Land Use Policy 90 (2020) 104306

Contents lists available at ScienceDirect

Land Use Policy

journal homepage: www.elsevier.com/locate/landusepol

Effectiveness and feasibility of different management practices to reduce soil T

erosion in an agricultural watershed
G.F. Riccia,*, J. Jeongb, A.M. De Girolamoc, F. Gentilea
a
University of Bari Aldo Moro, Department of Agricultural and Environmental Sciences, Bari, Italy
b
Texas A&M AgriLife Research, Temple, TX, USA
c
National Research Council, Water Research Institute (IRSA-CNR), Bari, Italy

ARTICLE INFO ABSTRACT

Keywords: Erosion is the most widespread form of soil degradation in Europe. EU Member States are called to identify areas
Best Management Practices (BMPs) prone to high risk of soil erosion and to adopt Best Management Practices (BMPs) to decrease land degradation.
SWAT This study is aimed at identifying effective BMPs and their economic feasibility for controlling soil erosion in
Common Agricultural Policy south-central Italy where lands are largely cultivated with winter wheat. The Soil and Water Assessment Tool
Rural Development Program
(SWAT) was applied to simulate the baseline hydrologic and soil erosion processes of the Carapelle basin in the
Economic feasibility
Puglia region, Italy. Calibrated sediment loads were reasonably accurate when statistically evaluated against
measured data (R2 = 0.5, NSE = 0.5, PBIAS=−2.8 %). The model performed equally well for simulating stream
flow rates (R2 = 0.6, NSE = 0.6, PBIAS = 5.3 %). The model maintained reliable performance during the va-
lidation period as well. Average annual specific sediment load was estimated 5.95 t ha-1 yr-1 mostly contributed
by cultivated croplands. Based on regional agricultural policies, four management scenarios were implemented
using the calibrated SWAT model: contour farming (BMP1), no tillage (BMP2); reforestation (BMP3) and contour
farming and reforestation (BMP4). A threshold of sediment yield greater than 10 t ha-1yr-1 was selected to
discretize target treatment areas where these BMPs were applied. Result show that combining contour farming
and reforestation (BMP4) was the most effective (38 % reduction; from 5.95 to 3.70 t ha-1) for erosion control,
followed by BMP2 (29 %; from 5.95 to 4.20 t ha-1), BMP1 (22 %; from 5.95 to 4.61 t ha-1) and BMP3 (15 %; from
5.95 to 5.04 t ha-1). An analysis of the farmer return-production cost ratio (FR/PC) indicated that the baseline
(conventional tillage) and BMP1 were both economically sustainable in areas with slope < 20 % (FR/PC = 1.12
and 1.11, respectively). BMP2 received the highest FR/PC rating of 1.67 in areas with slope < 20 %. The
baseline scenario had no economic advantage (FR/PC = 0.93) in steep slope areas. BMP3 was ranked at the top
(FR/PC = 1.49) followed by BMP2 (FR/PC = 1.41) in areas with slope > 20 %. The results show that a program
of measures can be effective for controlling soil erosion but it must be implemented over long time frames and it
requires relevant investments from the public and private sectors.

1. Introduction Centre (ESDAC) (Panagos et al., 2012), over 20 % of European soils are
estimated to be eroded by water and wind and about 275 ha of soil per
Protecting soils from eroding is a key element of the European day is lost to permanent soil sealing (Jones et al., 2012).
Management Policies (European Commission (EC, 2006) due to the Mediterranean areas are characterized by a higher amount of sedi-
multi-functionality of soils (Wiggering et al., 2006). Lands degraded by ment yield than other areas in Europe due to the presence of active
soil erosion are sensitive to the loss of nutrients and organic matter in badlands, rills, gullies and eroded torrential headwaters (Vanmaercke
the topsoil which leads to lower crop production, higher pollution of et al., 2011). The greater susceptibility in soil erosion in the Medi-
water bodies and destruction of wildlife habitats (Prager et al., 2011). terranean areas is also contributed by the unique pattern of rainfall that
Pimentel and Burgess (2013) evidenced that inappropriate agricultural is concentrated in winter and intense cultivation on steep slopes
practices combined with heavy precipitation, steep topography, low (García-Ruiz et al., 2013). Italy, as part of the Mediterranean areas, is
vegetation cover, deforestation, and overgrazing are some of the main subject to soil erosion especially in croplands (Panagos et al., 2015a;
causes of soil erosion. In a study conducted by the European Soil Data Abdelwahab et al., 2014). Best management practices (BMPs) such as

⁎
Corresponding author.
E-mail address: [email protected] (G.F. Ricci).

https://doi.org/10.1016/j.landusepol.2019.104306
Received 21 November 2018; Received in revised form 30 September 2019; Accepted 12 October 2019
0264-8377/ © 2019 Elsevier Ltd. All rights reserved.
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

contour farming, hill ponds, and grassed waterways, no till farming performance and economic feasibility for both public and private sec-
systems, reforestation, and strip cropping have been widely used in tors.
croplands (Arabi et al., 2008; Mtibaa et al., 2018). Hence, one of the
objectives of the Soil Thematic Strategy (European Commission (EC, 2. Materials and methods
2006) led by Soil European Commission is to identify high risk areas of
soil erosion for every Member State, in order to limit soil erosion im- 2.1. Study area
pacts and to control non point sources (NPS) pollution using specific
conservation measures (Asres et al., 2010; Abdelwahab et al., 2016; The study area is the Carapelle, a watershed (506 km2) located in
Vigiak et al., 2016). the Puglia and the Campania regions (Southern Italy) (Fig. 1). The main
The Common Agricultural Policy (CAP) has been instrumental in channel flows for 52.16 km on 8.2 % slope between 120 and 1089 m
supporting agricultural production from the economic stand-point. a.s.l. of elevation. Climate is typical Mediterranean, characterized by
After the Second World War, however, the financial measures tended to warm and dry summers and wet winters. Indeed, August is the driest
accelerate soil erosion. As a result, the focus has shifted and govern- month while March and November are the wettest period (Abdelwahab
mental incentives to landowners in the last several decades were cen- et al., 2018). Similar to other Mediterranean streams, the Carapelle
tered on the implementation of BMPs to reduce soil losses (Coderoni River is characterized by semi-arid climate regime with seasonal in-
and Esposti, 2018). In addition, funding provided by the European termittent flows exhibiting extremely low flows in the summer months
Union (EU) to counteract soil erosion is subordinated to the main- and high flow in winter and early spring (Romano et al., 2018). Pre-
tenance of a series of mandatory measures: the Good Agricultural En- cipitation ranges between 450 and 800 mm year-1 and flash floods
vironmental Conditions (GAEC) (Panagos et al., 2016). It was estimated events are common from June to October. Rainfall is characterized by
that the application of the GAEC has reduced soil loss from arable lands high spatial variability. These peculiarities of the rainfall regime in-
by 20 % in the past decade (Panagos et al., 2015a). fluences watershed hydrology and sediment transport processes making
In Italy, agronomic measures have become mandatory by two reg- monitoring activities a challenge, especially at daily time scale (Ricci
ulations: the D.M.16809/2008 of the Ministry of Agricultural, Food and et al., 2018).
Forestry Policies (MiPAAF, 2008) about the Cross-Compliance Standard Streamflow and suspended sediment concentration (SSC) have been
Temporary Measures for runoff water control on sloping land; the De- monitored at the outlet of the watershed (41° 17′ 50.347″ N; 15° 36′
cree n. 30125 of 2009 (MiPAAF, 2009) about the minimum land 2.583″ E; Ordona Village) for five years between 2007 and 2011. The
management that meets specific local conditions (Standard 1.1 Creation monitoring report and instrument details can be found in Gentile et al.
of temporary ditches for the prevention of soil erosion) (Bazzoffi et al., (2008) and Gentile et al. (2010). In this area, the main economic ac-
2011). The Rural Development Programme (literally in Italian: Pro- tivity is agriculture with an extensive cultivation of winter wheat on
gramma di Sviluppo Rurale; PSR) is the regional instrument for plan- more than 75 % of the watershed area. In this area, winter wheat is
ning and financing the agricultural system. This seven-year period planted in November and then harvested in July. Traditionally, a 4-year
program is financed by the European Agricultural Fund for Rural De- crop rotation is adopted with mineral fertilizer applications in De-
velopment (EAFRD) and provides 1.6 billion of Euro in the Puglia re- cember (fertilizer grade: 25-15-00) and February (urea). Forest and
gion. In particular, the aims of the target item number 8 and 10 are to pasture are prevalent in the mountainous areas (Aquilino et al., 2014).
expand forests and increase sustainable management in agricultural Urban lands are limited to a few small villages. Due to these aspects of
lands. These planned interventions are related with reducing soil ero- land uses, the main erosion processes are sheet and rill erosion which
sion in agricultural lands. Thus, critical source areas need to be iden- are strongly related to tillage operations. Indeed, farmers usually
tified, and agricultural BMPs be evaluated for their functional effec- practice the conventional tillage over the area (Baseline), which consist
tiveness and economic feasibility (Mtibaa et al., 2018). of plowing up and down slopes (25-40 cm depth) (Fig. 1). Moreover,
A large number of hydrological models have been developed to face past studies carried out in this area found that morphology has a sig-
these challenges with various levels of complexity (Kauffeldt et al., nificant influence on soil erosion (Ricci et al., 2018). Therefore, the soil
2016). For example, the Agricultural Policy Environmental eXtender erosion control policies must be focused on land management as well as
(APEX) (Williams and Izaurralde, 2005), the Water Erosion Prediction topographic effects in source areas.
Project (Flanagan et al., 2012), the Soil Water Assessment Tool (SWAT)
(Arnold et al., 1998), the Annualized Agricultural Non point Source 2.2. BMPs type and applications
(AnnAGNPS) (Bingner and Theurer, 2005), and the Hydrological Si-
mulation Program-FORTRAN (HSPF) (Bicknell et al., 2001) are com- Based on policy guideline found in the National regulation, DM
monly accepted by the global modeling community. In particular, 2490/2017 (MiPAAF, 2017), and Rural Development Program (PSR
SWAT is one of the most used models for watershed assessment because 2014-2020) of the Puglia region (PSR, 2014), the most applied BMPs in
of its complete database enabling the simulation of various manage- the rural areas are: contour farming, no tillage, and reforestation. The
ment practices as well as its complete algorithms to simulate hydrology, Natural Resources Conservation Service (NRCS) conservation practice
sediment transport, and NPS pollutant transport (Arabi et al., 2008; standard (USDA-NRCS, 2017) was used to describe and analyze the
Parajuli et al., 2008; Yang et al., 2010; Tuppad et al., 2010; Betrie et al., effects of the BMPs studied.
2011; Ramos et al., 2015). However, like other hydrological models, Contour farming is a practice that aims at increasing the infiltration
SWAT needs long-term experimental data for calibration and validation and reducing the surface erosion processes (e.g. rills) tilling the land
(De Girolamo et al., 2017). along the contour lines. As a direct effect, there is a reduction of the
In this study, a five-year data for water discharge and suspended fertilizer loss and an increase of the crop yields. Liu et al. (2013)
sediment concentration collected at a stream gauge in the Carapelle pointed out that the major effect is obtained in the slope ranges be-
watershed, a medium-size watershed located in Southern Italy, is used tween 3 % and 8 % since in steep slope areas there could be a high risk
to evaluate the effectiveness of agricultural BMPs for controlling soil of tractor overturning (Abubakar et al., 2010). In Italy, contour farming
erosion. In particular, this paper aims to (i) quantify soil erosion in the is one of the standard land management practices in hilly areas
Carapelle watershed to identify critical source areas under current (Bazzoffi et al., 2011) with the local name of “Girapoggio”. This tech-
managements; (ii) identify specific BMPs for controlling soil erosion nique has been used since the beginning of the 1900 in the Apennine
based on regional policies and evaluate their effectiveness; and (iii) and Sub-Apennine areas especially in the regions of Central Italy.
provide a procedural guideline to choose effective BMPs based on their No tillage or Sod seeding consists of avoiding deep tillage (e.g.

2
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Fig. 1. Study area: the Carapelle watershed (Puglia, SE Italy).

plowing), leaving the residue on the soil and carrying out only in-row soil management practices at different timescales (Arnold et al., 2012a).
operations, such as drilling during the transplanting or furrowing and The SWAT2015 version Rev. 637 (Winchell et al., 2013) was run at
closing the soil during the seeding by means of a seed row/furrow daily timescale for an eight-year period (2004-2011) with the first three
closing device. This practice can lead to reduce surface runoff, increase years (2004-2006) used to warm up the model. The first step for
soil infiltration and, consequently, improve soil properties and reduce modeling the hydrological and sediment processes is the watershed
soil erosion (Ullrich and Volk, 2009). No tillage is a management delineation based on the outlet point defined by the user. In this study,
system still in development, and it is most used in some areas in Central we chose the location of streamflow gauge as the outlet for the purpose
and Northern Italy (Pianura Padana) (Cavalchini et al., 2013). No til- of model calibration. The model divides the watershed into sub-wa-
lage can potentially be adopted in Southern Italy as well due to its tersheds and further divides the sub-watersheds in homogenous areas
positive effects on crop yields and quality (De Vita et al., 2007; Vastola by slope, land use, soil characteristics and management defined as
et al., 2017). Hydrological response units (HRUs).
Reforestation is a management practice that falls into the category In the study watershed, sloping cultivated areas are known to be the
of land use change (LUC), as it is a change to a land use type that is less main source of sediment yield (Abdelwahab et al., 2018; Ricci et al.,
prone to soil erosion (Bakker et al., 2008), and guarantees many posi- 2018). Considering that the aim of the present study was to estimate the
tive effects, including improving water quality, new wildlife habitat and efficiency of the BMPs, a smaller spatial scale (i.e., higher resolution
wood production. In Italy, reforestation increased after the second with greater details) was set for the sub-watersheds definition. The
world war, based on the national environmental policies. Indeed, the number of sub-watersheds and HRUs was increased in comparison with
National Forest and Carbon Inventory (INFC) estimated that the surface previous studies (Abdelwahab et al., 2018; Ricci et al., 2018) in order to
area of forests increased by 20 % in the past 20 years (National Forest better determine the reduction of the annual sediment yield due to the
and Carbon Inventory (INFC, 2015). application of BMPs in upland areas. Hence, it was choosen a threshold
of 200 ha, instead of 2000 ha, as in the previous studies, resulting in
2.3. Modeling hydrology and sediment yield 115 sub-watersheds. In order to keep the proportion of the main soil
type, slope and land uses in every sub-watershed percentage threshold
The semi-distributed, continuous hydrological model SWAT (Arnold of 5 %, 25 %, and 25 % for soil type, slope and land uses, respectively,
et al., 1998, 1993) was developed by the United States Department of were chosen to create 451 HRUs. Subsequently, the model was recali-
Agriculture (USDA) to assess the hydrological, sediment and nutrient brated. The SWAT model estimates surface runoff volume and sediment
regime in a watershed and to examine the impact of the application of yield for each HRU then routes through the channel network to obtain

3
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Table 1
SWAT model input data, source and resolution.
Description Source and data resolution

Digital Elevation Model (DEM) Shuttle Radar Topography Mission (SRTM); resolution of 20 × 20 m
Data source: http://www.sinanet.isprambiente.it
Land use data Land Use Map (UDS) of Puglia and Land Agricultural Use Map (CUAS) of Campania; resolution of 100 m
Data source: http://sit.puglia.it; http://sit.regione.campania.it
Soil database data Agro-ecological Characterization of the Puglia Region ACLA2;
resolution 250 × 250 m; 9 soil profiles
Data source: Regione Puglia, 2001
Weather data 8 Gauging station
Daily data: precipitation, solar radiation, wind speed, relative humidity, min. and max temperature
Data source: http://www.protezionecivile.puglia.it
Measured Data for calibration and validation Four years of measured daily streamflow and sediment load (2007-2011).

the streamflow and sediment load (Neitsch et al., 2011). The surface used for validation. The period from April to December 2009 was dis-
runoff is calculated by the modified Soil Conservation Service‐Curve carded because of a lack of recorded data due to malfunctioning of the
Number method (SCS-CN) (United States Department of Agriculture, power supply system. A preliminary calibration was carried out by
1972), while the stream velocity and discharge relationship is de- adjusting Ground Water flow (GW) parameters (eg. Alpha_BF and
termined by Manning’s equation. The Modified Universal Soil Loss GW_delay) using the Baseflow Filter Program (Arnold and Allen, 1999)
Equation (MUSLE) (Williams, 1975), derived from the Universal Soil in order to have a good concordance between measured and simulated
Loss Equation (USLE), is used to estimate the sediment yield. Subse- values of base flow/surface runoff (Brouziyne et al., 2017). A sensitivity
quently, the processes of channel degradation or sediment deposition in analysis, which involved model parameters that influence runoff and
the channel are simulated with a simplified Bagnold’s stream power sediment prediction (Malagó et al., 2017; Arnold et al., 2015), was
relationship (Bagnold, 1977), where the maximum transport is esti- performed with the SWAT-CUP automated tool applying the Sequential
mated based on the peak channel velocity. Among the equations pro- Uncertainty Fitting version 2 (SUFI2) (Abbaspour et al., 2015). A de-
vided by the model for the potential evapotranspiration, the Hargreaves tailed description of the sensitivity analysis can be found in Ricci et al.
method (Hargreaves, 1975) was selected, which needs temperature and (2018). In addition, the Manning “n” roughness coefficient for the main
solar radiation as input data. A detailed description of the SWAT model channel (CH_N2) was considered in the sediment load calibration
can be found in Abdelwahab, (2018). (Abdelwahab et al., 2018). The USLE P-factor (erosion management
To set up the model, several types of data are required (Table 1). For practice factor varying 0-1) was set to 1.0 because in no conservation
watershed delineation, a digital elevation model (DEM) (20 × 20 m) practice was applied in the watershed. Since the hydrology can be
was used. Soil profiles and characteristics such as texture, field capa- considered the driving factor for all other watershed processes (Arnold
city, wilting point, and hydraulic conductivity were obtained from the et al., 2012b, 2015; Malagó et al., 2015; Brouziyne et al., 2017), the
Agro-ecological Characterization of the Puglia Region project or ACLA2 calibration and the validation were performed first for the hydrology
(250 m). The main soil types were silty clay loam, sand clay loam, and and then subsequently for sediment load. SWAT-CUP allows to conduct
loamy soils. A land use map was developed by merging the Land either manual or automatic calibration, so firstly selected parameters
Agricultural Use Map (CUAS) of Campania (100 m) and the Land Use
Map (UDS) of Puglia (100 m) for greater accuracy. Finally, a SWAT
Table 2
code was assigned to each land use type for database assimilation. As SWAT model parameters used for the calibration and their final values.
reported above, the main crop is winter wheat, which covers more than
75 % of the watershed. Other land uses are forests, olive groves, pas- Parameters Description Calibrated value

ture, and urban areas. Through field surveys and interviews with the Runoff
farmers, management information for winter wheat and olive groves CN2.mgt Curve Number 60-88
were included in the management file (.mgt). For winter wheat, a 4- GWQMN.gw Threshold depth of water in shallow aquifer 1281.62
year crop rotation (wheat, wheat, wheat, clover) was considered with GW_DELAY.gw Groundwater delay time 92.76
ALPHA_BF.gw Baseflow alpha factor 0.59
ploughing, which is perpendicular to the slope lines (25-40 cm depth), GW_REVAP.gw Groundwater “revap” coefficient 0.028
in August, harrowing in October, planting in November and harvesting REVAPMN.gw Threshold depth of water in the shallow 172.61
in July. Fertilizer applications were carried out in December (fertilizer aquifer for "revap" to occur
grade: 25-15-00) and February (urea). For olives trees, three shallow RCHRG_DP.gw Deep aquifer percolation fraction 0.38
SURLAG.bsn Surface runoff lag time [days] 4.00
tillage operations (ploughing and harrowing) were applied every two
SOL_AWC.sol Available water capacity of the soil layer 0.08-0.26
months starting in April, and two fertilizer applications were carried SOL_K.sol Saturated hydraulic conductivity 1.95-13.15
out in December (manure) and in spring (fertilizer grade: 26-00-00), CH_N1.sub Manning's "n" value for the tributary 0.08
while the plants are harvested in November (Abdelwahab et al., 2016). channels
Climate data for the studied period (daily maximum and minimum CH_K1.sub Effective hydraulic conductivity in tributary 1.00
channel
temperatures, daily precipitation, and solar radiation) were acquired CH_K2.rte Effective hyd. Cond. In the main channel 56.68
from eight weather stations located inside the watershed or in the OV_N.hru Manning's "n" value for overland flow 2.99
surrounding areas. Climate data were provided by the Civil Protection Sediment
Agency and by the Agrometeorological Service of Puglia Region. CH_N2.rte Manning’s “n” value for main channel 0.05-0.14
ADJ_PKR.bsn Peak rate adjustment factor for sediment 3.00
routing in the subbasin (tributary channels)
2.4. Model calibration and validation PRF_BSN.bsn Peak rate adjustment factor for sediment 2.9
routing in the main channel
The entire evaluation period (2007-2011) was split into two periods SPEXP.bsn Exponent parameter for calculating sediment 2.00
reentrained in channel sediment routing
(Arnold et al., 2012a) to calibrate and validate the model. The first
SPCON.bsn Maximum amount of sediment reentrained 0.001
period, from January 2007 to April 2009 was used for calibration, during channel sediment routing
whilst the second period, from January 2010 to December 2011, was

4
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Table 3
SWAT parameters modified in BMPs simulations.
BMP code BMP name Selected Criteria Number of Selected HRUs Modified parameters in BMP simulations

-1
BMP1 Contour farming 10 t ha Slope < 20 % (tractor 44 CNII table provided by Arnold et al. (2012b)
overturning risk) Contoured w/residue for every hydrologic group
USLE_P table provided by Arnold et al. (2012b)
values for slope classes
BMP2 No tillage Erosion > 10 t ha-1 59 Removing tillage operation in target HRUs
BIOMIX set to 0.4
OV_N set to 0.320
CNII decreased by 2
BMP3 Reforestation Erosion > 10 t ha-1 Slope > 20 % 15 Change of the land use in the target HRUs
CNII values for crops Contoured w/residue for every
hydrologic group provided in the table provided by Arnold et al.
(2012b)
BMP4 BMP1+BMP3 Same criteria used in BMP1 and BMP3 44 + 15 Same criteria used in BMP1 and BMP3

(Table 2) were changed manually one at a time to reach a close cor- parameters were edited following the recommendation by Arnold et al.
respondence between the simulated and the observed curves (Jeong (2012b). The CN parameter was fixed to the values referred to the
et al., 2010). In the next step, an automatic procedure was used to find contoured farming with residue because in the study area crop residues
the best parameters based on the objective function selected, that was are usually left in the field after harvesting grains. The effect of BMP2
the Nash and Sutcliffe (1970) Efficiency (NSE). In Table 2 the para- was simulated in SWAT by removing tillage operations in the target
meters used for the calibration and their final values are reported. HRUs, as recommended by Ullrich and Volk (2009), while CN was
The model efficiency at daily time scale was evaluated using the decreased by a two-point value as suggested by Mtibaa et al. (2018).
coefficient of determination (R2), NSE, and Percent Bias (PBIAS %) Biological soil mixing efficiency (BIOMIX) and Manning’s roughness
(Van Liew et al., 2003; Moriasi et al., 2007). Acceptable model per- coefficient for overland flow (OV_N) were also modified and increased
formance values are considered based on NSE and R2 > 0.5, PBIAS to 0.4 and 0.320, respectively (Neitsch et al., 2011). BMP3 was im-
≤ ± 25 % for streamflow and ± 55 % for sediment load (negative plemented in SWAT by changing the land use of the 15 selected target
PBIAS indicates overestimations). These values, suggested by Moriasi HRUs (Betrie et al., 2011) as well as the relative CN value which was
et al. (2007) are also adopted in other studies (Furl et al., 2015; switched from wheat to forest (Arnold et al., 2012b). In the BMP4
Nerantzaki et al., 2015; Briak et al., 2016; Zettam et al., 2017; Melaku scenario, a combination of BMP1 and BMP3 was simulated in order to
et al., 2018). take advantage of the effectiveness of joining the two BMPs (Mtibaa
et al., 2018).

2.5. BMPs scenarios evaluation and modeling

2.6. Economic feasibility of BMPs
Four scenarios of BMPs were designed and tested based on the
aforementioned National and Regional policy guideline: contour Two aspects should be taken into consideration to evaluate the
farming (BMP1), no tillage (BMP2), reforestation (BMP3), and a com- BMPs feasibility: the suitability at the field level (farmers) as well as at
bination between contour farming and reforestation (BMP4) (Table 3). the watershed scale. Due to price fluctuation of durum wheat and the
The effect of the implemented scenarios was evaluated by taking into variability of crop yields from one year to another, it is difficult to
account the reduction of both specific sediment load (t ha-1) at the forecast income of farmers precisely. For this reason, the economic
outlet and of landscape sediment yield (t ha-1) at subwatersheds and analysis conducted in this work refers to a historical period - the year
HRUs level. Target areas (HRUs) were selected based on the actual 2018 (Table 4). To make the evaluation more consistent, economical
sediment yield and an average annual threshold value of 10 t ha-1 at values derived from PSR that do not refer to 2018 were discounted with
which level soil erosion risk is known to be high (Kuhlman et al., 2010; an interest rate of 1.02 %. This value was derived from the harmonized
Mtibaa et al., 2018). Therefore, a total of 59 HRUs, all classified as European index suggested by the ISTAT (https://www.istat.it/it/
winter wheat fields, were identified. For BMP1, since contour farming archivio/17484) that takes into account the revaluation of the capital
cannot be carried out in steep slope areas (> 20 %), a second criteria of and the variation of the interests introduced by the European Central
discretization, based on the slope (< 20 %), was applied to narrow the Bank.
applicable areas down to 44 HRUs. The no tillage operation was applied The economic feasibility of the four BMPs scenarios for the private
to all the 59 HRUs in the BMP2 scenario. The BMP3 scenario was sector was evaluated by considering the subsidy policy available for
considered taking into account that in steep slope areas (slope > 20 %) every single farmer in the study area. The PSR data were compared with
it is difficult to till the land with traditional techniques, and so it could the four scenarios after the first year of implementation in terms of farm
be better to convert the land use to forest or to grassland (Zhang et al., returns (FR) and Production Cost (PC; Euro ha-1) ratio (FR/PC). For
2014). Hence, the reforestation was applied to 15 HRUs on steep slopes each BMP, the specific farm return (Euro ha-1) was calculated by mul-
(> 20 %). Finally, the BMP4 scenario is a combination between the tiplying the crop yield to the crop selling unit price and by including an
contour farming and the reforestation that was applied to all the 59 increased or decreased rate of crop yield, based on the slope and the
HRUs. The comparison of the effectiveness of the implemented BMPs BMP considered (Liu et al., 2014; Mtibaa et al., 2018). Any subsidy
scenarios was carried out considering as Baseline the current situation provided by the CAP or PSR was also added. The specific production
in which the HRUs are under the conventional tillage. costs (Euro ha-1), reported in the PSR (PSR, 2014) was increased or
SWAT parameters were modified to implement selected BMPs at decreased on the basis of slope and on the BMP considered (Liu et al.,
specific fields so that the model implementation reflects real world 2014).
practices (Betrie et al., 2011). To reflect the impact of BMP1 many The winter wheat production in 2018, under the conventional til-
authors suggest that parameters including CN and the USLE P factor lage and on slopes less than 20 %, increased by 5 % in the scenario
have to be modified (Arabi et al., 2008; Gevaert et al., 2008; Tuppad BMP1. In BMP2 the crop yield increased by 10 %, because of the re-
et al., 2010; Wang et al., 2018). In this study, values for these duced loss of fertile soil as a result of no tillage practice which was

5
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Table 4
Principal income and costs considered in the economic analysis.
Description Unit Value Data Source

Income
Winter wheat selling unit price Euro t-1 222 https://www.obiettivocereali.com
Wood selling unit price Euro t-1 25.7 PSR, 2014 (measure 8.1)
Subsidy conventional tillage Euro ha-1y-1 100 https://terraevita.edagricole.it
Subsidy contour farming Euro ha-1y-1 100 https://terraevita.edagricole.it
Subsidy No Tillage Euro ha-1y-1 322 PSR, 2014 (measure 10.1.3)
Subsidy Reforestationa Euro ha-1y-1 6000 PSR, 2014 (measure 8.1)
Subsidy Reforestation implantb Euro ha-1y-1 2500 PSR, 2014 (measure 8.1)
Subsidy reforestation lost incomesb Euro ha-1y-1 100 PSR, 2014 (measure 8.1)
Costs
Winter wheat conventional tillage Euro ha-1y-1 684 https://terraevita.edagricole.it
Transaction cost for No Tillage Euro ha-1 63 (10 % of the tillage cost) PSR, 2014 (measure 10.1.3)
Investment for upgrade agricultural machines (No Tillage) Euro 6000 Farmers survey
Forest medium-cycle investment costa Euro ha-1y-1 6051 PSR, 2014 (measure 8.1)
Forest maintainace operationb Euro ha-1y-1 2518 PSR, 2014 (measure 8.1)

a
for the first 12 years.
b
for 12 years.

evidenced by De Vita et al. (2007) and reported by local farmers. Lastly, calculated multiplying the sediment load (t yr-1) by the commercial
a reduction of 10 % in crop yield was applied to fields on slopes higher price of soil estimated in 20$ (Panagos et al., 2015b), converted and
than 20 %, based on the higher susceptibility of soil erosion. discounted at the time of this study (19.46 Euro).
Table 4 summarizes the income and costs data considered in the
economic analysis. The wheat crop selling unit price was set according 3. Results
to commercial reports, the subsidies for baseline and BMPs are based on
the CAP policy, and the PSR. The contour farming is not prescribed in 3.1. Streamflow and sediment load
the PSR; therefore the conventional value was used (100 Euro ha-1). The
unit cost for the conventional tillage (Table 4) was increased to 752 The model showed a satisfactory performance in the estimation of
Euro ha-1 for slopes higher than 20 % to account for the greater use of the streamflow and sediment load at daily time scale, in both calibra-
agricultural machines. The contour farming included additional tion and validation periods. Statistical performance indicators are
farming and harvesting costs as it is more time consuming, therefore a summarized in Figs. 2 and 3. In particular, SWAT underestimated the
unit cost of 717 Euro ha-1 was considered. For the no tillage scenario, streamflow observations in the calibration period (PBIAS + 5.3 %) and
the unit cost was 694 Euro ha-1 (which included also the project and overestimated them in the validation period (PBIAS −17.2 %); it
transaction costs defined as the 10 % of the unit cost) since the out- overestimated the observed sediment load in the calibration period
goings derived from the plowing and the harrowing were excluded. In (PBIAS −2.8 %) and underestimated them in the validation period
this scenario, an additional investment is required (Table 4), which is (PBIAS + 5.1 %). Fig. 2 demonstrates the performance of the model
the investment that the farmers must support to upgrade the agri- being able to reproduce the temporal variability in observed stream-
cultural machines (e.g. sod seeder), based on the ten-year period of flow. While the model performed well to simulate the streamflow,
depreciation. several peaks were underestimated especially during the calibration. In
In the BMP3 scenario, the costs to implant a medium-cycle forest 2011, which was a dry year, the model overestimated streamflow in
(12 years) and to the maintenance were considered as reported in the general during low flow periods. Peak flows were relatively well si-
PSR documents (measure 8.1). In addition, a further subsidy was mulated. For example, the model estimated the highest peak only by 6
guaranteed, for the first 12 years, taking into account the lost incomes % lower (observed 94 m3 s-1; simulated 79 m3 s-1) in the wettest month
caused by the conversion from wheat field to forest. Considering the during the simulation period (November 2010). The average yearly
economic analysis reported in the PSR documents after 12 years the rainfall over the study period was 662 mm with the difference of
forest could be subjected to a first cut that can produce a yield of about 346 mm between the driest year (542 mm in 2007) and the wettest year
87.50 t ha-1 of firewood. Therefore, for the evaluation of the farm re- (888 mm in 2010). Overall, 71 % of the rainfall is lost as evapo-
turns a selling price of 25 Euro t-1 and a cut operation price of 732 Euro transpiration (470 mm) that is a value similar to those obtained by
ha-1, both discounted at 2018, were considered. Romanazzi et al. (2015) in the region. Average yearly surface runoff
The suitability of the BMPs scenarios at the watershed scale was was estimated 86 mm, corresponding to 17 % of the rainfall, and the
evaluated including the implementation cost of the BMPs by the public average yearly total water yield was 180 mm, which was equivalent to
sector (subsidies), the costs related to the crop productivity loss (CPL), 27 % of the rainfall.
and the soil loss value. The total practice cost (TPC) for the public sector The analysis of Fig. 3 confirms that the simulated loads are in line
at the watershed scale was obtained summing the subsidies for each with the observed data, although some peaks are underestimated. The
HRU. The evaluation was carried out in the first year, in the second year highest sediment peak load (November 11, 2010) was underestimated
and after 12 years to take into account the changes in the subsidies for by 49.5 % (observed 3.07 t ha-1; simulated 1.52 t ha-1). A dominant
the BMP3 and BMP4. The economic value of CPL was calculated mul- erosion pattern is observed in the winter months (December to April)
tiplying the Land Productivity loss (LPL, %) by the Crop Area (CA, ha) caused by the high frequency of rainfall events. The average annual
and Crop Productivity (CP, t ha-1) and, finally, by the unit price (Euro t- specific sediment load was 5.95 t ha-1 yr-1, while the specific sediment
1
) (Panagos et al., 2018). LPL (%) is the ratio between the area of severe loads simulated for the driest (2007) and the wettest (2010) years were
erosion (> 10 t ha-1) (SEA, ha) and the total agricultural areas of the 1.46 t ha-1 and 8.41 t ha-1, respectively.
watershed (TAA, ha). This ratio is then multiplied by 0.08 that is the After the validation process, the watershed HRUs were classified
value of crop productivity loss in intensively cultivated agricultural according to sediment yields for further analysis (Fig. 4A). Among
fields (Panagos et al., 2018). The economic value of the soil loss was HRUs, winter wheat fields were found to be the highest contributor of

6
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Fig. 2. SWAT streamflow calibration (A) and validation (B) at daily time scale and statistical indices: the coefficient of determination (R2), the Nash and Sutcliffe
Efficiency (NSE), and the Percent Bias (PBIAS %).

sediment yield while forests and rangelands contributed the lowest. The while the sloping cultivated lands are the sediment source areas. In
yearly sediment yield rate ranged between 0.00 and 63.80 t ha-1 yr-1. order to maximize the effects of BMPs application in the watershed, the
Slope class was used as another factor influencing the sediment pro- sub-watershed representation (Fig. 4B) was used in the following Sec-
duction. As depicted in Fig. 4A and in B, the flat areas near the wa- tion 3.2 “Modeling BMPs”.
tershed outlet are characterized by the lowest sediment yield values,

Fig. 3. SWAT sediment load calibration (A) and validation (B) at daily time scale and statistical indices: the coefficient of determination (R2), the Nash and Sutcliffe
Efficiency (NSE), and the Percent Bias (PBIAS %).

7
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Fig. 4. Annual spatial distribution of the simulated sediment yield (t ha-1), within the HRUs (A) and the sub-watershed (B): Baseline. Numbers in (B) refers to the
SWAT sub-watersheds.

Table 5
Simulated monthly and annual average specific sediment load at the watershed
outlet (for the study period2007-2011).
Specific sediment load (t ha-1)

Months Baseline BMP1 BMP2 BMP3 BMP4

1 1.64 1.26 1.06 1.37 0.99

2 0.92 0.72 0.68 0.81 0.61
3 1.22 0.97 0.96 1.05 0.80
4 0.23 0.18 0.19 0.20 0.14
5 0.02 0.01 0.01 0.01 0.01
6 0.04 0.03 0.03 0.03 0.03
7 0.06 0.05 0.05 0.04 0.03
8 0.01 0.00 0.00 0.01 0.00
9 0.09 0.06 0.07 0.08 0.05
10 0.24 0.17 0.17 0.21 0.14
11 0.76 0.61 0.60 0.64 0.49 Fig. 5. Box plot of sediment yield including only the 59 target HRUs. The
12 0.72 0.54 0.41 0.59 0.41 horizontal line within the box and the numbers indicates the median values,
Total 5.95 4.60 4.23 5.04 3.70 boundaries indicate the 25th and the 75th, whiskers indicate the 5th and the 95th
percentile. The HRUs number within the brackets indicate the target HRUs for
each BMP.
3.2. BMPs scenarios modeling
Annual sediment yield (t ha-1) from the targeted HRUs was com-
Four BMPs scenarios were simulated and their effectiveness was pared between the Baseline scenario and scenarios for the four BMPs
examined at a monthly and a yearly time scale at the watershed outlet (Fig. 5). The Baseline scenario again showed the highest sediment yield,
from 2007 to 2011 (Table 5). BMP4 was the most effective to reduce both in terms of maximum and median value, while the implementation
specific sediment loads by 38 % (from 5.95 t ha-1 to 3.70 t ha-1), fol- of BMP1 resulted in a decrease of the median sediment yield in the
lowed by BMP2 with 29 % (from 5.95 t ha-1 to 4.20 t ha-1) and by BMP1 target HRUs of 5.8 t ha-1. As expected, BMP2 is the practice with the
with 22 % (from 5.95 t ha-1 to 4.61 t ha-1). BMP3 exhibited only a greater impact on managing sediment yield. BMP3 yielded a median
marginal reduction of 15 % (from 5.95 t ha-1 to 5.04 t ha-1). BMP4 and value similar to the Baseline though the values of the 25th percentile
BMP2 were good alternative scenarios particularly in winter months, and the minimum were lower in this scenario. This is due to the fact
when the frequency of rainfall was high, with a reduction of 37 % and that only 15 HRUs were considered for reforestation in the area of
30 %, respectively. Indeed, from December to April the biomass in the 1988 ha (18.6 % of total area of 59 target HRUs). An additional treat-
field leads to increase surface roughness and protects soils from eroding ment to BMP1 by adding BMP3 (as prescribed to BMP4) was appreci-
during intense rainfall events (Mtibaa et al., 2018). During the same able, as demonstrated by the sum of the positive impact of the two
period, BMP1 showed a reduction of 22 % and BMP3 a reduction of 15 BMPs resulting in a decrease of the median and 25th percentile values
%. (Fig. 5).

8
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Fig. 6. Annual spatial distribution of the sediment yield (t ha-1) for the different BMPs scenarios. Numbers refers to the SWAT sub-watersheds.

A visual comparison was carried out to analyze the spatial dis- 8.74 t ha-1 in the Baseline. After implementing BMP3, the sediment
tribution of soil erosion areas at the sub-watershed scale between dif- yield drastically reduced to 0.17 t ha-1 in the same sub-catchment
ferent BMPs scenarios. The baseline scenario (Fig. 4B) showed lower (Figs. 4B and 6).
and upper values of sediment yield varying from 0.01 to 32.76 t ha-1,
respectively. Based on these values, in order to obtain a good visual 3.3. Economic feasibility of BMPs
result, the sediment yield maps were reconstructed with 20 classes.
Fig. 6 depicts the four maps for the implemented BMPs scenarios For economic analysis, the total farmer returns and the costs are
showing the distribution of the sediment yield among the sub-water- referred to the year 2018, for which an average value of 3 t ha-1 was
sheds. If only the target 59 HRUs and their sub-watersheds were con- assumed for the winter wheat yield when managed with the conven-
sidered, the scenarios BMP1 and BMP2 produced a reduction of the tional tillage (PSR, 2014 Agricultural census 2010). On the other hand,
sediment yield by 36 % and 37 %, respectively, compared with the the SWAT model simulation showed a variability in crop yields among
Baseline. BMP4 resulted in a high reduction of sediment yield by 52 %. HRUs and between years, which was confirmed by local farmers.
BMP3 showed a higher reduction of sediment yield in particular sub- Hence, the value of 3 t ha-1 was adjusted for slope higher than 20 % to
watersheds. For instance, sub-watershed 21 has about 85 % of the area lower values but was increased for the scenarios of BMP1, BMP2, and
occupied by winter wheat from which the average sediment yield was BMP4. Table 6 shows the results derived from the economic feasibility

9
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Economic analysis comparing the Baseline and the four implemented BMPs scenarios. Baseline is for winter wheat production with conventional tillage; BMP1 is for contour farming; BMP2 is for No tillage; BMP3 is for

FR= (SY × UP) + S, C = PC + IC; b for the first 12 years; c transaction cost for the implementation of no tillage; Ad the cost of depreciation (6000 Euro) must be added for each farm; e cost for the forest cut after 12
analysis, which are presented in terms of surface unit (Euro ha-1). For

FR/PC
simplicity, the target HRUs were partitioned into two slope classes

1.67

1.11

0.93
1.12
1.11

1.41
1.49
1.49
(< 20 % or > 20 %) even for those scenarios where all the target HRUs
were managed with the same practice (e.g. Baseline and BMP2). Con-
sidering the Baseline, the analysis showed that the profit for the farmers
(FR + SFY)-Ca

corresponds to 82 Euro ha-1 in areas where slopes are less than 20 %

-1

and to a negative value (−53 Euro ha-1) for areas with slopes higher
Euro ha

361-Ad

217-Ad
1548
1548
−53
than 20 %. The BMP1, which was implemented only for 44 HRUs with

81
82
81

slope less than 20 %, showed a value almost similar to the Baseline (81
Euro ha-1). Values obtained for BMP2 were 361 Euro ha-1 in areas with
a slope less than 20 %. The value reduced to 217 Euro ha-1 where slope
Investments Cost

was higher than 20 %. The BMP3, which was applied only in steep slope
areas (> 20 %) resulted in a value of profit of 4849 Euro ha-1 by dis-
Euro ha-1

counting the further incomes despite the lack of harvest in the first 12
years. As expected, the unit profit obtained for BMP4 in the two slope
6051
6051
(IC)

63c

69c

classes was identical to the results highlighted for BMP1 and BMP3.
The analysis of the benefit cost ratio (FR/PC) for the private sector
confirmed the results obtained with the economic feasibility analysis.
reforestation and BMP4 is for the combination between BMP1 and BMP4. Where applicable the target HRUs were split into two slope classes (> or < 20 %).

The Baseline was economically profitable in non-steep slope areas (FR/

Production Costs

PC 1.12) while it was not economically advantageous in areas with

2518b+732e
2518b+732e

slope higher than the 20 % (FR/C 0.93). The BMP3 was economically
Euro ha-1

the most profitable option in steep slope areas (FR/C 1.49) followed by
(PC)

717

752

695
684
717
631

BMP1 (FR/C 1.41). For slopes lower than 20 %, the BMP2 was the most
profitable (FR/PC 1.67) option followed by BMP1 (1.11).
Table 7 shows the economic suitability at the watershed scale per-
formed considering the value of CPL and of soil loss and, for each BMPs
Farm returns

Euro ha-1y-1

scenario, the implementation costs for the first, the second and after 12
years. From the economic point of view, the highest values of CPL and
(FRa)

1055

4849
4849
799

699

981
766
799

soil loss were obtained for the baseline, while the lowest were estimated
for the BMP4 followed by BMP2, BMP1, and BMP3. Regarding the costs
for the implementation, at the first year BMP3 and BMP4 showed
higher costs compared to all other practices, while they are econom-
Subsidy first year

ically advantageous after 12 years when the forest is subjected to the

Euro ha-1y-1

first cut. After 12 years, no subsidies are devoted to the reforestation,

therefore the TPC at the watershed scale become lower than the base-
(SFY)

6000
6000

line.

4. Discussion
2500b+100b
2500b+100b
-1 -1
Subsidy (S)

Euro ha y

4.1. Modeling streamflow and sediment load

100

100

322
100
100
322

Modeling streamflow and sediment load using a hydrological model

in the Mediterranean climate areas is a challenge (De Girolamo et al.,
2015a,b). Nonetheless, the statistical results obtained in the Carapelle
Euro ha-1y-1

watershed were satisfactory for both streamflow and sediment load,

Unit Price

showing that the SWAT model is a useful tool for simulating hydro-
(UP)

25.7
25.7
222
222
222

222

222
222

logical processes in these environments (Abouabdillah et al., 2014;

Mtibaa et al., 2018). In particular, during the validation period (2010-
2011) the model showed slightly lower performances for the stream-
flow mainly due to the extraordinary flood event recorded on 10th
Specific Yield

November 2010 when a daily runoff of 15.89 mm was recorded, that

corresponds to 94 % of average November monthly amount (period:
Euro t-1

87.50
87.50
3.15

2.70

2.97
3.00
3.15
3.30
(SY)

1987-2007). The simulation of this extreme event caused an error of the

model in predicting low flow in the subsequent period due to an
overestimation of the baseflow (Fig. 2B). In SWAT, the values of CN are
adjusted to take into account the antecedent soil moisture conditions,
8708
8708

8708

1988

1988
1988
8708

1988
Area

ha

which are in this case close to saturation, and consequently, the in-
filtration capacity decreases, resulting in a streamflow increase (Niraula
et al., 2012; Qiu et al., 2012).
Num. of

The model showed a tendency to underestimate sediment load

HRUs

during the validation period. Among the various possible causes of the
44

15

15
15
44
44
44

15
–

underestimation, some processes typical of the Mediterranean streams,

such as bank collapse (De Girolamo et al., 2018) might not have been
correctly quantified in SWAT with its Modified USLE equation and the
> 20 %
< 20 %

Baseline
Baseline
Practice
Table 6

BMP1
BMP2
BMP3
BMP4

BMP1
BMP2
BMP3
BMP4

Bagnold’s stream power function (Duvert et al., 2012).

years.

Input data can influence the estimation of streamflow and sediment

a

10
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Table 7
Economic suitability at watershed scale at the first year, second year and after 12 years for each BMP.
Practice LPL CPL CPL*UP Specific sediment load Soil loss value TPC first yeara TPC econd yeara TPC after 12 yeara
Euro (t ha-1) Euro Euro Euro Euro

Baseline 0.021 2510 557115 5.95 4692233 4018699 4018699 4018699

BMP1 0.012 1508 334698 4.61 3597378 4018699 4018699 4018699
BMP2 0.010 1273 282619 4.20 3284563 6393211 6393211 6393211
BMP3 0.017 2043 453554 5.04 3910194 20916699 8988699 3819899
BMP4 0.009 1041 231138 3.70 2893544 20916699 8988699 3819899

* LPL is for Land Production Loss, CPL is for Crop Production Loss, UP is for Unit Price.
a
TPC is for Total Practice Cost=(BMP subsidy × BMP ha) + (subsidy conventional tillage × conventional tillage ha). Soil loss replacement cost: 19.46 Euro t-1.

loads (Heathman et al., 2008) especially in areas with complex topo- follow the current study to take into account two main aspects: firstly,
graphy (Tuo et al., 2016; Zeiger and Hubbart, 2017). In order to im- the subsidy provided by the PSR may not be renewed in the new pro-
prove the performance of the model, the number of the sub-watersheds gramming; secondly, a reduction of the grain yield could occur in
and HRUs was increased taking into account the topography of the particular wet years (Pittelkow et al., 2015).
upland areas. Indeed, the sediment yield generated from the MUSLE is The BMP1 is the second single-BMP scenario in terms of sediment
strongly related to the slope and land use (Han et al., 2013) which are load reduction followed by the BMP3. In the BMP1, the reduction of
greatly influenced by changing the thresholds to discretize the sub- sediment load is caused by reduced surface runoff obtained by im-
watersheds and, consequently, the HRUs. Decreasing this threshold an pounding water in small depressions (Gassman et al., 2006; Arabi et al.,
increment of sub-basins and HRUs was obtained and consequently an 2008). Like the no tillage practice, the BMP1 promoted higher yield by
increase of average slope and increase of land uses present in the sub- reducing the loss of fertile soil and by keeping more moisture in soils.
basins. This study pointed out the relevance of the watershed delinea- For this reason, contour farming has been considered part of the
tion for simulating the impact of BMPs (Arabi et al., 2006). structural practices useful to decrease sheet and rill erosion, together
with strip cropping, contour buffer strips, terraces, grass terraces, and
4.2. Is it possible to reduce soil erosion? tile drain (Santhi et al., 2014; Ginzky et al., 2017). The BMP1 was se-
lected because it falls within the GAEC required by the EU (Borrelli
The main purpose of the modeling carried out with the SWAT model et al., 2016). In Italy, such practice is not so widespread because it is not
was to quantify the effect of conservation measures applied in areas included in the good agricultural conditions and so, usually, it is im-
prone to soil erosion for an integrated watershed management. It has plemented only if an evidence of soil erosion is noticed. Farmers con-
been demonstrated that the implementation of BMPs has to be chosen sider this practice as time consuming because of the creation of addi-
specifically according to the characteristic of individual sites (Xie et al., tional rows in corners and at the end of the field (USDA-NRCS, 2008).
2015; Lal, 2015), thus requiring a complex approach that must consider However, it does not require particular investments in agricultural
not only the reduction of sediment production, but also local environ- machinery (e.g. tractor) while ensuring a reduction of sediment yield by
mental policies and economic feasibility (Haas et al., 2017). Hence, the 36 %. Currently, the required public investment for the BMP1 im-
impact of three single and one combined BMPs scenarios were eval- plementation is the same as the baseline, meanwhile, the benefits at the
uated, considering the sediment budget at the watershed outlet as well watershed scale are estimated in approximately 1.09 × 106 Euro y-1, as
as at HRUs and sub-watershed scale. The implemented scenarios were a result of the minor productivity loss and of the minor soil loss value.
selected by taking into account the guidelines provided by CAP and Hence, this practice should be favored, making it mandatory by law, as
PSR. Generally, the model showed similar trends in terms of sediment many EU Countries have done (Spain, Romania, Belgium, Greece,
reduction at the outlet and the median sediment yield at the HRUs Malta, and Cyprus) (Borrelli et al., 2016), or increasing the subsidies.
scale. In areas with slope higher than 20 %, it was chosen to convert the
Overall, at the watershed scale (Table 5) every modeled scenario land use to forest (BMP3). It was demonstrated that the production of
gave a positive impact on the sediment load both at annual and sediment in the Carapelle watershed is strictly related to the slope in-
monthly time step. In particular, as expected, the monthly analysis crease and the type of land use (Ricci et al., 2018). Moreover, steep
showed that the biggest reduction of sediment load was achieved in the slope areas are the most sensitive to land use change, thus in order to
winter wet months. The BMP2 produced the lowest sediment load at the maximize the effect on soil erosion, these areas have to be converted to
outlet. The same scenario also presented a low value for the median grassland or forest (Zhang et al., 2014). The BMP3 showed a greater
sediment yield at HRUs level. Indeed, the peak runoff and sediment reduction of sediment yield referred to a single sub-watershed (Fig. 6)
yield reduced as surface roughness (OV_N) increased (Tripathi et al., because of the change in land use from one that is prone to soil erosion
2005; Ullrich and Volk, 2009). The positive effect of no tillage to se- to one that minimizes it. Forests reduce the production of sediment by
diment yield found in this study is also reported by Parajuli et al. decreasing the effect of the raindrop splashes and improving the soil
(2013); Maalim and Melesse (2013) and Liu et al. (2014). No tillage has protection from the detachment with surface litters (Xiao et al., 2015).
the advantage of easy applicability but involves an initial investment Despite this, it is considered unrealistic to apply the reforestation to the
for the replacement of the agricultural machineries (e.g. sod seeder). entire watershed; indeed, implanting a forest in an arable land can
Despite this, no tillage practice is attractive to farmers because it re- cause a loss of profit for the farmers. The regional policy in Puglia fi-
duces the production costs if switched from conventional tillage nances this practice through the measure 8.1 of the PSR, allowing the
(Pittelkow et al., 2015). Other benefits of no tillage include improve- use of fast-growing species that can give income in few years, and
ment in soil fertility and humidity in semi-arid areas for yield growth covers almost the 100 % of the cost of implant and maintenance. Fur-
(De Vita et al., 2007). In particular, this practice is financed by the PSR thermore, an additional subsidy is guaranteed for the first 12 years after
through the regional measure 10.1.3 as a part of the GAEC suggested by the implant in order to replace the lack of income. In addition, the
the CAP (De Vita et al., 2007); hence, the initial investment could be savings related to the fewer tillage to which forests are subject to with
easily regained. The BMP2 was found to be the most economically respect to winter wheat fields must be considered.
feasible based on the C-FR analysis referred to the first year of im- As resulted from the economic feasibility assessment, areas with
plementation. However, a long-term economic assessment should slope < 20 % managed with the conventional tillage and contour

11
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

farming showed similar levels of farm return, which means that this Despite the efforts and investments at EU, national and regional
practice is, although not subsidized, a good alternative because it can level to promote soil conservation measures and to support the agri-
reduce the sediment production and does not incur additional expenses. cultural system, the adoption of BMPs for soil erosion control has been
In areas with slope > 20 %, winter wheat production with conventional not significant in the Apulia region. Here, some specific regional
tillage was found to be unprofitable due to insufficient crop yield and characteristics, such as the small size of farms, and institutional factors,
higher expenses (e.g. fuel cost). The negative value is mainly caused by such as the low availability of the regional consultancy services for
the production costs data used for the economic analysis that are at best gathering information and guidance, play a key role in influencing the
average values for the Puglia region. A debate about the convenience in BMPs adoption. Thus, conventional tillage is perceived as more prof-
durum wheat production is not new. Pazienza and Zanni (2009) itable than contour farming. No tillage, which is the most widely ap-
pointed out that if the crop production was low this production could be plied conservation practice in the European countries due to its effec-
unprofitable when the market price goes down. It can be supposed that tiveness in preventing erosion and flooding (Panagos et al., 2016), is
the production costs (i.e. agricultural workers salary) in marginal areas not yet accepted by farmers in the Carapelle basin. This is probably due
is lower than that guaranteed in economically developed regions. to the fact that it needs some time after changing from conventional to
The land use change from wheat to forest in the Puglia region can no tillage before reaching the benefits of improved soil structure.
represent a good solution in steep slope areas where the conventional The present study suggests that additional actions are needed to
tillage is not sustainable (FR/PC 0.93). However, the lost income de- raise the awareness of the soil degradation problem and to remove the
rived from the implementation of this BMP, at least for the first 12 barriers that are limiting the adoption of BMPs. Some indirect inter-
years, could not find agreement among farmers. Hence, a broader ventions, such as information and technical assistance, should be up-
economic evaluation should be considered taking into account the graded to inform farmers and to support those producers who are
profit that can be earned from the first cut of the forest onwards (about willing but unable to change their agricultural systems to adopt BMPs.
12 years after the implant). Considering the economic analysis reported At the international level, platforms for raising awareness and pro-
in the PSR documents a further farm return of 4849 Euro ha-1 was posing actions to address the global soil degradation problem should be
obtained that generated the highest FR/C (1.49) for steep slope areas. sustained and supported as well as it is desirable the funding and re-
Moreover, an increase of subsidies should be devoted to this BMP in search actions to promote resilience to soil degradation factors
consideration of the reduction of the use of fertilizers and, nutrient (Panagos et al., 2016).
loads in surface and groundwater, which makes the environmental
objectives of Water Framework Directive more easily achievable. 5. Conclusions
Finally, as already evidenced by several studies (Strauch et al., 2013;
Ramos et al., 2015; Mtibaa et al., 2018), the combination of two BMPs The present study was aimed at developing a methodological ap-
resulted to be more effective than a single BMP. The implementation of proach to choose the most effective BMPs for sediment yield reduction
the BMP4 allowed obtaining a unit profit similar to the Baseline in areas and economic feasibility. To do this, sediment yield was estimated for
with slope < 20 % while the benefit of the BMP4 implementation was an agricultural watershed in Italy under the conventional tillage man-
greater in the steep slope areas (> 20 %). Indeed, the combination of agement and for four other BMPs scenarios using the SWAT model. The
BMPs suggested in the scenario BMP4 turned out highly effective in the specific farm returns including regional subsidies and costs were com-
region because of the significant reduction in sediment yield by almost puted based on public data available in national and regional rural
40 % on average as the practice focuses on managing steep slope areas policies documents. The SWAT model was proved to be a useful tool for
while limiting farmers’ capital investment. identifying source areas of sediment yield and for simulating various
BMP scenarios.
4.3. Adoption of soil conservation practices: potential barriers At the watershed outlet, no tillage was found to be the single BMP
scenario producing the highest sediment load reduction while refor-
The results of this study demonstrate that watershed planning and estation the least effective. However, the analysis of the sediment yield
management can be effective in responding to soil erosion. On the other revealed that reforestation was the best solution in steep slope areas
hand, the results show that the programs of soil conservation must be because it guarantees a constant ground cover throughout the year
implemented over long time frames and they require investments from preventing the soil particle detachment and, consequently, a lower loss
the public and private sectors. These investments can constitute a limit of soil. Moreover, for steep slope areas, reforestation showed the
in their applicability. The economic evaluation performed in this study highest farm returns considering the subsidies, which can be obtained
highlights that, especially in areas where the crop production is low from the rural development program (measure 8.1). Contour farming
(< 3 t ha-1), the public subsidies are necessary to sustain agriculture produced a fair reduction of sediment load, therefore, it could be con-
and to ensure environmental outcomes. Indeed, without subsidies, only sidered a good alternative scenario for smooth slope areas. Despite this
the BMP1 could be a viable solution in areas with slope < 20 %. practice is not financed by regional programs, it could represent a vi-
Therefore, a relevant public investment is necessary which can only be able solution even if there were not subsidies due to the low investment
tackled on a European scale through the Common Agricultural Policy. costs. Results obtained in this work highlights that there is no a “uni-
This requires studies at European scale operating with a common versal” BMP but the best solution must be identified studying the
methodology and standard datasets able to objectively identify areas characteristics of each target areas.
where remedial measures are needed. However, local studies for This study shows that the programs of measures can be effective for
monitoring human induced changes to the soil every 5-10 years are also responding to soil erosion, however, a relevant public investment is
needed to refine EU policies (Panagos et al., 2015a; Robinson, 2015). required to implement the measures. Moreover, additional actions are
Generally speaking, aside from income differences, several barriers needed to raise the awareness of the soil degradation problem and to
limit the adoption of BMPs both in Italy and in many other countries. remove the barriers that are limiting the adoption of BMPs.
Social and political factors, as well as individual characteristics (i.e.
farmer age and education) in addition to economic incentives, were Acknowledgments
found to influence farmer decisions to participate in environmental
practices (Karali et al., 2014). A lack of awareness of the causes and The authors would acknowledge Dr. Jeff Arnold for his precious
effects of soil erosion, which induces most farmers to not relate BMPs to suggestions concerning the modeling activities and the Grassland Soil
the indirect benefits, and the knowledge gaps in estimating the return of and Water Research Laboratory USDA-ARS (Temple, TX) group for
investments do not favor their adoption. sharing constructive discussions. Thanks are also due to the reviewers

12
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

for their useful comments and recommendations that helped us to im- manual calibration and parameters sensitivity analysis in a semi-arid watershed in
prove the manuscript. North-western Morocco. J. Arab. Geosci. 10, 427. https://doi.org/10.1007/s12517-
017-3220-9.
Cavalchini, A., Rognoni, G.L., Tangorra, F., Costa, A., 2013. Experimental tests on winter
References cereal: sod seeding compared to minimum tillage and traditional plowing. J. Agric.
Eng. 44 (2s). https://doi.org/10.4081/jae.2013.321.
Coderoni, S., Esposti, R., 2018. CAP payments and agricultural GHG emissions in Italy. A
Abbaspour, K.C., Rouholahnejad, E., Vaghefi, S., Srinivasan, R., Yang, H., Kløve, B., 2015.
farm-level assessment. Sci. Total Environ. 627, 427–437. https://doi.org/10.1016/j.
A continental-scale hydrology and water quality model for Europe: calibration and
scitotenv.2018.01.197.
uncertainty of a high-resolution large-scale SWAT model. J. Hydrol. 524, 733–752.
De Girolamo, A.M., Di Pillo, R., Lo Porto, A., Todisco, M., Barca, E., 2018. Identifying a
https://doi.org/10.1016/j.jhydrol.2015.03.027.
reliable method for estimating suspended sediment load in a temporary river system.
Abdelwahab, O.M.M., Bingner, R.L., Milillo, F., Gentile, F., 2014. Effectiveness of alter-
Catena 2018 (165), 442–453. https://doi.org/10.1016/j.catena.2018.02.015.
native management scenarios on the sediment load in a Mediterranean agricultural
De Girolamo, A.M., Barca, E., Pappagallo, E., Lo Porto, A., 2017. Simulating ecologically
watershed. J. Agric. Eng. 45 (3), 125–136. https://doi.org/10.4081/jae.2014.430.
relevant hydrological indicators in a temporary river system. Agric. Water Manag.
Abdelwahab, O.M.M., Bingner, R.L., Milillo, F., Gentile, F., 2016. Evaluation of alter-
180, 194–204. https://doi.org/10.1016/j.agwat.2016.05.034.
native management practices with the AnnAGNPS model in the Carapelle watershed.
De Girolamo, A.M., Lo Porto, A., Pappagallo, G., Tzoraki, O., Gallart, F., 2015a. The
Soil Sci. 181, 293–305. https://doi.org/10.1097/ss.0000000000000162.
hydrological status concept: application at a temporary River (Candelaro, Italy).
Abdelwahab, O.M.M., Ricci, G.F., De Girolamo, A.M., Gentile, F., 2018. Modelling soil
River Res. Appl. 31, 892–903. https://doi.org/10.1002/rra.2786.
erosion in a Mediterranean watershed: comparison between SWAT and AnnAGNPS
De Girolamo, A.M., Gallart, F., Pappagallo, G., Lo Porto, A., 2015b. Assessing flow regime
models. Environ. Res. 166, 363–376. https://doi.org/10.1016/j.envres.2018.06.029.
alterations in a temporary river. J. Hydrol. Hydromech. 3 (63), 263–272. https://doi.
Abouabdillah, A., White, M., Arnold, J.G., De Girolamo, A.M., Oueslati, O., Maataoui, A.,
org/10.1515/johh-2015-0027.
Lo Porto, A., 2014. Evaluation of soil and water conservation measures in a semi-arid
De Vita, P., Di Paolo, E., Fecondo, G., Di Fonzo, N., Pisante, M., 2007. No-tillage and
river basin in Tunisia using SWAT British. Soc. Soil Sci. 30 (4), 539–549. https://doi.
conventional tillage effects on durum wheat yield, grain quality and soil moisture
org/10.1111/sum.12146.
content in southern Italy. Soil Tillage Res. 92 (1–2), 69–78. https://doi.org/10.1016/
Abubakar, M.S., Ahmad, D., Akande, F.B., 2010. A review of farm tractor overturning
j.still.2006.01.012.
accidents and safety pertanika. J. Sci. Technol. 18 (2), 377–385 © Universiti Putra
Duvert, C., Nord, G., Gratiot, N., Navratil, O., Nadal-Romero, E., Mathys, N., et al., 2012.
Malaysia Press ISSN: 0128-7680.
Towards prediction of suspended sediment yield from peak discharge in small
Aquilino, M., Novelli, A., Tarantino, E., Iacobellis, V., Gentile, F., 2014. Evaluating the
erodible mountainous catchments (0.45-22 km2) of France, Mexico and Spain. J.
potential of GeoEye data in retrieving LAI at watershed scale. Proc. SPIE – Int. Soc.
Hydrol. 454, 42–55.
Opt. Eng. 9239https://doi.org/10.1117/12.2067185. Art. No. 92392B.
European Commission (EC), 2006. Commission Staff Working Document “Impact
Arabi, M., Frankenberger, J.R., Engel, B.A., Arnold, J.G., 2008. Representation of agri-
Assessment of the Thematic Strategy on Soil Protection” (SEC(2006)620). Avaliable
cultural conservation practices with SWAT. Hydrol. Process. 22, 3042–3055. https://
online: http://eur‐lex.europa.eu/legal‐content/ EN/TXT/?uri=CELEX:52006PC0232
doi.org/10.1002/hyp.6890.
(Accessed 15 December 2016). .
Arabi, M., Govindaraju, R.S., Hantush, M.M., Engel, B.A., 2006. Role of watershed sub-
Flanagan, D.C., Frankenberger, J.R., Ascough, J.C.I.I., 2012. WEPP: model use, calibra-
division on modeling the effectiveness of best management practices with SWAT. J.
tion, and validation. Trans. Asabe 55 (4), 1463–1477.
Am. Water Resour. Assoc. 42 (2), 513–528. https://doi.org/10.1111/j.1752-1688.
Furl, C., Sharif, H., Jeong, J., 2015. Analysis and simulation of large erosion events at
2006.tb03854.x.
central Texas unit source watersheds. J. Hydrol. 527, 494–504. https://doi.org/10.
Arnold, J.G., Allen, P.M., 1999. Automated methods for estimating baseflow andground
1016/j.jhydrol.2015.05.014.
water recharge from streamflow records. J. Am. Water Resour. Assoc. 35 (2),
García-Ruiz, J.M., Nadal-Romero, E., Lana-Renault, N., Beguería, S., 2013. Erosion in
411–424. https://doi.org/10.1111/j.1752-1688.1999.tb03599.x.
Mediterranean landscapes: changes and future challenges. Geomorphology 198,
Arnold, J.G., Engel, B.A., Srinivasan, R., 1993. In: A Continuous Time, Grid Cell
20–36. https://doi.org/10.1016/j.geomorph.2013.05.023.
Watershed Model. Proceedings of the Application of Advanced Information
Gevaert, V., Van griensven, A., Holvoet, K., Seuntjens, P., Vanrolleghem, P.A., 2008.
Technologies: Effective Management of Natural Resource Conference. ASAE, St.
SWAT developments and recommendations for modelling agricultural pesticide mi-
Joseph, MI.
tigation measures in river basins. Hydrol. Sci. J. Des Sci. Hydrol. 53 (5), 1075–1089.
Arnold, J.G., Moriasi, D.N., Gassman, P.W., Abbaspour, K.C., White, M.J., Srinivasan, R.,
https://doi.org/10.1623/hysj.53.5.1075.
Santhi, C., Harmel, R.D., van Griensven, A., Van Liew, M.W., Kannan, N., Jha, M.K.,
Gentile, F., Bisantino, T., Corbino, R., Milillo, F., Romano, G., Trisorio Liuzzi, G., 2008.
2012a. SWAT: model use, calibration, and validation. Trans. ASABE 55, 1491–1508.
Sediment transport monitoring in a Northern Puglia watershed. WIT Trans. Eng. Sci.
Arnold, J.G., Kiniri, J.R., Srinivasan, R., Williams, J.R., Haney, E.B., Neitsch, S.L., 2012b.
60, 153–161. https://doi.org/10.2495/DEB080161.
Soil & Water Assessment Tool: Input/Output Documentation Version 2012. Texas
Gentile, F., Bisantino, T., Corbino, R., Milillo, F., Romano, G., Trisorio Liuzzi, G., 2010.
Water Resource Institute.
Monitoring and analysis of suspended sediment transport dynamics in the Carapelle
Arnold, J.G., Srinivasan, R., Muttiah, R.S., Williams, J.R., 1998. Large area hydrologic
torrent (Southern Italy). Catena 80 (1), 1–8. https://doi.org/10.1016/j.catena.2009.
modeling and assessment—part 1: model development. J. Am. Water Resour. Assoc.
08.004.
34 (1), 73–89. https://doi.org/10.1111/j.1752‐1688.1998. tb05961.x.
Gassman, P., Osei, E., Saleh, A., Rodecap, J., Norvell, S., Williams, J., 2006. Alternative
Arnold, J.G., Youssef, M.A., Yen, H., White, M.J., Sheshukov, A.Y., Sadeghi, A.M.,
practices for sediment and nutrient loss control on livestock farms in northeast Iowa.
Moriasi, D.N., Steiner, J.L., Amatya, D.M., Skaggs, R.W., Haney, E.B., Jeong, J.,
Agric. Ecosyst. Environ. 117 (2–3), 135–144. https://doi.org/10.1016/j.agee.2006.
Arabi, M., Gowda, P.H., 2015. Hydrological processes and model representation:
03.030.
impact of soft data on calibration. Am. Soc. Agric. Biol. Eng. 58 (6), 24 1637-1660.
Ginzky, H., Dooley, E., Heuser, I.L., Kasimbazi, E., Markus, T., Qin, T., 2017. International
Asres, Awulachew, M.T., S.B, 2010. SWAT based runoff and sediment yield modelling: a
Yearbook of Soil Law and Policy 2017. Springer.
case study of the Gumera watershed in the Blue Nile basin. Ecohydrol. Hydrobiol. 10,
Haas, M.B., Guse, B., Fohrer, N., 2017. Assessing the impacts of Best Management
191–199. https://doi.org/10.2478/v10104-011-0020-9.
Practices on nitrate pollution in an agricultural dominated lowland catchment con-
Bagnold, R.A., 1977. Bed load transport by natural rivers. Water Resour. Res. 13,
sidering environmental protection versus economic development. J. Environ.
303–312. https://doi.org/10.1029/WR013i002p00303.
Manage. 196, 347–364. https://doi.org/10.1016/j.jenvman.2017.02.060.
Bakker, M.M., Govers, G., van Doorn, A., Quetier, F., Chouvardas, D., Rounsevell, M.,
Han, J.-C., Huang, G.-H., Zhang, H., Li, Z., Li, Y.-P., 2013. Effects of watershed subdivi-
2008. The response of soil erosion and sediment export to land-use change in four
sion level on semi-distributed hydrological simulations: case study of the SLURP
areas of Europe: the importance of landscape pattern. Geomorphology 98, 213–226.
model applied to the Xiangxi River watershed, China. Hydrol. Sci. J. 59 (1), 108–125.
https://doi.org/10.1016/j.geomorph.2006.12.027.
https://doi.org/10.1080/02626667.2013.854368.
Bazzoffi, P., Ciancaglini, A., Laruccia, N., 2011. Effectiveness of the GAEC cross-com-
Hargreaves, G.H., 1975. Moisture availability and crop production. Trans. ASAE 18 (5),
pliance standard Short-term measures for runoff water control on sloping land
980–984. https://doi.org/10.13031/2013.36722.
(temporary ditches and grass strips) in controlling soil erosion. Ital. J. Agron. 6 (s1),
Heathman, G.C., Flanagan, D.C., Larose, M., Zuercher, B.W., 2008. Application of the soil
10–24. https://doi.org/10.4081/ija.2011.6.s1.e3. 2011.
and water assessment tool and annualized agricultural non-point source models in
Betrie, G.D., Mohamed, Y.A., Griensven, A., Srinivasan, R., 2011. Sediment management
the St. Joseph River watershed. J. Soil Water Conserv. 63, 552–568. https://doi.org/
modelling in the Blue Nile basin using SWAT model. Hydrol. Earth Syst. Sci. 15,
10.2489/jswc.63.6.552.
807–818. https://doi.org/10.5194/hess‐15‐807‐2011.
Jeong, J., Kannan, N., Arnold, J.G., Glick, R., Gosselink, L., Srninvasan, R., 2010.
Bicknell, B., Imhoff, J., Kittle, J.Jr., Jobes, T., Donigian, A.Jr., Johanson, R., 2001.
Development and integration of subhourly rainfall-runoff modeling capability within
Hydrological Simulation Program-Fortran: HSPF Version 12 User’s Manual. EPA
a watershed model. Water Resour. Mgmt. 24 (15), 4505–4527.
National Exposure Research Laboratory, Athens, GA, USA.
Jones, A., Panagos, P., Barcelo, S., Bouraoui, F., Bosco, C., Dewitte, O., Gardi, C., Erhard,
Bingner, R.L., Theurer, F.D., 2005. AnnAGNPS Technical Processes Documentation,
M., Hervás, J., Hiederer, R., et al., 2012. The State of Soil in Europe: A Contribution
Version 3.2. USDA‐ARS National Sedimentation Laboratory, Oxford, MS, USA.
from JRC to the European Environmental Agency’s Environment State and Outlook
Briak, H., Moussadek, R., Aboumaria, K., Mrabet, R., 2016. Assessing sediment yield in
Report—SOER 2010. Publications Office, Luxembourg.
Kalaya gauged 16 watershed (northern Morocco) using GIS and SWAT model. Int.
Kauffeldt, A., Wetterhall, F., Pappenberger, F., Salamon, P., Thielen, J., 2016. Technical
Soil Water Conserv. Res. 4, 177–185. https://doi.org/10.1016/j.iswcr.2016.08.002.
review of large-scale hydrological models for implementation in operational flood
Borrelli, P., Paustian, K., Panagos, P., Jones, A., Schütt, B., Lugato, E., 2016. Effect of
forecasting schemes on continental level. Environ. Model. Softw. 75, 68–76. https://
Good Agricultural and Environmental Conditions on erosion and soil organic carbon
doi.org/10.1016/j.envsoft.2015.09.009.
balance: a national case study. Land Use Policy 50, 408–421. https://doi.org/10.
Karali, E., Brunner, B., Doherty, R., Hersperger, A., Rounsevell, M., 2014. Identifying the
1016/j.landusepol.2015.09.033.
factors that influence farmer participation in environmental management practices in
Brouziyne, Y., Abouabdillah, A., Bouabid, R., Benaabidate, L., Oueslati, O., 2017. SWAT
Switzerland. Hum. Ecol. 42 (6), 951–963. https://doi.org/10.1007/s10745-014-

13
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

9701-5. strips to abate fecal bacteria and sediment yield using SWAT. Agric. Water Manag.
Kuhlman, T., Reinhard, S., Gaaff, A., 2010. Estimating the costs and benefits of soil 95, 1189–1200. https://doi.org/10.1016/j.agwat.2008.05.006.
conservation in Europe. Land Use Policy 27 (1), 22–32. Pazienza and Zanni, 2009. Fare i conti per deciderese seminare il grano duro. Informatore
Lal, R., 2015. Restoring soil quality to mitigate soil degradation. Sustainability 7 (5), Agrario 44 (2009), 9–11.
5875–5895. https://doi.org/10.3390/su7055875. Pimentel, D., Burgess, M., 2013. Soil erosion threatens food production. Agriculture 3,
Liu, R., Zhang, P., Wang, X., Wang, J., Yu, W., Shen, Z., 2014. Cost-effectiveness and cost 443–463. https://doi.org/10.3390/agriculture3030443.
benefit analysis of BMPs in controlling agricultural nonpoint source pollution in Pittelkow, C.M., Linquist, B.A., Lundy, M.E., Liang, X., van Groenigen, K.J., Lee, J., van
China based on the SWAT model. Environ. Monit. Assess. 186, 9011–9022. https:// Gestel, N., Six, J., Venterea, Rt., van Kessel, C., 2015. When does no-till yield more? A
doi.org/10.1007/s10661-014-4061-6. global meta-analysis. Field Crops Res. 183, 156–168. https://doi.org/10.1016/j.fcr.
Liu, R., Zhang, P., Wang, X., Chen, Y., Shen, Z., 2013. Assessment of effects of best 2015.07.020.
management practices on agricultural non-point source pollution in Xiangxi River Prager, K., Schuler, J., Helming, K., Zander, P., Ratinger, T., Hagedorn, K., 2011. Soil
watershed. Agric. Water Manag. 117, 9–18. https://doi.org/10.1016/j.agwat.2012. degradation, farming practices, institutions and policy responses: an analytical fra-
10.018. mework. Land Degrad. Dev. 2011 (22), 32–46.
Maalim, F.K., Melesse, A.M., 2013. Modelling the impacts of subsurface drainage on Rural Development Programme (PSR) Puglia, 2014. Allegato Metodologia di calcolo di
surface runoff and sediment yield in the Le Sueur Watershed, Minnesota, USA. costi aggiuntivi e mancato guadagno PSR 2014-2020 della Regione Puglia (Reg. (UE)
Hydrol. Sci. J. 58 (3), 570–586. https://doi.org/10.1080/02626667.2013.774088. 808/2014, art. 11). http://psr.regione.puglia.it/).
Malagó, A., Pagliero, L., Bouraoui, F., Franchini, M., 2015. Comparing calibrated para- Qiu, L., Zheng, F., Yin, R., 2012. SWAT-based runoff and sediment simulation in a small
meter sets of the SWAT model for the Scandinavian and Iberian peninsulas. Special watershed, the loessial hilly-gullied region of China: capabilities and challenges. Int.
issue: evaluation of water resources with SWAT. Hydrol. Sci. J. 60 (5), 949–967. J. Sediment Res. 27 (2), 226–234. https://doi.org/10.1016/s1001-6279(12)60030-4.
https://doi.org/10.1080/02626667.2014.978332. Ramos, M.C., Benito, C., Martinez-Casasnovas, J.A., 2015. Simulating soil conservation
Malagó, A., Bouraoui, F., Vigiak, O., Grizzetti, B., Pastori, M., 2017. Modelling water and measures to control soil and nutrient losses in a small, vineyard dominated, basin.
nutrient fluxes in the Danube river basin with SWAT. Sci. Total Environ. 603–604, Agric. Ecosyst. Environ. 213, 194–208. https://doi.org/10.1016/j.agee.2015.08.004.
196–218. https://doi.org/10.1016/j.scitotenv.2017.05.242. Ricci, G.F., De Girolamo, A.M., Abdelwahab, O.M., Gentile, F., 2018. Identifying sediment
Melaku, N.D., Renschler, C.S., Holzmann, H., et al., 2018. Prediction of soil and water source areas in a Mediterranean watershed using the SWAT model. Land Degrad. Dev.
conservation structure impacts on runoff and erosion processes using SWAT model in 29, 1233–1248. https://doi.org/10.1002/ldr.2889.
the northern Ethiopian highlands. J Soils Sediments. J. Soils Sediments 18, 1743. Robinson, D.A., 2015. Moving toward data on soil change. Science 347 (6218), 140.
https://doi.org/10.1007/s11368-017-1901-3. https://doi.org/10.1126/science.347.6218.140.
MiPAAF, 2008. Decreto del Ministero delle Politiche Agricole Alimentari e Forestali Romanazzi, A., Gentile, F.G., Polemio, M., 2015. Modelling and management of a
(MiPAAF) n. 16809, del 24 novembre 2008, pubblicato sulla G.U.R.I. n. 302 del 29/ Mediterranean karstic coastal aquifer under the effects of seawater intrusion and
12/2008, che modifica ed integra il DM 12541 del 21 dicembre 2006, “Disciplina del climate change. Environ. Earth Sci. 74 (1), 115–128. https://doi.org/10.1007/
regime di condizionalità della PAC e abrogazione del Decreto Ministeriale del 15 s1266501544236.
dicembre 2005”. Romano, G., Abdelwahab, O.M.M., Gentile, F., 2018. Modeling land use changes and their
MiPAAF, 2009. Decreto del Ministero delle Politiche Agricole Alimentari e Forestali impact on sediment load in a Mediterranean watershed. Catena 163, 342–353.
(MiPAAF) n. 30125, del 22 dicembre 2009, pubblicato sulla G.U.R.I n. 303 del 31/ https://doi.org/10.1016/j.catena.2017.12.039.
12/2009. Disciplina del regime di condizionalità ai sensi del regolamento (CE) n.73/ Santhi, C., Kannan, N., White, M., Di Luzio, M., Arnold, J.G., Wang, X., Williams, J.R.,
2009 e delle riduzioni ed esclusioni per inadempienze dei beneficiari dei pagamenti 2014. An integrated modeling approach for estimating the water quality benefits of
diretti e dei programmi di sviluppo rurale. conservation practices at the River Basin Scale. J. Environ. Qual. 43 (1), 177. https://
MiPAAF, 2017. De Decreto del Ministero delle Politiche Agricole Alimentari e Forestali doi.org/10.2134/jeq2011.0460.
(MiPAAF) n. 2490 del 25 gennaio 2017. Pubblicato sulla G.U.R.I n. 74 del 29/3/ Strauch, M., Lima, J.E.F.W., Volk, M., Lorz, C., Makeschin, F., 2013. The impact of best
2017, Disciplina del regime di condizionalità ai sensi del regolamento (UE) n. 1306/ management practices on simulated streamflow and sediment load in a Central
2013 e delle riduzioni ed esclusioni per inadempienze dei beneficiari dei pagamenti Brazilian catchment. Environ. Manag. 127, S24–S36. https://doi.org/10.1016/j.
diretti e dei programmi di sviluppo rurale. jenvman.2013.01.014.
Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harmel, R.D., Veith, T.L., Tripathi, M.P., Panda, R.K., Raghuwanshi, N.S., 2005. Development of effective man-
2007. Model evaluation guidelines for systematic quantification of accuracy in wa- agement plan for critical subwatersheds using SWAT model. Hydrol. Process. 19 (3),
tershed simulations. Trans. ASABE 50, 885–900. 809–826. https://doi.org/10.1002/hyp.5618.
Mtibaa, S., Hotta, N., Irie, M., 2018. Analysis of the efficacy and cost-effectiveness of best Tuo, Y., Duan, Z., Disse, M., Chiogna, G., 2016. Evaluation of precipitation input for
management practices for controlling sediment yield: a case study of the Joumine SWAT modeling in Alpine catchment: a case study in the Adige river basin (Italy). Sci.
watershed, Tunisia. Sci. Total Environ. 616–617 (2018), 1–16. https://doi.org/10. Total Environ. 573, 66–82. https://doi.org/10.1016/j.scitotenv.2016.08.034.
1016/j.scitotenv.2017.10.290. Tuppad, P., Kannan, N., Srinivasan, R., Rossi, C.G., Arnold, J.G., 2010. Simulation of
National Forest and Carbon Inventory (INFC), 2015. Third National Forest Inventory. agricultural management alternatives for watershed protection. Water Resour.
Detailed Information Can Be Found at. https://www.sian.it/inventarioforestale/jsp/ Manag. 24 (12), 3115–3144. https://doi.org/10.1007/s11269-010-9598-8.
metodo_introa2015.jsp?menu=2. Ullrich, A., Volk, M., 2009. Application of the soil and water assessment tool (SWAT) to
Nash, J.E., Sutcliffe, J.V., 1970. River flow forecasting through conceptual models part I – predict the impact of alternative management practices on water quality and quan-
a discussion of principles. J. Hydrol. 10, 282–290. https://doi.org/10.1016/0022- tity. Agric. Water Manag. 96, 1207–1217. https://doi.org/10.1016/j.agwat.2009.03.
1694(70)90255-6. 010.
Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R., 2011. Soil and Water Assessment United States Department of Agriculture - Natural Resources Conservation Service, 2008.
Tool: Theoretical Documentation V. 2009. Texas Water Resources Institute Technical Conservation Practice Standard for Contour Farming (Acre), Code 330.
Report No. 406 Texas A&M University System College Station, Texas 77843-2118. United States Department of Agriculture - National Conservation Practice Standards,
Nerantzaki, S.D., Giannakis, G.V., Efstathiou, D., Nikolaidis, N.P., Sibetheros, I.A., 2017. National Conservation Practice Standards-. Available online at. https://www.
Karatzas, G.P., Zacharias, I., 2015. Modeling suspended sediment transport and as- nrcs.usda.gov/wps/portal/nrcs/detailfull/national/technical/cp/ncps/?cid=
sessing the impacts of climate change in a karstic Mediterranean watershed. Sci. Total nrcs143_026849.
Environ. 538, 288–297. United States Department of Agriculture–Soil Conservation Service, 1972. National
Niraula, R., Norman, L.M., Meixner, T., Callegary, J.B., 2012. Multi-gauge calibration for Engineering Handbook, Section 4, Hydrology. USDA Soil Conservation Service.,
modeling the semi-arid Santa Cruz Watershed in Arizona-Mexico Border area using Washington, DC.
SWAT. Air Soil Water Res. 5https://doi.org/10.4137/aswr.s9410. ASWR.S9410. Van Liew, M.W., Arnold, J.G., Garbrecht, J.D., 2003. Hydrologic simulation on agri-
Panagos, P., Borrelli, P., Poesen, J., Ballabio, C., Lugato, E., Meusburger, K., cultural watersheds: choosing between two models. Trans. ASAE 46, 1539–1551.
Montanarella, L., Alewell, C., 2015a. The new assessment of soil loss by water erosion Vanmaercke, M., Poesen, J., Verstraeten, G., De Vewnte, J., Ocakoglu, F., 2011. Sediment
in Europe. Environ. Sci. Policy 54, 438–447. https://doi.org/10.1016/j.envsci.2015. yield in Europe: spatial patterns and scale dependency. Geomorphology 130,
08.012. 142–161.
Panagos, P., Borrelli, P., Robinson, D.A., 2015b. Tackling soil loss across Europe. Nature Vastola, A., Zdruli, P., D’Amico, M., Pappalardo, G., Viccaro, M., Di Napoli, F., Cozzi, M.,
526 (7572). https://doi.org/10.1038/526195d. 195–195. Romano, S., 2017. A comparative multidimensional evaluation of conservation
Panagos, P., Imeson, A., Meusburger, K., Borrelli, P., Poesen, J., Alewell, C., 2016. Soil agriculture systems: a case study from a Mediterranean area of Southern Italy. Land
conservation in Europe: wish or reality? Land Degrad. Dev. 27 (6), 1547–1551. Use Policy 68, 326–333. https://doi.org/10.1016/j.landusepol.2017.07.034.
https://doi.org/10.1002/ldr.2538. Vigiak, O., Malagó, A., Bouraoui, F., Grizzetti, B., Weissteiner, C.J., Pastori, M., 2016.
Panagos, P., Van Liedekerke, M., Jones, A., Montanarella, L., 2012. European Soil Data Impact of current riparian land on sediment retention in the Danube River Basin.
Centre: response to European policy support and public data requirements. Land Use Sustain. Water Qual. Ecol. 8, 30–49. https://doi.org/10.1016/j.swaqe.2016.08.001.
Policy 2012 (29), 329–338. Wang, W., Xie, Y., Bi, M., Wang, X., Lu, Y., Fan, Z., 2018. Effects of best management
Panagos, P., Standardi, G., Borrelli, P., Lugato, E., Montanarella, L., Bosello, F., 2018. Cost practices on nitrogen load reduction in tea fields with different slope gradients using
of agricultural productivity loss due to soil erosion in the European Union: from di- the SWAT model. Appl. Geogr. 90, 200–213. https://doi.org/10.1016/j.apgeog.2017.
rect cost evaluation approaches to the use of macroeconomic models. Land Degrad 08.020.
Dev. 29, 471–484. https://doi.org/10.1002/ldr.2879484. Wiggering, H., Dalchow, C., Glemnitz, M., Helming, K., Müller, K., Schultz, A., Stachow,
Parajuli, P.B., Jayakody, P., Sassenrath, G.F., Ouyang, Y., Pote, J.W., 2013. Assessing the U., Zander, P., 2006. Indicators for multifunctional land use—linking socio-economic
impacts of crop-rotation and tillage on crop yields and sediment yield using a mod- requirements with landscape potentials. Ecol. Indic. 6 (1), 238–249. https://doi.org/
eling approach. Agric. Water Manag. 119, 32–42. https://doi.org/10.1016/j.agwat. 10.1016/j.ecolind.2005.08.014.
2012.12.010. Williams, J.R., 1975. Sediment routing for agricultural watersheds. JAWRA J. Am. Water
Parajuli, P.B., Mankin, K.R., Barnes, P.L., 2008. Applicability of targeting vegetative filter Resour. Assoc. 11, 965–974. https://doi.org/10.1111/j.1752-1688.1975.tb01817.x.

14
G.F. Ricci, et al. Land Use Policy 90 (2020) 104306

Williams, J.R., Izaurralde, R.C., 2005. Chapter 18: the APEX model. Watershed Models. flow diversion terraces. J. Environ. Qual. 39, 220. https://doi.org/10.2134/jeq2009.
CRC Press, Boca Raton, FL, USA, pp. 437–482. 0157.
Winchell, M., Srinivasan, R., Di Luzio, M., Arnold, J.G., 2013. ArcSWAT interface for Zeiger, S.J., Hubbart, J.A., 2017. An assessment of mean areal precipitation methods on
SWAT2012 user’s guide blackland research and extension center, temple. TXTR simulated stream flow: a SWAT model performance assessment. Water 9, 459.
439–464. https://doi.org/10.3390/w9070459.
Xiao, L., Yang, X., Chen, S., Cai, H., 2015. An assessment of erosivity distribution and its Zettam, A., Taleb, A., Sauvage, S., Boithias, L., Belaidi, N., Sanchez-Perez, J.M., 2017.
influence on the effectiveness of land use conversion for reducing soil erosion in Modelling hydrology and sediment transport in a semi-arid and anthropized catch-
Jiangxi, China. Catena 125, 50–60. https://doi.org/10.1016/j.catena.2014.10.016. ment using the SWAT model: the case of the Tafna River (Northwest Algeria). Water 9
Xie, H., Chen, L., Shen, Z., 2015. Assessment of agricultural best management practices (3), 216. https://doi.org/10.3390/w9030216.
using models: current issues and future perspectives. Water 7 (12), 1088–1108. Zhang, S., Liu, Y., Wang, T., 2014. Howland use change contributes to reducing soil
https://doi.org/10.3390/w7031088. erosion in the Jialing River Basin, China. Agric. Water Manag. 133, 65–73. https://
Yang, Q., Zhao, Z., Benoy, G., Chow, T.L., Rees, H.W., Bourque, C.P.-A., Meng, F.-R., doi.org/10.1016/j.agwat.2013.10.016.
2010. A watershed-scale assessment of cost-effectiveness of sediment abatement with

15