18
Mean height of tree layer (m): 9
Mean number of species: 17
Number of relevés: 3
Ch. & diff. spp. of association:
Euonymus japonicus S,H 2(1)
Rosa multiflora S,H 2(+)
Paederia scandens H 2(+)
Pittosporum tobira S,H 2(2-4)
Farfugium japonicum H 2(+-1)
Cocculus orbiculatus S 2(+)
Pinus thunbergii T1,S 1(3)
Neolitsea sericea S,H 1(1)
Rhaphiolepis umbellata S 1(+)
Cyrtomium falcatum H 1(+)
Ch. & diff. spp. of Camellietea japonicae:
Camellia japonica T1,S,H 3(2-4)
Machilus thunbergii S 2(1-2)
Ilex integra T1,S 3(+-3)
Castanopsis cuspidata var. sieboldii T1,T2,S,H 2(1-2)
Ligustrum japonicum S,H 2(+)
Kadsura japonica S,H 2(+)
Companions:
Wisteria floribunda T1,S 3(+)
Celtis sinensis var. japonica S,H 2(+-1)
Smila china S,H 2(+)
Aucuba japonica var. borealis S,H 2(+-1)
etc.
Location: Hashitate-machi, Kashima-no-mori

windbreaks. Most evergreen broad-leaved forests were recorded on the Kaga coast
areas (Table 4, running numbers 1–4, Kaga City 2002). These evergreen
broad-leaved forest species recover in old pine plantations over the years.
2) Pinus thunbergii-Pinus densiflora Plantations (Table 6)
Historical treatments describe the planting of Pinus thunbergii and P. densiflora on
back dunes, for protection against moving sand, since 1911. Nowadays these
plantations have developed into somewhat more natural pine forests and have
become stable. Pine plantations are normally 5 m to 20 m tall, depending on
plantation year and site conditions.
The pine plantations have developed into two types of forest. One is already the
potential vegetation (evergreen broad-leaved forest: Camellietea japonicae), with
many naturally recovering species, e.g. M. thunbergii, I. integra, I. pedunculosa,
The Kaga Coast (Japan) … 437

Fig. 6 Old Machilus (Persea) thunbergii in remnant of the natural forest around the Kashima
Shintō shrine

Fig. 7 Native akate-gani crabs (Chiromantes haematocheir) climbing in arborescent evergreen

Aucuba japonica var. borealis, on the Kaga coast. (Photo by Prof. Seiji Yanai, Ishikawa
Prefectural University)
438 K. Fujiwara et al.

Table 6 Pinus thunbergii and Pinus densiflora plantation

The Kaga Coast (Japan) … 439

Table 6 (continued)
440 K. Fujiwara et al.

Eurya japonica, Neolitsea sericea, Ardisia japonica, and Hedera rhombea. At the
front of the back dunes and along the contact with the foredunes, these forests are
relatively short and may have Juniperus taxifolia var. lutchuensis as dominant,
along with other dune species such as Arabis stelleri var. japonica and Linaria
japonica. Less built-up areas have Miscanthus sinensis grassland species, such as
Pleioblastus chino var. viridis, M. sinensis, Iris ensata var. spontanea,
Hemerocallis flava var. littorea, Campanula punctata and Thalictrum minus var.
hypoleucum. The most stable areas have four-layer forests, with evergreen
broad-leaved forest species (* mark in Table 6) and secondary-forest species.
Sometimes P. thunbergii covers more than 30% in the sub-canopy. These forests
have a higher mean number of species, perhaps 44 species. On the other hand, areas
of undeveloped soil support only short pine forests, and mosses and lichens
dominate on the forest floor; the mean number of species is around 24.
3) Crab Communities and Biology
Although Akate-gani is highly adapted to land life, it must live temporarily in the
sea during its growth process. Akate-gani is omnivorous. It can climb trees to eat
fresh leaves, such as of Aucuba japonica var. borealis (Fig. 7), and can eat the
dropped leaves of Machilus thunbergii, etc. [18]. It can also use tree trunk flow to
get down to the seacoast [19]. Kashima Shintō shrine forest thus seems somewhat
like a terrestrial mangrove forest.

4 Current Protection Activities and Results

1. Collaboration with traditional and European technology for stability of sand

dune
For sand-defense plantings, the Japanese combined traditional technology with
European (such as French) technology to make the planting succeed.
After the Meiji Restoration, the government carried out a big project called “the
coastal sand defense project of beach national forest” from 1911. This project took
13 years, from 1911 to 1924, and drastically solved the sand problem (cf Fig. 4).
Following traditional methods (see above), they made 10 meter-high foredunes and
back dunes behind. On the foredunes they planted coastal plants such as Vitex
rotundifolia and sea bells (Calystegia soldanella), stabilized by bamboo mats. On
the Kaga Coast, they utilized both shrubs and herbs. These survived in each habitat,
and even endangered species came into these habitats (see next section).
2. Pine and sand-dune herb plantations, and natural forests made large ecosystems
Black pine (Pinus thunbergii) plantations and coastal plant communities made
by the project have become “nature created by humans” reflecting the character-
istics of this region. This is the result of interaction between the human approach,
which considers natural laws, and natural activity such as succession.
The Kaga Coast (Japan) … 441

Fig. 8 Plantation activity on the Kaga coast, with current view of pine plantation on the back
dune

The area of black-pine plantation did not only come into existence as a defense
against sand and wind. It has also become a beautiful forest with an excellent
landscape that reflects the characteristics of this area. This forest has been created
over about one century by both human engineering and plant successional pro-
cesses. Today the forest is also a place for recreation of children and adults (Fig. 8).
The tree overstorey of the black-pine area is black pine, but depending on the
location, red pines (P. densiflora) were mixed with black pines. There are also a few
individuals of exotic Pinus pinaster (P. mariana), which remain from the early
experimental planting. Adjacent to black pines there are plantations of Robinia
pseudoacacia and Alnus sieboldiana, which have the same origin.
Euonymo-Pittosporetum tobirae is the natural windbreak shrub community on the
coast of Ishikawa prefecture. Euonymus japonica shrubs were planted by the
Daishōji domain. Recently some black pines have been damaged by pine-wood
nematodes, where Machilus thunbergii occurred and succession to
Polysticho-Perseetum thunbergii has been in progress. Citizen planting activities
based on the potential natural vegetation have pushed this succession forward.
Coastal plants other than Vitex rotundifolia and Elaeagnus umbellata, originally
planted by humans, have also grown up on the foredune. Today the community is
evaluated as the richest habitat for rare species, such as Viola senamiensis, V.
mandshurica var. crassa, and Iris ensata. The Kaga community of Viola
senamiensis in particular is the largest such colony in the world (Fig. 5). The plants
in this community take turns blooming over the whole period from March to
October. Each plant species is integrated into the community by succession. In the
sandy-beach shrub community, there are some other communities, such as the
Linaria-Vitex community. Of herbaceous communities, about 20 are differentiated
(Kaga City 2002). The planted herbaceous species have developed their own
communities and occur on appropriate habitats (Fig. 9) [9].
442 K. Fujiwara et al.

Fig. 9 Restored Sand-Dune Vegetation on the Kaga Coast

It is confirmed that there are 17 species of larger animals, such as serow. There
are also over 290 bird species that come to or live all year on the Kaga Coast. Wild
ducks and wild geese migrate seasonally to Katano-Kamoike and its Ramsar-listed
surrounding wetlands. It is also confirmed that Nipponia nippon, the Japanese
national bird, fly along the Kaga Coast.
There are also about 450 insect species recorded just in the coastal plant com-
munities. There are 18 “important species,” which are listed in the Red Data Book
of Japan or of Ishikawa prefecture. Insects on the Kaga Coast mainly live in three
areas: the coastal plant community, the Kashima-no-mori forest, and the wetlands
around Katano-Kamoike. It is ascertained that over 900 species of insects are living
in the whole Kaga-Coast area. The most important species are Agrotis ripae and
Cicindela laetescripta, which live in only a few places. The Kaga Coast is also a
stable habitat of Bembix niponica (a moth of the family Noctuidae) and
Herpetogramma albipennis (a moth in the family Crambidae), both of which were
found only recently. Spiders Araneida and many Lycosa ishikariana also live in the
coastal plant communities [8].
The Kashima-no-mori, a National Natural Monument of Japan, is an evergreen
broad-leaved forest dominated by Machilus thunbergii. The forest is the largest
habitat of the Akate-gani crabs (Chiromantes haematocheir). Akate-gani has a
unique mode of living—it lives in the forest and hibernates in burrows. The scene
of Akate-gani laying eggs shows a good example of a food chain.
The Kaga Coast (Japan) … 443

Why are there so many animals and plants living on the Kaga Coast? Because
there is a large area of coastal dune vegetation, and this forest and foredune support
large, diverse ecosystems.
3. The Kaga Coast was the last remaining undeveloped coastal area in Japan
The dunes along the Kaga Coast represent a rare case in coastal Japan, since the
Kaga dunes had not been diverted to farmland, factory or residential area,
amusement park, sports facility, or land for public use such as park and school.
Such diversion is generally seen in sand-defense or windbreak forests in Japan. In
addition, even if such forest remains, the landscape often gets destroyed, as by
highway construction, installation of tetrapods or other breakwaters, or develop-
ment of ports or other waterfront areas. Destruction of the sand-defense forest is
now accepted by society because sand disasters now occur less frequently, since
less sand is now delivered by rivers where dams are constructed [16]. For these
reasons, from a national viewpoint, the value of the “Kaga Coast” as a cultural
landscape is increasing year after year.

References

1. Anonymous (1794) A summary memorandum of local office

2. Anonymous (1816) The memorandum of official maps of Kaga, Ecchu, Noto
3. Braun-Blanquet J (1964) Pflanzensoziologie, Grundzüge der Vegetationskunde, 3rd edn
4. Editorial Committee for Flora of Ishikawa Prefecture (ed) (1997) Flora of Ishikawa prefecture.
Ishikawa Prefecture, Kanazawa
5. Editorial Committee for The History of Forestry in Ishikawa Prefecture (ed) (1997) The
history of forestry in Ishikawa prefecture. Forest association of Ishikawa prefecture
6. Editorial Office of Urban Kubota (1992) Feature Article: Hills and plains of Hokuriku, Urban
Kubota, Issue 31, Kubota Corp., Tokyo
7. Enuma Country Office of Ishikawa Prefecture (1925) The topography of Enuma country
8. Executive Committee of Present State Survey of Hashitate Hills of Kaga City (ed) (2007) The
present-state survey report on the preparation project of Hashitate Natural Park, Kaga City
9. Institute for Global Environmental Strategies-Japanese Center for International Studies in
Ecology (ed) The vegetation of Kaga City, Kaga City, Kaga (in Japanese with English
summery)
10. Ishikawa Prefecture (1925) The history of Ishikawa prefecture, 3rd volume, Kanazawa.
Ishikawa Prefecture, Kanazawa
11. Kanazawa Forestry Office (1952) The outline of the coastal sand defense project of beech
national (revised edition), Kanazawa Forestry Office
12. Kano T, year unknown, Sozan Iko
13. Konno Y (ed.) (1993) The geological History of Ishikawa prefecture, Ishikawa Prefecture and
Hokuriku Research Institute of Geology, Kanazawa
14. Kozuka H (1844) The topography of Kaga and Enuma
15. Mueller-Dombois D, Ellenberg H (1974) Aims and methods of vegetation ecology. Wiley,
p 547
16. Murai H et al (ed) Coastal forests in Japan, Soft Science Corp., Tokyo
17. Tsukaya S (1803) Bakkei Kibun
444 K. Fujiwara et al.

18. Yamagishi A, Yanai S (2012) Defoliation by terrestrial crabs in a laurel forest on the coast of
the Sea of Japan. ESJ59/EAFES5 Abstract. P3–305J (in Japanese)
19. Yanai S (2017) The ecology of Aka-te-gani (Chiromantes haematocheir) inhabiting coastal
forests. Inochi-Shiki-Saiji (Life, seasonal color) D & U, vol 24 (in Japanese). http://www.
sizenken.jp/info/inochi_sikisaiji/vol_24.pdf
20. https://www.pref.ishikawa.lg.jp/sizen/reddata/rdb_2010/data/documents/hamabouhuu521.pdf
21. https://www.pref.ishikawa.lg.jp/sizen/reddata/rdb_2010/data/documents/isosumire306.pdf
Measurement and Conservation
of Species and Habitat Biodiversity
at Different Scales on Small
Mediterranean Islands

Gianniantonio Domina, Anna Maria Mannino,

Francesco M. Raimondo, and Patrizia Campisi

Abstract In this paper we discuss current methodologies for measuring biodi-

versity, analyse pros and cons at different scales, and finally propose distribution
mapping as the most suitable method for measuring floristic biodiversity on small
Mediterranean islands. These territories often host endangered flora and habitats,
and have peculiarities that distinguish them from the mainland and the larger
islands.

Keywords Remote sensing
Distribution mapping
Floristic lists
Population

size

1 Introduction

Small islands are geographically and ecologically well-defined areas in which bio-
logical processes are easier to schematize than on the mainland [15]. Size is the best
geographical feature for characterizing the different Mediterranean islands biologi-
cally. Larger islands show biological traits similar to other coastal areas of the
Mediterranean [7, 8]. Other features, such as distance from the mainland and shape,
have been shown to be less influential for the study of biological colonization [8].

G. Domina (&)
Department SAAF, University of Palermo, viale delle Scienze Bldg. 4, 90128 Palermo, Italy
e-mail: [email protected]
A. M. Mannino
F. M. Raimondo
P. Campisi
Department STEBICEF, University of Palermo, via Archirafi 38, 90123 Palermo, Italy
e-mail: [email protected]
F. M. Raimondo
e-mail: [email protected]
P. Campisi
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 445
F. Pedrotti and E. O. Box (eds.), Tools for Landscape-Scale Geobotany and
Conservation, Geobotany Studies, https://doi.org/10.1007/978-3-030-74950-7_21
446 G. Domina et al.

The size limit for distinguishing small versus large Mediterranean islands may be
about 80 km2, approximately the size of Formentera, Pantelleria, and Santorini.
According to Gaston [12], biodiversity can be considered from the point of view
of ecology, of organisms and of genetics. It is therefore clear that several methods
exist to measure biodiversity but all agree on measuring the number of elements and
their variability. The range of spatial, temporal and taxonomic scales used increases
further the variables to be measured. Many existing methods are well designed and
informative. But, very often they are inadequate for purposes beyond those for
which they were specifically designed. Species diversity indexes, for example, are
based on counting species, and do not take into account their total or relative
abundance distribution. Knowledge on distribution of species and their population
trends are essential for reliable plans of conservation and management of the
resources. Knowledge on trends in biodiversity loss is hindered by the absence of
reliable basic data for most groups of organisms as well as habitats. Plants are
primary producers and key structural elements for most ecosystems and islands are
among the best floristically known territories in the Mediterranean. Greuter [14]
reports that in Mediterranean islands the percentage of records of taxonomic
doubtful taxa per total number of records is about half of that recorded in mainland,
whereas the percentage of records of doubtful occurrence per total number of
records is about doubled. These data can be explained by a higher attention of
researcher for taxa occurring in the islands and with a greatest difficulty in ascer-
taining historical records. From these considerations comes the need to choose a
suitable method for measuring biodiversity in small islands. In this contribution, we
review the current methods for the measurement of biodiversity and propose dis-
tribution mapping as the most suitable method for small Mediterranean islands.

2 Materials and Methods

A preliminary survey of current methods used to assess biodiversity was made from
available literature. The approaches used concern: habitat measurement, population
size measurement, taxon lists, and distribution mapping. Pros and cons of each
method, at different scales, are summarized in Table 1.

3 Discussion

Habitat Measurement. Advances in remote sensing and GIS software have greatly
aided large-scale habitat measurement. For example, the extent of forest fires in
Southeast Asia and losses of primary forest have been monitored with satellite
images [18]. Nowadays, it is possible to obtain satellite images with ground res-
olution of 40 cm. This resolution can be sufficient for trees or habitats well dif-
ferentiated from their surroundings, but it is not yet adequate for many other
Measurement and Conservation of Species and Habitat Biodiversity … 447

Table 1 Methods to measure biodiversity, their pros and cons and their scale of applicability
Method Pros Cons Most suitable
for
Remote – low cost and high – mapped phenomena must be Broad scale
habitat speed recognizable on available
measurement – automatic recognition images
– there is no need to visit
all sampling sites
– satellite data available
for the whole earth
– high interest in current
scientific literature
Measures of – highly informative – need to visit all sampling sites Fine scale
population – adopted as – long time requirement
size international standard – few historical data available
Floristic lists – low cost – no reliable information on the Both broad
– large historical phenomenon size and fine scale
data-sets available – little interest in current
scientific literature
Distribution – gives information – geo-referenced data are needed Both broad
mapping about the phenomenon – need to visit all sampling sites and fine scale
size
– large data-sets
available

purposes, such as monitoring many habitats and plants that occur on islands.
Typical examples are Mediterranean temporary ponds, a highly threatened
ecosystem [24], and ephemeral therophytic grasslands [6], which occur on surfaces
so small that they cannot be recognized on satellite or aerial photos (Fig. 1).
Population size for threatened taxa, together with analysis of population trends
[11], are the bases for assessments according to IUCN methodologies (IUCN 2012).
Unfortunately, historical data on population size are not available for most species,
and new measures are expensive and not practical in many cases. The problem on

Fig. 1 Temporary ponds in the island of Lampedusa from a satellite image (a) and a live photo (b)
448 G. Domina et al.

Fig. 2 Habitat of Silene hicesiae Brullo & Signor. in the island of Alicudi. The instability of steep
rocky slopes makes this area inaccessible for measurements

islands is accentuated by the occurrence of sea cliffs and other areas of difficult
access that make impracticable the counting of single individuals or the measure-
ment of surfaces occupied by populations (Fig. 2).
Floristic Lists. Information on the presence of higher plants is available for many
Mediterranean islands and often from different periods. This is because islands have
always attracted naturalists who have historically produced reliable records that can
be found dispersed within national and supranational floras. For example, several
references to the flora of the East Aegean Islands can be found in Boissier’s Flora
Orientalis (Boissier 1867–1884) or in targeted contributions (e.g. [5]. Nowadays,
visits by individual researchers have been joined by formal surveys undertaken by
governments and NGOs. The PIM (Petites Iles de Méditerranée) initiative was set
up in 2005 by the Conservatoire du Littoral in Aix-en-Provence (France), in order
to promote the conservation of biodiversity on small Mediterranean islands. It is a
virtuous example of how it is possible to carry on naturalistic exploration of the
small Mediterranean islands [17], involving local administrations and recording all
the results into a large virtual library (https://www.initiative-pim.org) (Fig. 3).
In addition, on islands the lack of records for a species is more easily correlated
to its disappearance from that territory than it is on the mainland. Particular
attention has to be paid to measurement of rarity and extinction risk. For example,
Arbutus unedo L. and Erica multiflora L. ssp. multiflora occupied large areas on
Lampedusa until the XIX Century, before the systematic deforestation of the island.
At present, only few, isolated individuals are present [20]. Most extinctions, after a
Measurement and Conservation of Species and Habitat Biodiversity … 449

Fig. 3 An example of freely distributes dot mapping for Elatine gussonei (Sommier) Brullo et al.,
an endangered species living on Mediterranean temporary ponds [19], Fig. 7), in the Maltese
islands

first peak due to some specific phenomenon, have long-term after-effects, whereby
the species may persist at low numbers with a negligible chance of recovery and a
severely diminished role in the ecosystem. In addition, actual presence is no longer
comparable to that of an earlier time in terms of genetic resources [4].
Distribution mapping of taxa is probably the most commonly used surrogate for
overall biodiversity, at both local and broader scales. The first well known
European example of an atlas covering an entire regional flora was carried out for
the British Isles [21]. Several projects have been undertaken in the last years to map
flora in Europe and in the Mediterranean area, both at large and regional scale.
Coastal biodiversity of Sicily was assessed by dividing the territory into elementary
areas. These cells were used to map the presence of the biogeographically most
important taxa (fauna and vascular, bryophytic and algal flora), in order to evaluate
punctual and general levels of biodiversity conservation diachronically [10].
The WIKIPLANTBASE system [2] was set up in four Italian administrative regions
(Toscana, Sardinia. Liguria, and Sicily) and allows free entry of historical and
modern records as dots on a base map, based on field observations, bibliographic
450 G. Domina et al.

references or herbarium specimens. This system has the great advantage of being
open to collaborative work greatly enriched by amateur contributions [22].
The species level can be an accepted standard for Mediterranean island moni-
toring and conservation plans. This is because the concept of a species is well
understood also by the public, stakeholders and policy makers. Varieties or sub-
species, although basic to understanding ongoing speciation, risk creating confusion
and giving an unclear idea of the phenomena, diverting important resources from
the entities that need conservation most. It must be kept in mind, though, that there
are several species, such as Pancratium maritimum L., that show great interpopu-
lation variability, so they deserve protection at as many sites as possible [13].

4 Conclusion

From the above discussion it is intuitive that remote habitat measurement is the best
solution for broad-scale phenomena that can be recognized on photos. Distribution
mapping of taxa is probably the most commonly used surrogate for overall bio-
diversity, at both local and broader scales. Freely distributed dot mapping can be
used when highly detailed distribution data are available. This allows identification
of the geographical position of a taxon on a cartographic support with the highest
possible precision (e.g. [1]. For further statistical elaboration with commensurable
units of investigation free-dot conversion in a fixed matrix is advisable [13].

References

1. Barina Z (2017) Distribution atlas of vascular plants in Albania. Hungarian Natural History
Museum, Budapest
2. Bedini G, Pierini B, Roma-Marzio F, Caparelli KF, Bonari G, Dolci D, Gestri G,
D’Antraccoli M, Peruzzi L (2016) Wikiplantbase #Toscana, breaking the dormancy of
floristic data. Pl. Biosyst. 150(3):601–610
3. Boissier PE (1867–1884) Flora Orientalis, 1–5. H. Georg, Basileae
4. Boivin NL, Zeder MA, Fuller DQ, Crowther A, Larson G, Erlandson JM, Denham T,
Petraglia MD (2016) Ecological consequences of human niche construction: examining
long-term anthropogenic shaping of global species distributions. PNAS 113(23):6388–6396
5. Candargy CA (1889) Flore de l’ile de Lesbos. Plantes sauvages et cultivées, Diggelmann,
Zürich
6. Čarni A, Matevski V, Šilc U, Ćušterevska R (2014) Early spring ephemeral therophytic
non-nitrophilous grasslands as a habitat of various species of Romulea in the southern
Balkans. Acta Bot Croat 73(1):107–129
7. Celesti-Grapow L, Bassi L, Brundu G, Camarda I, Carli E, D’Auria G, Del Guacchio E,
Domina G, Ferretti G, Foggi B, Lazzaro L, Mazzola P, Peccenini S, Pretto F, Stinca A,
Blasi C (2016) Plant invasions on small Mediterranean islands: an overview. Pl Biosyst 150
(5):1119–1133
8. Domina G, Mazzola P (2011) Sulla presenza di specie aliene nella flora vascolare del
Mediterraneo. Biogeographia 30:269–276
Measurement and Conservation of Species and Habitat Biodiversity … 451

9. Domina G, Venturella G, Gargano ML (2018a) Synthetic cartography for mapping

biodiversity in the Mediterranean region: Sicily as a case study. PhytoKeys 109:77–92
10. Domina G, Campisi P, Mannino AM, Sparacio I, Raimondo FM (2018b) Environmental
quality assessment of the Sicilian coast using a multi-disciplinary approach. Acta Zool Bulg
Suppl 11:11–18
11. Ercole S, Fenu G, Giacanelli V, Pinna MS, Abeli T, Aleffi M, Bartolucci F, Cogoni D,
Conti F, Croce A, Domina G, Foggi B, Forte T, Gargano D, Gennai M, Montagnani C,
Oriolo G, Orsenigo S, Ravera S, Rossi G, Santangelo A, Siniscalco C, Stinca A, Sulis E,
Troia A, Vena M, Genovesi P, Bacchetta G (2017) The species-specific monitoring protocols
for plant species of Community interest in Italy. Pl Sociol 2:77–83
12. Gaston KJ (2010) Biodivesity. In: Sodhi NS, Ehrlich PR (eds) Conservation Biology for All.
Chapter 2, Oxford University Press, Oxford, pp 27–44
13. Giovino G, Domina G, Bazan G, Campisi P, Scibetta S (2015) Taxonomy and conservation of
Pancratium maritimum (Amaryllidaceae) and relatives in the Central Mediterranean. Acta Bot
Gallica BotLett 162:289–299
14. Greuter W (1995) Origin and peculiarities of Mediterranean island floras. Ecol Medit 21:1–10
15. Greuter W (2001) Diversity of Mediterranean island floras. Bocconea 13:55–64
16. IUCN (International Union for Conservation of Nature) (2012) IUCN Red List Categories and
Criteria. Version 3.2. Second Edition. IUCN Species Survival Commission. IUCN, Gland,
Switzerland. https://jr.iucnredlist.org/documents/redlist_cats_crit_en.pdf. Accessed 08 Aug
2018
17. Médail F, Charrier L, Charrier M, Kashta L (2013) Vegetation map of Sazani Island
(Albania). Conservatoire du littoral, Aix-en-Provence
18. Miettinen J, Stibig H-J, Achard F (2014) Remote sensing of forest degradation in Southeast
Asia - aiming for a regional view through 5–30 m satellite data. Global Ecol Conserv 2:24–36
19. Mifsud S (2006) A comparative study between Elatine gussonei (from Malta) and Elatine
macropoda (from Majorca), Aquatic Gardeners. https://www.aquatic-gardeners.org/Articles/
elatine/Comparitive_study.pdf. Accessed 16 Oct 2018
20. Pasta S (2001) Contributi alla conoscenza botanica delle isole minori circumsiciliane.
I. Sintesi aggiornata delle conoscenze botaniche sull’isola di Lampedusa finalizzata alla
conservazione delle sue emergenze floristico-vegetazionali. Naturalista Sicil., s. 4, 25
(Suppl.):19–70
21. Perring F, Walters SM (1962) Atlas of the British Flora. Botanical Society of the British Isles,
London
22. Peruzzi L, Bagella S, Filigheddu S, Pierini B, Sini M, Roma-Marzio F, Caparelli KF,
Bonari G, Gestri G, Dolci D, Consagra A, Sassu P, Caria MC, Rivieccio G, Marrosu M,
D’Antraccoli M, Pacifico G, Piu V, Bedini G (2017) The Wikiplantbase project: the role of
amateur botanists in building up large online floristic databases. Fl. Medit. 27:117–129
23. Whittaker RH (1972) Evolution and measurement of species diversity. Taxon 21:213–251
24. Zaccharias I, Zamparas M (2010) Mediterranean temporary ponds. a disappearing ecosystem.
Biodiv Conserv 19(14):3827–3834