Ecology Letters, (2013) 16: 1424–1435 doi: 10.1111/ele.

12189

IDEA AND
PERSPECTIVE Predicting species distributions for conservation decisions

Abstract
Antoine Guisan,1,2,3,4* Reid Species distribution models (SDMs) are increasingly proposed to support conservation decision making.
Tingley,5 John B. Baumgartner,5 However, evidence of SDMs supporting solutions for on-ground conservation problems is still scarce in
Ilona Naujokaitis-Lewis,6 Patricia the scientific literature. Here, we show that successful examples exist but are still largely hidden in the grey
R. Sutcliffe,3 Ayesha I. T. Tulloch,3 literature, and thus less accessible for analysis and learning. Furthermore, the decision framework within
Tracey J. Regan,5 Lluis Brotons,7,8 which SDMs are used is rarely made explicit. Using case studies from biological invasions, identification of
Eve McDonald-Madden,3,4 Chrystal critical habitats, reserve selection and translocation of endangered species, we propose that SDMs may be
Mantyka-Pringle,4,9 Tara G.
tailored to suit a range of decision-making contexts when used within a structured and transparent deci-
Martin,3,4 Jonathan R. Rhodes,9
sion-making process. To construct appropriate SDMs to more effectively guide conservation actions, mod-
ellers need to better understand the decision process, and decision makers need to provide feedback to
Ramona Maggini,3 Samantha A.
modellers regarding the actual use of SDMs to support conservation decisions. This could be facilitated by
Setterfield,10 Jane Elith,11 Mark W.
individuals or institutions playing the role of ‘translators’ between modellers and decision makers. We
Schwartz,12 Brendan A. Wintle,5
encourage species distribution modellers to get involved in real decision-making processes that will benefit
Olivier Broennimann,1 Mike
from their technical input; this strategy has the potential to better bridge theory and practice, and contrib-
Austin,13 Simon Ferrier,13 Michael
ute to improve both scientific knowledge and conservation outcomes.
R. Kearney,14 Hugh P.
Possingham3,15 and Yvonne M. Keywords
Buckley3,16 Biological invasions, conservation planning, critical habitats, environmental suitability, reserve selection,
species distribution model, structured decision making, translocation.

Ecology Letters (2013) 16: 1424–1435

or that may be colonised in the future following climate change

SETTING THE SCENE: SPECIES DISTRIBUTION MODELS FOR
(Hoegh-Guldberg et al. 2008) or biological invasions (Thuiller et al.
CONSERVATION APPLICATIONS
2005; Baxter & Possingham 2011; Giljohann et al. 2011). However,
Species ranges are shifting, contracting, expanding and fragmenting this information is important for making robust conservation man-
in response to global environmental change (Chen et al. 2011). The agement decisions and can be provided by predictions of species
emergence of global-scale bioinformatic databases has provided new occurrences derived from environmental suitability models that
opportunities to analyse species occurrence data in support of con- combine biological records with spatial environmental data.
servation efforts (Jetz et al. 2012) and has paved the way toward Species distribution models (SDMs; also commonly referred to as
more systematic and evidence-based conservation approaches (Mar- ecological niche models, ENMs, amongst other names; see Appen-
gules & Pressey 2000; Sutherland et al. 2004). However, records of dix S1) are currently the main tools used to derive spatially explicit
observed species occurrence typically provide information on only a predictions of environmental suitability for species (Guisan & Thuil-
subset of sites occupied by a species (Rondinini et al. 2006). They ler 2005; Elith & Leathwick 2009; Franklin 2010; Peterson et al.
do not provide information on sites that have not been surveyed, 2011). They typically achieve this through identification of statistical

1 10
Department of Ecology and Evolution, University of Lausanne, 1015, Lau- Research Institute for Environment and Livelihoods, Charles Darwin Univer-
sanne, Switzerland sity, Darwin, NT, 0909, Australia
2 11
Institute of Earth Surface Dynamics, University of Lausanne, 1015, Lausanne, School of Botany, The University of Melbourne, Parkville, Vic, 3010, Australia
12
Switzerland John Muir Institute of the Environment, University of California, Davis,
3
ARC Centre of Excellence for Environmental Decisions (CEED), School of Biologi- 95616, USA
13
cal Sciences, The University of Queensland, St Lucia, Brisbane, Qld, 4072, Australia CSIRO Ecosystem Sciences, GPO Box 1700, Canberra, ACT 2601, Australia
4 14
CSIRO Ecosystem Sciences, Ecosciences Precinct, Dutton Park, Brisbane, Qld, Department of Zoology, The University of Melbourne, Parkville, Vic, 3010,
4102, Australia Australia
5 15
ARC Centre of Excellence for Environmental Decisions (CEED), School of Imperial College London, Department of Life Sciences, Silwood Park, Ascot
Botany, The University of Melbourne, Parkville, Vic, 3010, Australia SL5 7PY, Berkshire, England, UK
6 16
Ecology & Evolutionary Biology, University of Toronto, Toronto, Canada Present address: Zoology Department, School of Natural Sciences, Trinity
7

gica i Aplicacions Forestals (CREAF), Bellaterra, Spain
Centre de Recerca Ecolo College, Dublin 2, Ireland
8

gic Forestal de Catalunya (CTFC - CEMFOR), Solsona, Spain
Centre Tecnolo *Correspondence: E-mail: [email protected]
9
ARC Centre of Excellence for Environmental Decisions (CEED), School of
Geography, Planning and Environmental Management, The University of
Queensland, St Lucia, Brisbane, Qld, 4072, Australia

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
Idea and Perspective SDMs for conservation decisions 1425

relationships between species observations and environmental de- clearly lacks the perspective of practitioners and decision makers on
scriptors, although more mechanistic modelling approaches, and how SDMs can contribute to solving environmental problems,
approaches involving expert opinion, also exist (Appendix S1). despite SDM construction often being justified based on their
SDMs have the potential to play a critical role in supporting spatial potential utility for decision making. As a result, there are a wide
conservation decision making (Margules & Pressey 2000; Addison variety of tools published, but little guidance on how SDMs – and
et al. 2013; Appendix S2), but their applicability and relative utility other models (Addison et al. 2013) – could be used to support deci-
across the breadth of conservation contexts remains unclear, as sion making in relation to clear conservation objectives (Possingham
does the extent of their adoption in aid of conservation decision et al. 2001). More practice-oriented assessments of the use of mod-
making. els to support conservation are urgently needed.
The last decade has seen a surge in the development of SDMs Here, we investigate instances outside the peer-review literature
(Fig. 1a, Appendix S3). However, despite large numbers of SDM- where SDMs have been used to guide conservation decisions, how
based studies published in the peer-reviewed literature, and wide- they were constructed when used, and how they could be used more
spread claims of applicability to conservation problems (Guisan & effectively in the future. We do not propose a review of SDMs, or
Thuiller 2005; Rodriguez et al. 2007; Cayuela et al. 2009; Elith & their use in conservation, nor do we undertake an exhaustive quanti-
Leathwick 2009; Franklin 2010; Peterson et al. 2011), evidence of tative assessment of the grey literature, which is difficult to access in
the practical utility of these models in real-world conservation man- many countries. Rather, based on chosen examples in different coun-
agement remains surprisingly sparse. An indicative assessment of tries (including developed and developing ones), we emphasise the
keywords in ISI suggests that < 1% of published papers using importance of clearly articulating the decision context to determine
SDMs are specifically targeted at conservation decisions (Fig. 1b, where and how SDMs may be useful. We examine how closer con-
Appendix S3). A recent review of SDMs used in tropical regions sideration of the decision-making context and better collaboration
(Cayuela et al. 2009) similarly concluded that < 5% of studies with decision makers may encourage the development and use of
addressed conservation prioritisation. Furthermore, in the few pub- SDMs for guiding decisions. Our primary focus is on statistical
lished applications of SDMs to conservation decision making (e.g. SDMs, as they are the most frequently and readily applied, although
Brown et al. 2000; Sober on et al. 2001; Ferrier et al. 2002; Leathwick other approaches, such as mechanistic SDMs (Kearney & Porter
et al. 2008), the importance of their contribution to the decision- 2009), may also provide input for conservation decision making.
making process and implementation of actions is often unclear (but
see Pheloung et al. 1999). The bulk of the peer-reviewed literature
FROM PROBLEMS TO DECISIONS: HOW CAN SDM CONTRIBUTE
TO DECISION MAKING?
337
The potential of SDMs to guide conservation actions is best
5

2546 assessed by first considering the full decision-making process, a step

(a) rarely taken. Structured decision making (Gregory et al. 2012; Fig. 2)
30

256
provides a rigorous framework for this process and is increasingly
4

proposed to address environmental problems (Wintle et al. 2011;

20

572 192 Addison et al. 2013). This approach is usually sequential (Possing-
ham et al. 2001), with potential roles for SDMs at most stages of
of publications

10

182
the decision process (Fig. 2, Table 1), as outlined below.
3

10
140
0

1995 2000 2005 2010 103 Identifying a problem

2

76
The need to make a conservation decision arises from the identifi-
(b) cation of a conservation problem (Fig. 2a). SDMs could play a role
43 by highlighting likely shifts of suitable habitat for a species due to
33
climate change (Araujo et al. 2011), or by identifying areas likely to
1

22
9 11
13 be invaded by a pest species (Thuiller et al. 2005; Araujo et al.
5 5
2 3 5 2011), and therefore allow the identification of potential conflict
5 9 10 10 10 15 18 areas if species may not be able to migrate across human-modified
0

1 1 1 1 2 2 2 3 5
landscapes, or if the native communities at threat of being invaded
1995 2000 2005 2010 shelter threatened species (e.g. Vicente et al. 2011).
Year
Defining the objectives
Figure 1 Cumulative trends over the last 20 years extracted from the Web of
Science (WoS), showing the increasing number of peer-reviewed papers related Once a problem is identified, the definition of conservation objec-
to SDMs (keyword search). Curves are drawn as proportions ( ) of the tives is usually the realm of decision makers and stakeholders. How-
cumulative number of papers published in the WoS category ‘Ecology’. The ever, scientific input may be used to ensure objectives are realistic,
cumulative number of papers for each year is indicated on the curves. (a) All
given the current, or projected, state of the environment. SDMs
SDM papers. (b) Only SDM papers in the four important conservation domains
(biological invasions, critical habitat, reserve selection, translocation) discussed in may be used as a frame of reference for setting objectives retro-
the paper, without (solid line) or with (dashed line) the keyword ‘decision’. For spectively from the identified problem, or interactively by refining
choice of keywords see Appendix S3. conservation objectives within an adaptive framework (Runge et al.

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
1426 A. Guisan et al. Idea and Perspective

(a)

(b)
(d)

(c)

Figure 2 A structured decision-making process (Gregory et al. 2012) with indication of potential entry points for the use of SDMs. See main text and Table 1 for details.
The black arrows indicate where SDMs can contribute to steps in the decision-making process.

Table 1 Examples of ways to increase the utility of SDMs within four conservation domains and the structured decision analysis process (DAP). The first five rows corre-
spond to specific DAP steps, whereas the final three rows describe general issues requiring consideration.

Biological invasions Critical habitat Reserve selection Translocation

Problem A new invader is likely to Particular habitat patches drive Inappropriate habitat protection leads The rate of climate change may exceed
identification impact particular habitats. species’ extinction vulnerabilities. to higher extinction vulnerabilities. species’ capacity to respond.
Defining the Reduce harmful impacts by Provide adequate habitat protection Provide adequate habitat protection Increase persistence probabilities of
objectives prevention or mitigation for threatened species. for threatened species. climate vulnerable species.
of invasion.
Defining When and where to carry Strengthen protection, acquire new Acquire reserves, private landowner Translocate species, manage dispersal
possible out quarantine, surveillance, reserves, foster migration, incentives, restoration, reserve corridors, passive migration
actions eradication, containment or translocation. management. management.
local control.
Consequences Estimating the extent to Estimating extent of opportunity Estimating which subset of at risk Selecting subset of at risk taxa for
of actions which potential impacts may costs for other habitat uses, taxa may be conserved. action, risk of creating invasion
be prevented or mitigated estimation of extinction risk. problem.
through actions.
Trade-off Cost efficiency of surveillance Social and economic conflict over Social and economic conflict over Cost-benefit and potential conflicts
analysis and management vs. risk of land use. land use. of placing species in novel
adverse impacts. environments.
Decision that Predicting areas of potential Determining most favourable habitats. Model diversity at a landscape level Identify target locations for managed
can be occupancy to target to set priorities. relocation.
informed by surveillance and management.
SDM
How SDM Under-prediction may miss Distribution model error misidentifies Uncertain suitable environments may Spatial scale constraints limit the
uncertainty critical surveillance, over- optimal habitats leading to excess lead to suboptimal reserve selection. specificity of targeting locations.
influences prediction may waste opportunity costs or species
decisions management resources. extinction.
Key issues for Biotic interactions may play a Careful integration of population Project regional diversity hotspots Apply SDMs to assess future
integrating strong role in determining persistence processes into under global change models. distributions for species targeted for
science and environmental suitability in management decision. dispersal assistance.
management novel habitats.

2011). For example, initial objectives may be set based on low qual-
Defining possible alternative actions
ity data but through the course of subsequent conservation and
research actions, better quality data may inform an SDM and lead The definition of feasible actions (Fig. 2b) may be informed by
to changes in the initial objectives. It is essential that the outcomes SDMs. For example, when making decisions about where to trans-
of any subsequent action (see the following two points) be evalu- locate a threatened species (Chauvenet et al. 2012) or where to tar-
ated against the objectives (Chauvenet et al. 2012). get control of an invasive species (Baxter & Possingham 2011),

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
Idea and Perspective SDMs for conservation decisions 1427

SDMs may be used to identify candidate locations as alternative conservation domains, with differences in use intensity. We discuss
actions that may subsequently be evaluated in greater detail. Infor- four areas where SDMs have been used to guide management deci-
mation about the costs of management actions, logistical constraints sions: the use of climate-matching SDMs in some invasive species
(e.g. distance) or conflicting conservation priorities (e.g. various land risk assessments (Managing biological invasions), the use of SDMs
ownerships) for example will ultimately determine the feasibility of to guide the legal identification of critical habitats for threatened spe-
different actions, but the SDM provides a suite of options. cies (Identifying and protecting critical habitats), the use of SDMs in
regional conservation planning (Reserve selection) and the use of
SDMs for informing translocation of threatened or captive-bred
Evaluating the consequences of alternative actions
populations (Translocation) (Table 1, Fig. 3).
Species distribution models can be used to evaluate the implementa-
tion of alternative actions (Fig. 2c) in terms of predicting resultant
Managing biological invasions
changes to species’ distributions, or to the quality of habitat. For
example, use of SDMs has been proposed to assess alternative In some countries, SDMs are commonly used to guide decisions
reserve designs and their role in conserving biodiversity under cur- about invasive species management. For instance, Australia has
rent and possible future climates (Hannah et al. 2007). implemented advanced detection, prevention and impact mitigation
programmes that include SDMs. Pre-border weed risk assessment
encourages the use of SDMs to aid decisions about whether to
Assessing the trade-offs between benefits and costs of actions
allow the import of new plant species (Pheloung et al. 1999; see
This important step builds on the identified consequences of Defining possible actions, Fig. 2b). Post-border weed risk assessments
actions (Fig. 2). SDMs can be used to quantify benefits to be traded use maps of potential distributions, developed using SDMs, to assist
off against costs of actions, such as in prioritising competing wet- in the identification of potentially widespread, high impact, invaders
land bird management options ranging from adding artificial habitat and to apportion control costs among potentially affected regions.
features to controlling disease outbreaks and changing pond inunda- SDMs are systematically used to contribute to the classification of
tion regimes (Sebastian-Gonzalez et al. 2011), or in optimising vari- species as weeds of national significance (NTA 2007). At the regio-
ous control actions for invasive species across space (Giljohann nal scale, such an approach recently contributed to the official list-
et al. 2011). ing of gamba grass (Andropogon gayanus) as a weed in the Northern
Territory of Australia (NTA 2009; Fig. 3a). In Mexico, SDMs were
used to predict the potential impact of the invasive cactus moth
Assessing and dealing with uncertainty
(Cactoblastis cactorum) on native cacti (Opuntia spp) to facilitate plan-
All conservation decisions are made in the presence of some uncer- ning and mitigation of future impacts (Sober on et al. 2001).
tainty, and most involve the implicit or explicit specification of an
acceptable level of risk (Fig. 2d). Assessment of risk includes esti-
Identifying and protecting critical habitats
mation of the differential cost to biodiversity of errors associated
with under-protection vs. over-protection (Schwartz 2012). In par- Critical habitats are typically defined as habitats necessary for the
ticular, the type (Barry & Elith 2006) and magnitude (Carvalho et al. persistence, or long-term recovery, of threatened species (Greenwald
2011) of uncertainty that are acceptable need to be based on the et al. 2012), and their identification is required by law in some coun-
needs of decision makers, and incorporated into the definition of tries (e.g. Canada, USA, Australia). SDMs are one tool for differen-
the objectives (Richardson et al. 2009; Fig. 2a). SDMs enable the tiating habitat quality at a range-wide scale, and can be combined
quantification of some types of uncertainties in the spatial predic- with other sources of information, such as population dynamics, to
tions of environmental suitability (Barry & Elith 2006), and these define critical habitat (Heinrichs et al. 2010). In Canada, hybrid
can be explicitly incorporated in conservation prioritisation pro- SDM-population dynamics models were used to determine critical
cesses (Moilanen et al. 2006). However, some other types of uncer- habitat for the Ord’s kangaroo rat (Dipodomys ordii; Heinrichs et al.
tainties are not directly retrievable from SDMs (Appendix S1) but 2010). In Catalonia (Spain), SDMs were used to identify critical hab-
need to be recognised and where possible considered. When decid- itats for four threatened bird species to guide land-use decisions in
ing whether to invest in reducing uncertainty, it is useful to consider a farmland area affected by a large-scale irrigation plan. In the latter
whether the uncertainty is reducible (Barry & Elith 2006) and case, SDMs were first developed by scientists (Brotons et al. 2004),
whether a reduction in uncertainty might lead to decisions that yield explained to practitioners (CTFC 2008) and finally influenced policy
better management outcomes (Regan et al. 2005), a concept gener- and were considered in a legal decree in the framework of the Na-
ally known as value of information (Runge et al. 2011). tura 2000 network management plan (DMAH 2010; Fig. 3b; see
Appendix S4). In Australia, the Victorian State Government devel-
oped SDMs for use in regulating vegetation-clearing applications
EXAMPLES OF USING SDM FOR GUIDING CONSERVATION
(DEPI 2013).
DECISIONS

Despite the numerous potential conservation applications proposed

Reserve selection
for SDMs, examples where SDMs have explicitly guided decisions
relating to the management of natural resources are difficult to find The delineation and establishment of protected areas often forms
in the scientific literature. We searched the grey literature (partially the cornerstone on which conservation plans are built (Margules &
based on our own linkages with practitioners) and found various Pressey 2000). An early example of the use of SDMs in systematic
examples of the practical use of SDMs to guide decisions in different conservation planning involved the development of SDMs for over

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
1428 A. Guisan et al. Idea and Perspective

(a) (b)

(c) (d)

Figure 3 Four examples of maps used in conservation decision making based on SDMs. (a) Declaration of gamba grass (Andropogon gayanus, picture by Samantha
Setterfield) as a weed using the weed risk assessment process in the Northern Territory of Australia (NTA 2009). (b) Identifying critical habitats (red) for three
endangered bird species in Catalonia, Spain, as used in a legal decree (DMAH 2010) (picture of Tetrax tetra by Blake Matheson). (c) E-RMS tool windows and spatial
query result for an endangered frog (Philoria loveridgei), as used in the conservation planning project for northeast New South Wales forests (Brown et al. 2000). (d)
Identification of habitat use by the Bighorn sheep (Ovis canadensis sierra, picture by Lynette Schimming) in the Sierra Nevada, California, based on historical records
only (NPS Seki 2011); SDM were not used to plan current translocation efforts but to predict the future distribution of potential translocation sites (Johnson et al. 2007).

2300 species of plants and animals throughout the northeast for- sal, cross-validation, and independent field testing), and to commu-
ests of New South Wales, Australia (results first presented in a nicating this uncertainty to decision makers (e.g. through mapping
report in 1994, cited in Brown et al. 2000; Ferrier et al. 2002). This of confidence limits for predicted distributions). In another exam-
region was the focus of a long-running conflict between the needs ple in Madagascar, SDMs for large numbers of species in the main
of commercial forest harvesting and the protection of exceptionally biodiversity groups (mammals, birds, reptiles, amphibians, freshwa-
high biodiversity. The SDM outputs were integrated with data on ter fishes, invertebrates, plants) were developed by scientists and
other conservation and timber values in an environmental deci- managers, and used to define priority areas for conservation (Kre-
sion-support system by a team of negotiators representing all rele- men et al. 2008) using the Zonation software (Moilanen et al.
vant government agencies and non-government stakeholders (see 2009). These were then combined with other ‘priority areas’ using
example in Fig. 3c). The aim was to identify areas of high conser- the Marxan software (Watts et al. 2009) and put on the map of
vation value for exclusion from logging, thereby resulting in major ‘potential sites for conservation’. Following a legal decree (Arr^ete
additions to the regional network of protected areas (Ferrier et al. Interministe riel n18633/2008/MEFT/MEM, renewed in 2013), no
2002). This SDM application also provides an early demonstration mining and forestry activities can be permitted in these priority
of various approaches to evaluating and quantifying some sources areas for conservation as long as the decree remains in force
of uncertainty in predictions (e.g. through expert ecological apprai- (Appendix S5).

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
Idea and Perspective SDMs for conservation decisions 1429

Translocation Decision context

The active transport of species by humans has been proposed as a The example from the northeast forests of New South Wales
measure to mitigate the threats species face under present or future (Brown et al. 2000; Ferrier et al. 2002) provides a rare documented
conditions (Richardson et al. 2009; Chauvenet et al. 2012). SDMs can case where all necessary conditions for building SDMs in a conser-
potentially inform the translocation decision process at three key vation context were met. Foresight by planners in the state environ-
stages. First, SDMs can identify suitable habitat under current and mental agency and funding by both commonwealth and state
future climates to reveal whether habitat suitability is likely to decline governments, along with data availability and sufficient lead-time for
in regions currently occupied by the species (Fig. 2a), thereby sup- skilled staff to develop SDMs appropriate for the conservation
porting the decision of whether translocation is necessary (Hoegh- objectives, made the use of SDMs in the decision-making process
Guldberg et al. 2008; Thomas 2011). Second, if translocation is possible. The Madagascar case is another example where careful
deemed necessary, SDMs can identify potential recipient sites, which evaluation of the decision needs led to appropriate decisions for
may be climate refugia within the current range, or sites that are pro- building SDMs, in this case by: ensuring species-environment tem-
jected to become newly suitable (Chauvenet et al. 2012; McLane & poral matching, using models above some validation threshold only,
Aitken 2012; Fig. 2b). Third, SDMs can be used to identify which correcting for biogeographical overprediction and adding expert val-
local species may be at risk of impact from the introduction of a idation. In some cases, however, an SDM could be constructed for
translocated species through predicted overlapping distributions, in a species in the context of a conservation action to be taken, but
the same way as they are used to identify conflict areas between the desired outputs (e.g. spatial predictions, ecological response
native and invasive species (Vicente et al. 2011; Fig. 2c). An example curves) may not meet the criteria (e.g. spatial accuracy, level of cer-
of the identification of suitable translocation sites in present and/or tainty) necessary for its contribution to a final decision. Hence, early
future climates exist for the bighorn sheep (Ovis canadensis sierrae) in awareness of decision criteria increases the chance of developing
the Sierra Nevada (Johnson et al. 2007; NPS Seki 2011; Fig. 3d). An SDMs that are useful for decision makers. This requires a close
SDM was used to identify suitable sites for reintroductions and association between decision makers and SDM-developers from the
translocation by avoiding areas of overlap with existing grazing stock onset of SDM development (McAlpine et al. 2010). Collaboration
allotments and areas of high predator densities. between decision makers and SDM-developers also offers opportu-
These four groups of examples show that SDMs can be used to nities for evaluation of other sources of ecological knowledge and
guide different decision-making steps in different conservation con- data as a substitute for or complement to SDMs.
texts (Table 1, Figure 2). Yet, the bulk of SDMs currently remains
primarily developed for scientific purposes. However, as we show
Time
below, the way SDMs are built may vary depending on the require-
ments of the decision-making context, which are primarily influ- Many threatened species have restricted distributions and specific
enced by the conservation objectives and the decisions to be made habitat requirements, so decisions to protect critical habitat may
(often – but not necessarily – defined independently of the SDMs; need to be made with some urgency to avoid extinction (Martin &
e.g. select reserves to minimise biodiversity loss below some arbi- Maron 2012). This urgency often leads to protection of minimum
trary threshold). amounts of habitat based on occurrence data alone. For example,
the endangered Banff Springs snail (Physella johnsoni) is found in only
five thermal springs, all of which are designated as critical habitat
TOWARD A DECISION-MAKERS PERSPECTIVE: HOW CAN THE
for this species (Lepitzki & Pacas 2010). In such cases, allocating
DECISION-MAKING CONTEXT GUIDE SDM DEVELOPMENT?
time to collect more data and build accurate SDMs or more com-
Many methodological choices are made when building and using an plex spatially explicit population models may not necessarily
SDM (Guisan & Thuiller 2005; Elith & Leathwick 2009; Franklin improve predictions but may delay the action of protection. How-
2010; Peterson et al. 2011), often with very general, research-ori- ever, deciding to build a simple SDM, or to not build one at all,
ented objectives in mind, such as answering macro-ecological ques- may overlook some potentially critical habitats for the species
tions, predicting range shifts under climate change (Keith et al. (Heinrichs et al. 2010). There is thus a trade-off between allocating
2008; Carvalho et al. 2011; Fordham et al. 2012) or assessing the conservation resources to model construction or to immediate
potential spread of invasive species (Thuiller et al. 2005). The use of action with uncertain consequences (McDonald-Madden et al. 2008).
SDMs is conditional on the availability of suitable data, skilled staff, For situations where time is less critical, more sophisticated SDMs
modelling tools, funds and time. Many methodological factors, such might suggest new sites where a threatened species could be found,
as error in locational or temporal accuracy, or biased data, also or areas that could be recolonised (Fig. 2b), as demonstrated in the
potentially affect SDMs and their predictions (Kadmon et al. 2003; cases of the Sierra Nevada bighorn sheep (Ovis canadensis sierra; NPS
Cayuela et al. 2009; Appendix S1). Using an inappropriate modelling Seki 2011; see above) and the whitebark pine (Pinus albicaulis) in
method or disregarding influential methodological factors can have western North America (McLane & Aitken 2012).
consequences for the intended use of an SDM. The utility of an
SDM for decision makers is therefore highly context sensitive.
Population dynamics
Below, we present examples that show why choices of various
options for building/using an SDM may require more careful atten- Modelled probabilities of occurrence from SDMs may not always
tion in a decision-making context where modelling methods should correlate with the population processes necessary for species’ persis-
be determined by the nature of the conservation problem at hand tence (Fordham et al. 2012). In such cases, it may be necessary to
and the decision to be made (Table 1). combine process-models such as population viability analyses with

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
1430 A. Guisan et al. Idea and Perspective

SDMs to better evaluate the effects of management actions on choice of threshold are continuously updated until decision-makers
long-term species’ persistence (Keith et al. 2008; Wintle et al. 2011; are satisfied with the balance of both types of errors.
Fordham et al. 2012). Such an approach was recently used to assess
critical habitats for Ord’s kangaroo rat in Alberta, Canada (Hein-
Uncertainty
richs et al. 2010) and revealed that 39% of habitat predicted as suit-
able for this species is unlikely to contribute to population viability. Given the large variability in output resulting from using different
These habitats are therefore unlikely to support long-term species SDM techniques, data or environmental change scenarios (Appendix
persistence and should not be given high conservation priority. This S1), it is important to quantify uncertainty in environmental suitabil-
study highlights the importance of using, e.g. hybrid SDM-popula- ity predictions used to make decisions (Moilanen et al. 2006; Carv-
tion models and/or the use of proximal environmental variables alho et al. 2011). However, it is critical that conservation scientists
(Austin 2007) directly relevant to the species’ demography (Eckhart specify which components of uncertainty are estimated (Barry &
et al. 2011) when predictions of species’ persistence are the primary Elith 2006) and which are not. For example, using an ensemble of
modelling output. global climate models (GCMs) to project future distributions will
provide a suite of projections from which means and variances of
suitability can be calculated. This measure of uncertainty, however,
Type of error
can only capture the uncertainty derived from different projections
Species distribution model predictions are susceptible to two types of future climate and does not include uncertainty that derives from
of errors (Franklin 2010): suitable habitat predicted as unsuitable different model constructions, errors in the species data used to fit
(false negatives) and unsuitable habitat predicted as suitable (false the model, in the estimation of current climate, or in the goodness-
positives). Both errors can be costly when using SDMs to support of-fit of the SDM. In addition, this uncertainty estimate assumes
conservation decisions. For example, for biological invaders, false that the ensemble model captures the spectrum of potential future
negatives are considered more serious than false positives at the climates: an attribute that the current suite of GCMs is not designed
pre-border stage, as underestimating the extent of a species’ poten- to have (Schwartz 2012). New structured approaches for dealing
tial distribution could lead to an incorrect decision to allow import with uncertainty associated with SDM outputs (Barry & Elith 2006;
(Pheloung et al. 1999), which might subsequently lead to high Appendix S1) exist in conservation decision support tools such as
impact and mitigation costs (Yokomizo et al. 2009). However, for Marxan (Carvalho et al. 2011) and Zonation (Moilanen et al. 2006).
established invaders, both types of errors can matter. False negatives These generally involve some form of assessment of the robustness
may result in invaders being incorrectly labelled as harmless in a of decisions to large errors in key data, models or assumptions (Re-
given area, leading to a failure to establish appropriate surveillance gan et al. 2005; Wintle et al. 2011). For instance, info-gap decision
or containment measures. Alternatively, false positives can lead to theory has been used to identify reserve networks that achieve con-
wasted surveillance effort, or concentration of management effort servation targets with the highest robustness to uncertainty (Moila-
in inappropriate areas (Baxter & Possingham 2011). Deciding how nen et al. 2006). Because much uncertainty about the predictions of
to balance both types of error will thus vary from one decision- SDMs is irreducible (Regan et al. 2005; Barry & Elith 2006), meth-
making context to another, depending on the consequences of the ods for explicitly dealing with this uncertainty in decision making
errors in relation to the conservation objective. Errors can emanate will be critical for successful application.
from several sources (e.g. data, algorithm, parameterisation options),
but one factor that has a direct effect on error rates is the choice of
WHY HAVE SUCCESSFUL EXAMPLES OF SDM SUPPORTING
a threshold to classify continuous predictions of environmental suit-
DECISION MAKING BEEN SO POORLY REPORTED?
ability as either ‘unsuitable’ or ‘suitable’ (Franklin 2010). Several cri-
teria exist that depend on the type of species data. For SDMs built We have found evidence that SDMs can help guide decisions (e.g.
with presence-only data, predictions of environmental suitability are Brown et al. 2000; Sober on et al. 2001; NTA 2007; US Fish & Wild-
not probabilities of occupancy but rather relative surrogates of life Service 2007; CTFC 2008; Cayuela et al. 2009; NTA 2009;
occupancy, as the baseline probability of occupancy (i.e. prevalence) DMAH 2010; Lepitzki & Pacas 2010; Environment Canada 2011;
is typically unknown and cannot be used as the criterion. For pres- NPS Seki 2011), but most examples are hidden in the grey literature
ence–absence SDMs, the decision to set a certain threshold can be and only rarely reported in the peer-reviewed literature. Our keyword
formally considered by explicitly accounting for the respective con- search (Fig 1 and Appendix S3) suggested that applications to deci-
sequences of each type of error (omissions, commissions) when sion problems are rare compared to the breadth of published SDM-
choosing a threshold, or by using different thresholds for different based conservation papers. This suggests that reporting, to the scien-
decisions (e.g. when to monitor, when to eradicate, when to change tific community, of successful use of SDMs to support decision mak-
categorisation of threat; Field et al. 2004; Royle & Link 2006). A ing is sparse, and leaves open the question as to how many of these
promising alternative is to base decisions on the continuous envi- successful applications actually exist but remain largely hidden? A
ronmental suitability predictions derived from SDMs and incorpo- useful perspective in this regard would be to assess comprehensively
rate the uncertainty directly, rather than categorising ‘suitable’ and how frequently and how effectively SDMs have been used in practice
‘unsuitable’ habitat using specific thresholds (Moilanen et al. 2005). to support conservation decisions in a large number of countries.
The important point is that decision makers need to specify the Greater clarity in these issues is incumbent upon both scientists,
intent of SDM predictions so that modellers can understand the who need to better explain the potential value of their models to
implications of the different types of errors. Ideally, this would be managers, and managers, who need to feed the results of existing
an iterative process involving modellers and decision makers, model applications back to scientists. This viewpoint considers the
whereby methodological decisions such as model complexity and whole conservation decision-making framework and process as one

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
Idea and Perspective SDMs for conservation decisions 1431

within which these two groups should have ideally been involved. A researchers and decision makers to ensure that SDMs are designed
variety of decision-making systems exist. Here, we have outlined a to meet the needs of, and constraints faced by decision makers
decision process that entails defining a problem, defining objectives, (Cash et al. 2003; Addison et al. 2013). This could partly be achieved
identifying potential actions, describing consequences of those by making SDMs compliant with the Open Standards for the Practice of
actions, assessing associated uncertainty and considering trade-offs Conservation (Schwartz et al. 2012), an operationalised multi-criteria
among these consequences (Gregory et al. 2012; Schwartz et al. framework used to plan and prioritise conservation actions. In many
2012; Addison et al. 2013; Fig. 2). Having a common, transparent instances, however, decision making does not proceed in a linear
framework that both decision makers and modellers can access is fashion (as in Fig. 2), or managers may object to the use of models
part of the solution to making better conservation decisions. How- (Addison et al. 2013), making it difficult for researchers to design
ever, considerable barriers remain which must be overcome. the most appropriate SDMs. Therefore, the greater the transparency
Broader inclusion of SDMs in decision-making processes seems lim- in the decision-making process (Gregory et al. 2012; Schwartz et al.
ited by engagement impediments (see below). The published cases 2012), the more likely researchers will be able to provide models
of SDMs developed for conservation purposes highlight the need and outputs that are actually useful in that process. In turn, the
for scientists to do a better job of engaging decision makers early in greater the transparency in the modelling tools, and their linkage to
the development of SDMs but also conversely for decision makers ecological theory (Appendix S1), the more likely managers will be
to involve scientists early in the decision process. It is easy for sci- able to use them (Schmolke et al. 2010). We have observed that
entists to become focused on developing and improving tools with SDM applications and their explicit conservation objectives, particu-
relatively little attention to the information needs of decision mak- larly in the grey literature, tend to be insufficiently documented and,
ers. In turn, SDMs remain difficult for non-experts to use confi- therefore, are difficult to assess and reproduce, with some notable
dently, because there are many methodological options, high output exceptions (e.g. the Madagascar case study in Appendix S5, Nature-
variability and many nuances to consider for their targeted applica- Print in S7). Developing SDMs with a clear understanding of the
tions (Addison et al. 2013). Consequently, although scientists and decision problem at hand fosters the development of SDMs that
decision makers often need similar information to solve their deal appropriately with issues such as spatial scale, species consid-
respective questions (e.g. spatially explicit distribution data), these ered, variables to include in the model, time frame for the study and
communities can remain disconnected, with results from research the use of projections of environmental change (Schwartz 2012).
left unread and unused by decision makers, and constraints faced Developing more useful SDMs to assist conservation decisions is
by decision makers not known or not considered by researchers a necessary condition, but obviously not sufficient to have SDMs
(Sober on 2004; Sutherland & Freckleton 2012). routinely used by decision makers. Communication, translation and
There are also cultural differences between researchers and deci- mediation between scientists and decision makers are reported as
sion makers arising from differences in sources of funding, career necessary functions to better bridge the research/management gap
aspirations, temporal contingencies to solve problems, or differences in other fields (Cash et al. 2003), and reported as particularly critical
in the philosophy of the evaluation of the work done (i.e. economic in the case of SDMs (e.g. Schwartz et al. 2012; Addison et al. 2013).
vs. peer-reviewed; Laurance et al. 2012). This disparity results in As suggested by Sober on (2004), these functions could be per-
researchers too rarely communicating with decision makers, and formed by intermediate institutions playing the role of ‘translator’
decision makers too often not inviting researchers (and especially (or facilitators) between scientists and decision makers (Fig. 4), but
modellers) to participate in the decision-making process (Cash et al. the concept can also be expanded to individuals, groups or consor-
2003; Sober on 2004; Addison et al. 2013). The lack of information tia (e.g. BI/FAO/IUCN/UNEP; see van Zonneveld et al. 2011;
exchange across the research/management boundary reflects a fail- Appendix S6). These translators would synthesise, standardise and
ure of researchers to answer real conservation management ques- communicate the most recent scientific insights useful for solving
tions (Knight et al. 2008), and a failure of decision makers to identified problems to managers (Fig. 4), and mediate the different
capitalise on useful research outputs (Schmolke et al. 2010; Addison steps of a structured decision process (Fig. 2) to ensure that model-
et al. 2013). This problem is exacerbated by the almost overwhelm- lers and managers are jointly involved where needed. It is an impor-
ing peer-reviewed science literature, the bulk of which can be hard tant aim of our paper to promote this linkage. Such institutions
to access and/or not directly relevant to management needs (Haines already exist in some countries (see Table 1 in Sober on 2004;
et al. 2004; Sutherland et al. 2004; Pullin & Knight 2005; Knight Appendix S6), but could be promoted in other countries and their
et al. 2008), controversy surrounding terminology and modelling role as translator institutions clarified and made more systematic.
philosophy (Appendix S1) and by the often confidential communi- Such institutions could ensure that modellers are informed on pre-
cation streams that drive agency and organisational decisions (Cash cisely how SDMs are used in particular decision contexts so that
et al. 2003; Schwartz et al. 2012). Finally, SDMs may be used, but their development can be adjusted and improved in future applica-
their conservation application not reported, since practitioners often tions (Fig. 4). Such translators could also ensure that SDMs comply
lack the time or incentive for publishing their findings in the scien- with the Open Standards for conservation discussed above (Sch-
tific literature. wartz et al. 2012). Institutions playing this translator role may stand
alone as governmental or non-governmental bodies (e.g. CONA-
BIO in Mexico or the Future Earth programme; Appendix S6), be
BRIDGING THE GAP BETWEEN MODELLERS AND DECISION
nested within institutions with other primary functions (e.g. univer-
MAKERS
sities, government departments; e.g. Centre for Evidence-Based
Making SDMs more useful in decision making requires improved Conservation; Appendix S6), or be virtual web-based entities such
communication, appropriate translation of scientific and decision- as the recent Environmental Evidence initiative (Pullin & Knight
context knowledge, mediation and timely collaboration between 2005; Appendix S6). Individuals need to be trained, encouraged and

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
1432 A. Guisan et al. Idea and Perspective

Figure 4 Proposed role of ‘Translators’ (being individuals, groups or institutions; Cash et al. 2003; Sober
on 2004) as bridges between SDM development and conservation
decision making. See Figure 2 for details of the steps of the structured decision-making process and where SDM can provide support.

rewarded for taking on ‘translator’ roles and engaging directly with enough to read about a conservation problem, it is incumbent upon
modellers and decision makers. scientists to reach out to decision makers to understand their needs
Translators can provide a valuable service in promoting and sup- in making a decision, and it is incumbent upon decision makers to
porting the development of appropriate tools for management. report to modellers how SDMs have been used to support decisions
However, although an increasing number of online initiatives are to enable iterative improvement of models. More visibility of part-
making it easier for non-experts to directly access biodiversity data nerships between researchers and decision makers in the scientific
and build SDMs through user-friendly web interfaces (Graham et al. literature will motivate the development of better-integrated SDM
2010; Jetz et al. 2012), these web tools only afford – in their current approaches that have a higher chance of being used to inform
implementation – a limited ability to explore different data sets and important conservation decisions. Finally, a better integration of
model settings (Table 2; Appendix S7). They therefore currently SDM science and management would be beneficial to conservation
cannot be considered sufficient alternatives to the direct involve- decision making but would also advance our understanding of basic
ment of professional modellers in a decision process, ideally medi- ecological processes.
ated by translators. For example, key components of the model
building process (e.g. use of a combination of techniques, evaluation
THE OUTLOOK
of model fit and performance, uncertainty assessment, inspection of
response curves) are currently not available in most of the popular This study was motivated by our observation that conserving biodi-
applications (Table 2), although potentially crucial to support deci- versity is important, that SDMs may contribute to this aim, but that
sion making. While we hope that options to refine biodiversity data more useful SDMs can be developed through practice-oriented case
sets and SDM settings become more widely available in the future studies. Conservation science has made significant progress in devel-
(Jetz et al. 2012), we cannot advocate the use of overly simplified oping an applied arm that helps managers make better decisions
tools to support conservation decisions (e.g. the use of box-like (Sutherland et al. 2004; Pullin & Knight 2005; Gregory et al. 2012;
envelopes may inflate areas identified as critical habitat requiring Schwartz et al. 2012; Sutherland & Freckleton 2012). At the same
protection, and thus conservation cost). The increasing availability time, SDMs have benefitted from over two decades of development
of these tools in the future will therefore make close collaboration as a set of tools with many potential conservation applications (Gui-
between modellers and decision makers even more critical, as there san & Thuiller 2005; Rodriguez et al. 2007; Franklin 2010; Peterson
is the potential for perverse conservation decisions to be made on et al. 2011), but have remained largely the purview of academic
the basis of poorly developed and understood models. What we studies that inform other academic scientists. These tools are now
need is not simpler implementations of SDMs, but a wider recogni- sufficiently mature to take on a larger role in supporting conserva-
tion that SDMs should be developed by experts with a clear conser- tion decisions. Yet, although successful SDM applications exist, they
vation objective in mind and a clear knowledge of the decision remain poorly reported in the scientific literature, suggesting the
process in which they take part. Translators, participatory or co- linkage between SDM science and practice is still weak. We identi-
design principles (Appendix S6) may all be involved in achieving fied three critical components likely to better bridge these two com-
useful and appropriately used SDMs. munities. First, SDM scientists need to better engage decision
Better understanding of the decision process and its constraints makers and understand the decision-making process, to better assess
would allow modellers to determine whether or not an SDM can be how and when SDMs could be used to guide conservation deci-
used, and if so, which type of SDM is best suited. It is usually not sions. Second, SDMs must be designed to meet the spatial and tem-

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
Idea and Perspective SDMs for conservation decisions 1433

Table 2 Examples of online SDM tools (web information acquired May 2013) for predicting the distributions of a large number of species. All examples allow users to
upload occurrence data and fit models online, but with very little flexibility in model parameterisation and evaluation. See also Appendix S7.

OpenModeller (OM) coupled with

Atlas of Living Australia National Institute of Invasive Global Biodiversity Information
Programme (ALA) LifeMapper (LM) Species Science (NIISS) Facility (GBIF)

1. Name of supporting Atlas of Living Australia, Consortium of US Universities National Institute of Invasive Species Centro de Refer^encia em Informacß~ao
organisation(s) Canberra (Australian and University of Goias in Science (US consortium of govern. Ambiental (CRIA), Escola Politecnica da
branch of GBIF) Brasil and non-govern. organisations) USP (Poli), and Instituto Nacional de
Pesquisas Espaciais (INPE), Brasil
2. Can occurrence data Yes No No Yes
be vetted for accuracy?
3. Predictors available Climate, topography, Climate Climate Terrestrial – climate; Marine –climate,
land-use bathymetry and satellite data

4. Modelling techniques MaxEnt, GDM BIOCLIM, GARP* Maxent, BRT Envelope Score
5. Spatial coverage Australia Global USA Global
6. Temporal extent of Current Current + Future Current + Future (1 scenario/GCM) Current
predictor variables (3 IPCC scenarios)
7. Uncertainty No No Yes (SD across 3 runs) No
assessment?
8. Website link http://www.ala.org.au/ http://lifemapper.org/ www.niiss.org http://data.gbif.org/http://openmodeller.
sourceforge.net/

9. Link to an official ALA GBIF NIISS GBIF

occurrence database

10. Reference – Stockwell et al. 2006; Graham et al. 2010; Munoz et al. 2011
(if available)

*ANN, Aquamaps, CSM, SVM and ED to be included in future versions.

poral needs of the conservation problems using transparent meth- National Centre for Competence in Research (NCCR) ‘Plant
ods (e.g. Open Standards) that incorporate uncertainties and recog- Survival’ in Neuch^atel. LB benefitted from support from the Cata-
nise model limitations, especially given potential legal consequences lan Government (CARTOBIO and 2010-BE-272 projects) and the
of decisions. Third, decision makers must in turn provide feedback EU-FP7 SCALES (#226852) to attend the workshops.
to modellers about the success or failure of SDMs used to guide
conservation decisions (i.e. practical limitations, key features of suc-
AUTHORSHIP
cess). To achieve progress, we support the role of ‘translators’ (insti-
tutions, groups or individuals) to facilitate the link between AG organised the three workshops and study design, with support
modellers and decision makers. We strongly encourage species dis- from YMB and HPP. All co-authors attended at least one of the
tribution modellers to get involved in real decision-making pro- workshops and/or interacted by videoconference with the group.
cesses that will benefit from their technical input. This strategy has All authors helped outlining the manuscript and contributed sub-
the potential to better bridge theory and practice, and to contribute stantially to its writing. RT and YMB led the invasive literature
to improve both scientific knowledge and conservation outcomes. search with help from SAS, OB, JE, LB and AG. AITT, PRS and
HPP led the reserve selection literature search, with help from LB,
CMP, JRR, SF, JE, LB, INL and AG. JBB and TJR led the translo-
ACKNOWLEDGEMENTS
cation literature search, with help from EMM, CMP, TGM, MRK
We thank Giorgio Ulrich for help with bibliographic searches, C. and AG. INL and TGM led the critical habitat literature search,
Kremmen & T. Allnutt for details on the Madagascar case study, with help from RM, AITT, LB and AG. MWS and BAW contrib-
and Jorge Sober on, A. Townsend Peterson, Craig McInerny and an uted substantially to the bridge with practitioners section. MWS and
anonymous referee for useful comments. AG’s stay in Brisbane, YMB drafted Table 1. AG and OB prepared all figures.
Australia, was supported by the CSIRO McMaster Foundation. The
three workshops (held on December 2011, April and May 2012) REFERENCES
that led to this publication were organised with financial support
and within the framework of the Australian Research Council Centre Addison, P.F.E., Rumpff, L., Bau, S.S., Carey, J.M., Chee, Y.E., Jarrad, F.C. et al.
of Excellence for Environmental Decisions (CEED; http://www.ceed.edu. (2013). Practical solutions for making models indispensable in conservation
decision-making. Divers. Distrib., 19, 490–502.
au) led by HPP. AG benefitted from insights from a project on Araujo, M.B., Alagador, D., Cabeza, M., Nogues-Bravo, D. & Thuiller, W.
applying SDMs to invasive management in Switzerland granted by (2011). Climate change threatens European conservation areas. Ecol. Lett., 14,
the Swiss Federal Office of the Environment (FOEN) and the 484–492.

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
1434 A. Guisan et al. Idea and Perspective

Austin, M. (2007). Species distribution models and ecological theory: a critical Graham, J., Newman, G., Kumar, S., Jarnevich, C., Young, N., Crall, A. et al.
assessment and some possible new approaches. Ecol. Model., 200, 1–19. (2010). Bringing modeling to the masses: a web based system to predict
Barry, S.C. & Elith, J. (2006). Error and uncertainty in habitat models. J. Appl. potential species distributions. Future Int., 2, 624–634.
Ecol., 43, 413–423. Greenwald, N.D., Suckling, K.F. & Pimm, S.L. (2012). Critical habitat and the
Baxter, P.W.J. & Possingham, H.P. (2011). Optimizing search strategies for role of peer review in government decisions. Bioscience, 62, 686–690.
invasive pests: Learn before you leap. J. Appl. Ecol., 48, 86–95. Gregory, R., Failing, L., Harstone, M., Long, G., McDaniels, T. & Ohlson, D.
Brotons, L., Manosa, S. & Estrada, J. (2004). Modelling the effects of irrigation (2012). Structured Decision Making: A Practical Guide to Environmental Management
schemes on the distribution of steppe birds in Mediterranean farmland. Choices. Wiley-Blackwell, Oxford, UK.
Biodivers. Conserv., 13, 1039–1058. Guisan, A. & Thuiller, W. (2005). Predicting species distribution: offering more
Brown, D., Hines, H., Ferrier, S. & McKay, K. (2000). Establishment of a biological than simple habitat models. Ecol. Lett., 8, 993–1009.
information database for regional conservation planning in north-east New South Wales. Haines, A., Kuruvilla, S. & Borchert, M. (2004). Bridging the implementation
NSW National Parks and Wildlife Service Hurtsville, Hurtsville, NSW, Australia. gap between knowledge and action for health. Bull. World Health Organ., 82,
Carvalho, S.B., Brito, J.C., Crespo, E.G., Watts, M.E. & Possingham, H.P. 724–731.
(2011). Conservation planning under climate change: toward accounting for Hannah, L., Midgley, G., Andelman, S., Araujo, M., Hughes, G., Martinez-Meyer,
uncertainty in predicted species distributions to increase confidence in E. et al. (2007). Protected area needs in a changing climate. Front. Ecol.
conservation investments in space and time. Biol. Conserv., 144, 2020–2030. Environ., 5, 131–138.
Cash, D.W., Clark, W.C., Alcock, F., Dickson, N.M., Eckley, N., Guston, D.H. Heinrichs, J.A., Bender, D.J., Gummer, D.L. & Schumaker, N.H. (2010).
et al. (2003). Knowledge systems for sustainable development. Proc. Natl Acad. Assessing critical habitat: evaluating the relative contribution of habitats to
Sci. USA, 100, 8086–8091. population persistence. Biol. Conserv., 143, 2229–2237.
Cayuela, L., Golicher, D.J., Newton, A.C., Kolb, M., de Alburquerque, F.S., Hoegh-Guldberg, O., Hughes, L., McIntyre, S., Lindenmayer, D.B., Parmesan,
Arets, E.J.M.M. et al. (2009). Species distribution modeling in the tropics: C., Possingham, H.P. et al. (2008). Assisted colonization and rapid climate
problems, potentialities, and the role of biological data for effective species change. Science, 321, 345–346.
conservation. Trop. Conserv. Sci., 2, 319–352. Jetz, W., McPherson, J.M. & Guralnick, R.P. (2012). Integrating biodiversity
Chauvenet, A.L.M., Ewen, J.G., Armstrong, D.P., Blackburn, T.M. & Pettorelli, distribution knowledge: toward a global map of life. Trends Ecol. Evol., 27,
N. (2012). Maximizing the success of assisted colonizations. Anim. Conserv., 16, 151–159.
161–169. Johnson, H., Bleich, V.C. & Stephenson, T.R. (2007). Modelling Sierra Nevada
Chen, I.C., Hill, J.K., Ohlemuller, R., Roy, D.B. & Thomas, C.D. (2011). Rapid Bighorn Sheep habitat: applying resource selection functions to species recovery. California
range shifts of species associated with high levels of climate warming. Science, Department of Fish and Game, Bishop, CA.
333, 1024–1026. Kadmon, R., Farber, O. & Danin, A. (2003). A systematic analysis of factors
CTFC (2008). Informe cientıfico sobre la identificacion de zonas de habitat adecuado para la affecting the performance of climatic envelope models. Ecol. Appl., 13, 853–867.
carraca, la terrera comun, la calandria comun y el sison en el a mbito de las IBAs 142 Kearney, M. & Porter, W. (2009). Mechanistic niche modelling: combining
(secans de Lleida y 144(Cogul-Alfe s). Informe in
edit. Centre Tecnol
ogic Forestal physiological and spatial data to predict species ranges. Ecol. Lett., 12, 334–
de Catalunya i Generalitat de Catalunya, Barcelona, Spain. 350.
DEPI (2013). Biodiversity information tools for use in native vegetation decisions. The State Keith, D.A., Akcakaya, H.R., Thuiller, W., Midgley, G.F., Pearson, R.G., Phillips,
of Victoria Department of Environment and Primary Industries, Department S.J. et al. (2008). Predicting extinction risks under climate change: coupling
of Environment and Primary Industries, Melbourne, Australia. stochastic population models with dynamic bioclimatic habitat models. Biol.
DMAH (2010). RESOLUCIO  MAH/3644/2010 de 22 d’octubre per la qual es Lett., 4, 560–563.
fa p ublic l’Acord de declaraci o d’impacte ambiental del Projecte de regadiu i Knight, A.T., Cowling, R.M., Rouget, M., Balmford, A., Lombard, A.T. &
concentraci o parcel l
aria del Segarra-Garrigues. Transformaci o en regadiu, Campbell, B.M. (2008). Knowing but not doing: selecting priority
obres de distribuci o i concentracio parcel l
aria a diversos termes municipals. conservation areas and the research-implementation gap. Conserv. Biol., 22,
Departament de Medi Ambient i Habitatge (DMAH). Diari Oicial de la 610–617.
Generalitat de Catalunya, 5759, 84690–84743. Kremen, C., Cameron, A., Moilanen, A., Phillips, S.J., Thomas, C.D., Beentje, H.
Eckhart, V.M., Geber, M.A., Morris, W.F., Fabio, E.S., Tiffin, P. & Moeller, et al. (2008). Aligning conservation priorities across taxa in Madagascar with
D.A. (2011). The geography of demography: long-term demographic studies high-resolution planning tools. Science, 320, 222–226.
and species distribution models reveal a species border limited by adaptation. Laurance, W.F., Koster, H., Grooten, M., Anderson, A.B., Zuidema, P.A.,
Am. Nat., 178, S26–S43. Zwick, S. et al. (2012). Making conservation research more relevant for
Elith, J. & Leathwick, J.R. (2009). Species distribution models: ecological conservation practitioners. Biol. Conserv., 153, 164–168.
explanation and prediction across space and time. Annu. Rev. Ecol. Evol. Syst., Leathwick, J., Moilanen, A., Francis, M., Elith, J., Taylor, P., Julian, K. et al.
40, 677–697. (2008). Novel methods for the design and evaluation of marine protected
Environment Canada (2011). Recovery Strategy for the Ord’s Kangaroo Rat (Dipodomys areas in offshore waters. Conserv. Lett., 1, 91–102.
ordii) in Canada [Proposed]. Species at Risk Act Recovery Strategy Series. Lepitzki, D.A.W. & Pacas, C. (2010). Recovery strategy and action plan for the banff
Environment Canada, Ottawa, Canada. springs snail (Physella johnsoni) in Canada. Species at Risk Act Recovery Strategy
Ferrier, S., Watson, G., Pearce, J. & Drielsma, M. (2002). Extended statistical Series. Parks Canada Agency, Ottawa, Canada.
approaches to modelling spatial pattern in biodiversity in northeast New Margules, C.R. & Pressey, R.L. (2000). Systematic conservation planning. Nature,
South Wales. I. Species-level modelling. Biodiv. Conser., 11, 2275–2307. 405, 243–253.
Field, S.A., Tyre, A.J., Jonzen, N., Rhodes, J.R. & Possingham, H.P. (2004). Martin, T.E. & Maron, J.L. (2012). Climate impacts on bird and plant communities
Minimizing the cost of environmental management decisions by optimizing from altered animal-plant interactions. Nat. Clim. Change, 2, 195–200.
statistical thresholds. Ecol. Lett., 7, 669–675. McAlpine, C.A., Seabrook, L.M., Rhodes, J.R., Maron, M., Smith, C., Bowen,
Fordham, D.A., Akcakaya, H.R., Araujo, M.B., Elith, J., Keith, D.A., Pearson, R. M.E. et al. (2010). Can a problem-solving approach strengthen landscape
et al. (2012). Plant extinction risk under climate change: are forecast range ecology’s contribution to sustainable landscape planning? Landscape Ecol., 25,
shifts alone a good indicator of species vulnerability to global warming? Glob. 1155–1168.
Change Biol., 18, 1357–1371. McDonald-Madden, E., Bode, M., Game, E.T., Grantham, H. & Possingham,
Franklin, J. (2010). Mapping Species Distribution: Spatial Inference and Prediction. H.P. (2008). The need for speed: informed land acquisitions for conservation
Cambridge University Press, Cambridge. in a dynamic property market. Ecol. Lett., 11, 1169–1177.
Giljohann, K.M., Hauser, C.E., Williams, N.S.G. & Moore, J.L. (2011). McLane, S.C. & Aitken, S.N. (2012). Whitebark pine (Pinus albicaulis) assisted
Optimizing invasive species control across space: willow invasion management migration potential: testing establishment north of the species range. Ecol.
in the Australian Alps. J. Appl. Ecol., 48, 1286–1294. Appl., 22, 142–153.

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS
Idea and Perspective SDMs for conservation decisions 1435

Moilanen, A., Franco, A.M.A., Early, R.I., Fox, R., Wintle, B.A. & Thomas, C.D. viability analysis to prioritize wetland bird conservation actions. Biol. Conserv.,
(2005). Prioritizing multiple-use landscapes for conservation: methods for large 144, 2354–2361.
multi-species planning problems. Proc. Biol. Sci., 272, 1885–1891. Seki, N.P.S (2011). Sierra Nevada Bighorn sheep environmental assessment: research and
Moilanen, A., Wintle, B.A., Elith, J. & Burgman, M. (2006). Uncertainty analysis recovery actions. Sequoia and Kings Canyon National Parks, Park Service, U.S.
for regional-scale reserve selection. Conserv. Biol., 20, 1688–1697. Department of the Interior, Three Rivers, CA, USA.
Moilanen, A., Wilson, K.A. & Possingham, H. (2009). Spatial conservation Sober on, J.M. (2004). Translating life’s diversity: can scientists and policymakers
prioritization: Quantitative methods and computational tools. Oxford University Press, learn to communicate better? Environ. Sci. Policy Sustain. Develop., 46, 10–20.
Oxford, U.K. Sober on, J., Golubov, J. & Sarukhan, J. (2001). The importance of Opuntia in
Munoz, M.E.D., de Giovanni, R., de Siqueira, M.F., Sutton, T., Brewer, P., Mexico and routes of invasion and impact of Cactoblastis cactorum
Pereira, R.S. et al. (2011). openModeller: a generic approach to species’ (Lepidoptera : Pyralidae). Fla. Entomol., 84, 486–492.
potential distribution modelling. Geoinformatica, 15, 111–135. Stockwell, D.R.B., Beach, J.H., Stewart, A., Vorontsov, G., Vieglais, D. &
NTA (2007). Northern Territory weed risk management user guide. Natural Resources Pereira, R.S. (2006). The use of the GARP genetic algorithm and Internet grid
Division, Department of Natural Resources, Environment, The Arts and computing in the Lifemapper world atlas of species biodiversity. Ecol. Model.,
Sport, Northern Territory of Australia, Palmerston, NT, Australia. 195, 139–145.
NTA (2009). NT weed risk assessment report: Andropogon gayanus (gamba grass). Sutherland, W.J. & Freckleton, R.P. (2012). Making predictive ecology more
Natural Resources Division, Department of Natural Resources, Environment, relevant to policy makers and practitioners. Philos. Trans. R. Soc. B, 367, 322–
The Arts and Sport, Northern Territory of Australia, Palmerston, NT, 330.
Australia. Sutherland, W.J., Pullin, A.S., Dolman, P.M. & Knight, T.M. (2004). The need
Peterson, A.T., Sober on, J., Pearson, R.G., Anderson, R.P., Martinez-Meyer, E., for evidence-based conservation. Trends Ecol. Evol., 19, 305–308.
Nakamura, M. et al. (2011). Ecological Niches and Geographic Distributions. Thomas, C.D. (2011). Translocation of species, climate change, and the end of
Princeton University Press, Princeton, USA. trying to recreate past ecological communities. Trends Ecol. Evol., 26, 216–221.
Pheloung, P.C., Williams, P.A. & Halloy, S.R. (1999). A weed risk assessment Thuiller, W., Richardson, D.M., Pysek, P., Midgley, G.F., Hughes, G.O. &
model for use as a biosecurity tool evaluating plant introductions. J. Environ. Rouget, M. (2005). Niche-based modelling as a tool for predicting the risk of
Manage., 57, 239–251. alien plant invasions at a global scale. Glob. Change Biol., 11, 2234–2250.
Possingham, H.P., Andelman, S.J., Noon, B.R., Trombulak, S. & Pulliam, H.R. US Fish and Wildlife Service (2007). Recovery Plan for the Sierra Nevada Bighorn
(2001). Making Smart Conservation Decisions. In Conservation Biology: Research Sheep. U.S. Fish and Wildlife Service, Sacramento, CA.
Priorities for the Next Decade. (eds Soule, M.A., Orians, G.H.). Island Press, Vicente, J., Randin, C.F., Goncalves, J., Metzger, M.J., Lomba, A., Honrado, J.
Washington, pp. 225–244. et al. (2011). Where will conflicts between alien and rare species occur after
Pullin, A.S. & Knight, T.M. (2005). Assessing conservation management’s climate and land-use change? A test with a novel combined modelling
evidence base: a survey of management-plan compilers in the United approach. Biol. Invasions, 13, 1209–1227.
Kingdom and Australia. Conserv. Biol., 19, 1989–1996. Watts, M.E., Ball, I.R., Stewart, R.S., Klein, C.J., Wilson, K., Steinback, C. et al.
Regan, H.M., Ben-Haim, Y., Langford, B., Wilson, W.G., Lundberg, P., (2009). Marxan with Zones: software for optimal conservation based land-
Andelman, S.J. et al. (2005). Robust decision-making under severe uncertainty and sea-use zoning. Environ. Model. Softw., 24, 1513–1521.
for conservation management. Ecol. Appl., 15, 1471–1477. Wintle, B.A., Bekessy, S.A., Keith, D.A., van Wilgen, B.W., Cabeza, M.,
Richardson, D.M., Hellmann, J.J., McLachlan, J.S., Sax, D.F., Schwartz, M.W., Schroder, B. et al. (2011). Ecological-economic optimization of biodiversity
Gonzalez, P. et al. (2009). Multidimensional evaluation of managed relocation. conservation under climate change. Nat. Clim. Change, 1, 355–359.
Proc. Natl Acad. Sci., 106, 9721–9724. Yokomizo, H., Possingham, H.P., Thomas, M.B. & Buckley, Y.M. (2009).
Rodriguez, J.P., Brotons, L., Bustamante, J. & Seoane, J. (2007). The application Managing the impact of invasive species: the value of knowing the density-
of predictive modelling of species distribution to biodiversity conservation. impact curve. Ecol. Appl., 19, 376–386.
Divers. Distrib., 13, 243–251. van Zonneveld, M., Thomas, E., Galluzzi, G. & Scheldeman, X. (2011).
Rondinini, C., Wilson, K.A., Boitani, L., Grantham, H. & Possingham, H.P. Mapping the ecogeographic distribution of biodiversity and GIS tools for
(2006). Tradeoffs of different types of species occurrence data for use in plant germplasm collectors. In: Collecting plant genetic diversity: Technical guidelines -
systematic conservation planning. Ecol. Lett., 9, 1136–1145. 2011 Update (eds. Guarino, L., Rao, V.R. & Goldberg, E.). CAB International
Royle, J.A. & Link, W.A. (2006). Generalized site occupancy models allowing for & Biodiversity International. Available at: http://cropgenebank.sgrp.cgiar.org/
false positive and false negative errors. Ecology, 87, 835–841. index.php?option=com_content&view=article&id=662.
Runge, M.C., Converse, S.J. & Lyons, J.E. (2011). Which uncertainty? Using
expert elicitation and expected value of information to design an adaptive
program Biol. Conserv., 144, 1214–1223. SUPPORTING INFORMATION
Schmolke, A., Thorbek, P., DeAngelis, D.L. & Grimm, V. (2010). Ecological
models supporting environmental decision making: a strategy for the future.
Additional Supporting Information may be downloaded via the online
Trends Ecol. Evol., 25, 479–486. version of this article at Wiley Online Library (www.ecologyletters.com).
Schwartz, M.W. (2012). Using niche models with climate projections to inform
conservation management decisions. Biol. Conserv., 155, 143–156.
Schwartz, M.W., Deiner, C., Forrester, T., Grof-Tisza, P., Matthew, J., Santos, Editor, Hector Arita
M. et al. (2012). Perspectives on the open standards for the practice of Manuscript received 3 June 2013
conservation. Biol. Conserv., 155, 169–177. First decision made 28 June 2013
Sebastian-Gonzalez, E., Sanchez-Zapata, J.A., Botella, F., Figuerola, J., Hiraldo, Second decision made 07 September 2013
F. & Wintle, B.A. (2011). Linking cost efficiency evaluation with population

© 2013 The Authors. Ecology Letters published by John Wiley & Sons Ltd and CNRS