1 s2.0 S1470160X2100042X Main

Ecological Indicators 123 (2021) 107377
Contents lists available at ScienceDirect
Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind
Major drivers of land degradation risk in Western Serbia: Current trends

and future scenarios
Veljko Perović a, *, 1, Ratko Kadović b, Vladimir Đurđević c, Dragana Pavlović a, Marija Pavlović a,
Dragan Čakmak a, Miroslava Mitrović a, Pavle Pavlović a
a
Department of Ecology, Institute for Biological Research ‘Siniša Stanković’ – National Institute of the Republic of Serbia, University of Belgrade, Bulevar despota Stefana
142, Belgrade 11060, Serbia
b
Faculty of Forestry, University of Belgrade, Kneza Višeslava 1, Belgrade 11030, Serbia
c
Institute of Meteorology, Faculty of Physics, University of Belgrade, Dobracina 16, Belgrade 11000, Serbia
A R T I C L E I N F O A B S T R A C T
Keywords: Land degradation and desertification (LDD) is one of the greatest ecological challenges of today, with climate
LDD change resulting from anthropogenic factors a major cause of it. Recent projections of LDD in the Mediterranean
MEDALUS region indicate a gradual widening of arid areas due to increased aridity and global warming by the end of the
Drivers
21st century. Therefore, this study used the MEDALUS method to identify sensitivity to LDD in Western Serbia
Climate change
SPEI
between 1986 and 2005 and to assess possible effects of climate change (RCP4.5 and RCP8.5 scenarios) on land
degradation processes by the end of the 21st century. Likewise, analysis of possible major drivers of degradation
was conducted using principal component analysis (PCA) and multiple linear regression analysis (MLRA). The
study revealed that degradation processes in the study area were found to be most influenced by anthropogenic
drivers (34.4%), less so by natural/anthropogenic ones (23.5%), and least by natural factors (20.1%). Results
also showed that critical areas of LDD susceptibility account for nearly 37% of the study area, transitional areas
cover 35%, while 27% constitutes potentially safe areas. Additionally, critical areas were projected to expand by
33.6% (RCP4.5) and 51.7% (RCP8.5) by 2100 as a result of predicted temperature increases and a reduction in
precipitation in the study area. This study also revealed that the Standardised Precipitation-Evapotranspiration
Index (SPEI) better explains the impact of climate change on LDD than other indices, bearing in mind the ca
pacity of this index to detect temporal oscillations in drought in the context of climate change, and it is therefore
a reliable climate parameter for this method.
1. Introduction coexistence, political stability, social equality, etc. (Kadović et al.,

2016).
At present, land degradation and desertification (LDD) is one of the Recent data shows that over 75% of the Earth’s land surface is
greatest environmental challenges caused by climate change resulting already degraded, and this could surpass 90% by 2050. Those regions
mainly from anthropogenic activities. LDD leads to a series of economic with the highest sensitivity to LDD include states in the Sahara region
and social problems due to its negative impact on soil productivity and and large areas stretching between East and South Africa. LDD also af
food availability, biodiversity and ecosystem functioning (Vieira et al., fects a large part of Eastern and Central Asia, parts of South America,
2015; Prăvălie et al., 2017). Desertification is defined as the process of and relatively large expanses of Western Australia and North America
land degradation in arid, semi-arid and dry sub-humid areas occurring (Cherlet et al., 2018; Ferrara et al., 2020). Likewise, LDD is also
as a result of various factors, including climatic variations and human considered a major environmental problem in Europe, particularly in
activities (UNCCD, 1994). The main processes in LDD include the the Mediterranean region because land aridity expansion is 70% higher
chemical, physical and biological degradation of the environment. Be than previously estimated. Within the region, Spain is most at risk, with
sides the environment, LDD also affects society, e.g. the quality of social land that is highly sensitive to degradation covering 49% of the
* Corresponding author.
E-mail address: veljko.perovic@ibiss.bg.ac.rs (V. Perović).
1
ORCID: 0000-0003-3315-6936.
https://doi.org/10.1016/j.ecolind.2021.107377
Received 23 September 2020; Received in revised form 17 December 2020; Accepted 7 January 2021
Available online 22 January 2021
1470-160X/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
V. Perović et al. Ecological Indicators 123 (2021) 107377
country’s total surface area, followed by Greece (34%), Portugal (28%), multiple linear regression analysis (MLRA); (iii) to estimate the possible
Bulgaria (29%), Romania (11%) and Italy (10% of the total area) effects of climate change on LDD processes (using two scenarios of the
(Prăvălie et al., 2017). regional climate model: RCP4.5 and RCP8.5), and (iv) to identify critical
At the national level, in the Republic of Serbia, 86.4% of the total areas of LDD susceptibility using differential local Moran’s I analysis.
surface area is at risk from different types and intensities of land
degradation; approximately 13% of the area is affected by degradation 2. Material and methods
from hazardous and harmful materials (MEP, 2015). However, so far
there has been no detailed assessment of areas threatened by LDD, apart 2.1. Study area
from two local studies focussing on the northern part of the country
(Kadović et al., 2016; Momirović et al., 2019). The Zlatibor District is located in south-west Serbia (Fig. 1). The
Given the significance of LDD-related issues, it is absolutely essential centre of the region is the town of Užice, situated at latitude 43◦ 51′ 21′′
to understand these processes on a spatial and temporal scale, as well as and longitude 19◦ 50′ 28′′ . The district stretches across an area of 6140
to detect major drivers of degradation. To that effect, an array of km2, which is 6.9% of the surface area of Serbia and, in terms of area, is
methodologies and parameters have been developed in many countries the largest district in the country. It borders the Mačva and Kolubara
across the world in the past decades to provide a complex analysis of Districts to the north, the Moravica District to the east, the Raška District
LDD (Prăvălie et al., 2017). The most commonly applied methodologies to the southeast, the Republic of Montenegro to the south and southwest,
include the FAO/UNEP methodology (FAO-UNEP, 1984, 1997), the and Bosnia and Herzegovina to the west. In terms of administration, the
DSPIR framework (GIWA, 2001), MEDALUS (Kosmas et al., 1999), Zlatibor District comprises the town of Užice and the municipalities of
LADA (Liniger et al., 2008) and DRAST (Karavitis et al., 2020). The Arilje, Bajina Bašta, Kosjerić, Nova Varoš, Požega, Priboj, Prijepolje,
Mediterranean Desertification and Land Use (MEDALUS) method was Sjenica and Čajetina.
designed specifically for this European region, with it being used, for The relief of the Zlatibor District rises gradually from north to south.
example, for Italy (Salvati and Bajocco, 2011; Ladisa et al., 2012; The mountains in the study area have typical features of tectonic relief
Smiraglia et al., 2019), Spain (Lavado Contador et al., 2009), Romania and have undergone major changes due to various exogenous processes.
(Prăvălie et al., 2020), and Turkey (Budak et al., 2018; Uzuner and Plateaus dominate this part of Serbia, intersected by gorges and canyons.
Dengiz, 2020), but also for other parts of the world, like Egypt (Bakr The mountains of Zlatibor, Tara and Zlatar, as well as the Pešter Plateau,
et al., 2012), Algeria (Boudjemline and Semar 2018), Iran (Sepehr et al., are notable for their importance and beauty. The climate is mostly
2007), and Morocco (Ait Lamqadem et al., 2018). It is based on the temperate-continental and is characterised by moderately cold to cold
principles of Environmental Sensitive Areas (ESAs), through which LDD winters and mild summers.
is analysed in a complex, multifactorial way (Prăvălie et al., 2017). This Traditionally, land use has been associated with livestock farming,
method involves analysing degraded areas using visual observations, but tourism-related activities are becoming an ever-increasing feature
measuring stations and fields, and the processing of environmental in (Dragović et al., 2008). All watercourses in this area belong to the Black
dicators through the application of statistical and mathematical models Sea basin, but to the Western Morava and Drina hydrological regions.
(Salvati and Bajocco, 2011; Salvati et al., 2016). In addition, in recent The Zlatibor District is characterised by a great diversity of geological
decades, LDD studies have included remote sensing and geographical substrates. Shales and sediments occupy a large area along the Drina
information science (GIS) techniques, primarily through analysing sat River. It is estimated that about 20% of the territory of Western Serbia is
ellite multispectral bands and various indices, along with the application covered by slates, phyllites, sandstones and conglomerates. Large areas
of geostatistical techniques, DEM analysis, etc. As a result, the analysis of massive limestones and alevrolites with inclusions of sandstone,
process has been simplified and the dynamics of change in LDD pro limestone, and ophiolite are found in the south of the region, while the
cesses can be followed more easily (Tavares et al., 2015; Gül and western part is characterised by a serpentinite substrate (Pavlović et al.,
Erşahin, 2019; Salunkhe et al., 2018; Kolios et al., 2018). Specifically, 2017). The land cover of Western Serbia is also very diverse. The most
the multiple ecological and socioeconomic relationships that charac widespread soils are leptosols, dystric cambisols, calcocambisols and
terise LDD call for the development and application of analytical calcomelanosols, followed by stagnosols and vertisols (Mrvić et al.,
frameworks and statistical methodologies that evaluate and quantify the 2013; Pavlović et al., 2017).
spatial and temporal evolution of complex systems (Salvati et al., 2015).
Previous research has shown the close correlation between LDD 2.2. The MEDALUS method and its components
sensitivity and soil characteristics, vegetation type, and climate features,
and also socio-economic factors, land management options and the The MEDALUS method identifies regions that are environmentally
quality of policy responses (Smiraglia et al., 2019; Ferrara et al., 2020). sensitive areas (ESAs). It provides a composite indicator that can be used
Including these factors allows time series to be analysed, trends to be to better understand factors causing LDD and comprises multiple pa
followed, the major drivers of degradation to be determined, and LDD rameters, such as relief, soil, geological substrate, vegetation, climate,
sensitivity in various scenarios to be predicted, particularly in the light and human activities. Such an approach can therefore be seen as a good
of future climate change (Ferrara et al., 2020; Zhang et al., 2020). In this ’early warning’ indicator of the level of sensitivity of soil to LDD and its
regard, projections to date indicate that increased aridity and global changes over time (Salvati and Bajocco, 2011). Each indicator is
warming, as well as rapid population growth, will heighten the risk of generated from several parameters, which combine to produce a quality
LDD in the near future (Huang et al., 2016) and may lead to the gradual indicator, and it should be emphasised that this method allows the
widening of arid areas in many parts of the Mediterranean region number of parameters and indicators used for quality assessment to be
(Giorgi, 2006). changed (Kadović et al., 2016). In this study, the MEDALUS method
Despite evidence of the permanent links between LDD and climate includes indicators and parameters adapted to the methodology
change, no previous study has integrated future projections of climate described by Kosmas et al. (1999).
change with the situation regarding LDD in the Western Balkans and According to the MEDALUS method, there are four types of ESAs
further afield. Bearing this in mind, this study analyses the spatial and based on the stage of LDD (Kosmas et al., 1999): Critical ESAs (C1, C2
temporal dynamics of LDD changes in Western Serbia and assesses the and C3), Fragile ESAs (F1, F2 and F3), Potential ESAs (P) and Non-
major drivers of degradation. Hence, the basic aims of this study are: (i) Threatened ESAs (N). In this study, analysis of ESAs was conducted on
to assess LDD in Western Serbia using the MEDALUS method, taking into the basis of five quality indicators: Climate Quality Index (CQI), Soil
account the socio-economic specificities of the study area; (ii) to identify Quality Index (SQI), Vegetation Quality Index (VQI), Management
the major drivers of LDD using principal component analysis (PCA) and Quality Index (MQI) and Social Quality Index (SoQI) according to the
2
Fig. 1. The location of the study area.
formula (Kosmas et al., 1999): the following formula:
ESAs = (CQI × SQI × VQI × MQI × SoQI) 1/5

(1) AI = P/PET (2)
where P is precipitation and PET the potential evapotranspiration, ob

2.2.1. Climate quality Index (CQI)
tained using the Penman-Monteith method. Rainfall Erosivity is a nu
The Climate Quality Index (CQI) is related to the impact of climate
merical value that quantifies the effects of a raindrop’s impact, but also
variation on LDD, and in this study, it was estimated based on the
provides relevant information on the amount and rate of runoff that
following five parameters: Rainfall, Aridity Index (AI), Rainfall
occurs after rain (Kadović, 1999). When calculating Rainfall Erosivity,
Erosivity, Standardised Precipitation-Evapotranspiration Index (SPEI),
the algorithm developed by Van der Knijff et al. (1999) and Grimm et al.
and Aspect (Table 1). The Aridity Index (AI) was calculated according to
(2003) was used. The SPEI has the capacity to detect temporal oscilla
tions of drought in the context of climate change (Lu et al., 2019;
Table 1 Myoung-Jin et al., 2020). The SPEI time series encompassed a six-month
Description of parameters used to obtain the Climate Quality Index (CQI). period (March-August), while the index was calculated using the SPEI
Indicator Parameter Description Score References package in R environment (https://CRAN.R-project.org/package =
SPEI). The package uses Gamma and log-logistic distributions to stan
CQI Rainfall (mm) >1000 1 Tavares et al.
650–1000 1.5 (2015) dardise SPEI values, while the monthly PET was obtained using the
280–650 1.8 Penman-Monteith method, which can better characterise drought epi
<280 2 sodes. Aspect determines the distribution of solar radiation and tem
Aridity Index (mm/ Humid (>0.65) 1 Prăvălie et al. peratures, which have a significant impact on vegetation growth
mm) Dry sub-humid 1.5 (2017)
(0.5–0.65)
(Kostadinov, 2006). This parameter was calculated using ASTER Global
Semi-arid (<0.5) 2 Digital Elevation Model V003. CQI was calculated according to the
SPEI* SPEI < -2 2 Vicente-Serrano formula by Kosmas et al. (1999):
− 2 < SPEI < -1 1.8 et al. (2010)
− 1 < SPEI < 1.5 CQI = (rainfall × aridity index × rainfall erosivity
-0.50
×Standardised Precipitation − Evapotranspiration index × aspect)1/5
− 0.5 < SPEI < 1
+0.5 (3)
Rainfall* erosivity <610 1 Prăvălie et al.
(MJ mm/ha h yr) 610–730 1.5 (2017) 2.2.2. Soil quality Index (SQI)
2
Soil organic matter (SOM) as the primary indicator of soil quality
>730
Aspect N, NE, NW, W, 1 Kosmas et al.
flat areas (1999) affects the aggregation and stability of the soil structure, the infiltration
S, SE, SW, E 2 rate and available water capacity, as well as resistance to erosion from
water and wind (Salunkhe et al., 2018). Soil texture/granulometric
* Modified MEDALUS parameters.
3
composition, particularly the size and shape of the particles, affects the Table 3
loss of soil by wind or water and also the rate of water infiltration, which Description of parameters used to obtain the Vegetation Quality Index (VQI).
again influences the amount of surface runoff. In this study, SOM con Indicator Parameter Description Score References
tent and granulometric composition were ascertained in 130 soil sam
VQI Fire risk* 212, 331 1 Prăvălie et al.
ples collected from the study area (Fig. 1). An SOM map was obtained 211, 221, 231,241, 1.3 (2017)
using the ordinary kriging interpolation method as an effective linear 242, 311, 321
unbiased estimator (Yao et al., 2019), while the spatial distribution of 324, 243 1.6
textural classes was obtained using the Soil texture plugin in QGIS 222 2
Erosion 311 1 Prăvălie et al.
software (QGIS Development Team, 2020). A 1:100,000 geological map protection* 222, 231, 243, 321, 1.3 (2017)
was taken in vector form from the University of Belgrade’s Faculty of 324
Mining and Geology (Serbia), on the basis of which different types of 241, 242 1.6
geological substrates were grouped into several classes according to 211, 212, 221, 331 2
Drought 212, 311, 324 1 Prăvălie et al.
their petrological and mineralogical composition. Topography and slope
resistance* 321 1.2 (2017)
data was obtained from ASTER Global Digital Elevation Model V003. 221, 222, 243 1.4
SQI (Table 2) was calculated according to Kosmas et al. (1999) as: 231, 241, 242 1.7
211, 331 2
SQI = (texture × parental material × humus content × slope)1/4 (4) Plant cover => 0.80 1 Ferrara et al.
0.72 < 0.80 1.1 (2020)
0.62 < 0.72 1.2
2.2.3. Vegetation quality Index (VQI)
0.50 < 0.62 1.3
VQI was analysed on the basis of four standard parameters for the 0.38 < 0.50 1.4
MEDALUS method: Fire Risk, Erosion Protection, Drought Resistance 0.26 < 0.38 1.5
and Plant Cover. The geospatial data was obtained using the CORINE 0.18 < 0.26 1.6
database (CLC, 2018) and the Normalised Difference Vegetation Index 0.13 < 0.18 1.7
0.11 < 0.13 1.8
(NDVI), which are most commonly used for regional assessments of VQI 0.1 < 0.11 1.9
(Prăvălie et al., 2017, 2020; Budak et al., 2018; Symeonakis et al., 2016), < 0.1 2
and each parameter was created from land cover/use classes according
* Codes are nomenclature of the CORINE Land Cover database.
to the CORINE classification system and NDVI values based on certain
theoretical aspects found in scientific literature (Prăvălie et al., 2017;
Ferrara et al., 2020). Thus, Vegetation Cover (%) was obtained by erosion and soil structures. An agricultural intensity map was generated
classifying NDVI values into 11 categories, while Fire Risk, Erosion from the CORINE database using three categories - low, medium and
Protection and Drought Resistance were obtained using CORINE data by high intensity, while the agricultural policy implementation map,
the grouping of land cover types (Table 3). VQI was calculated according relating to the implementation of existing environmental regulations,
to the formula by Kosmas et al. (1999): was taken from the reference map of the protected natural resources of
the Zlatibor District. This region has been declared a protected area in
(
VQI = fire risk × erosion protection × drought resistance × plant cover)1/4 order to safeguard the geomorphological and hydrological phenomena
(5) of this highly dissected terrain with its striking canyons and gorges,
preserved ecosystems of karst terrain and the rocks of the serpentinite
2.2.4. Management quality Index (MQI) gorges with their particularly diverse and specific vegetation, and
The various anthropogenic environmental pressures contributing to coniferous forests, and rare and endangered animal species, as well as
LDD were quantified in relation to Agricultural Intensity and Policy extremely important, well preserved historical monuments and ethnic
Enforcement. Agricultural Intensity refers to LDD processes such as soil heritage (MEP, 2012). MQI (Table 4) was calculated according to the
formula by Kosmas et al. (1999):
Table 2 MQI = (intensity of land use × protection policies)1/2 (6)

Description of parameters used to obtain the Soil Quality Index (SQI).
Indicator Parameter Description Score References
2.2.5. Social quality Index (SoQI)
In more complex analyses of LDD potential, the index of socio-
SQI Texture L, SCL, SL, LS, CL 1 Kosmas et al.
economic quality is relevant since certain anthropogenic aspects can
SC, SiL SiCL 1.2 (1999)
Si, C, SiC 1.6 directly affect the condition of the soil (Prăvălie et al., 2017). Social
S 2 indicators are linked to LDD processes as a result of human activities,
Parent Shale, schist, basic, 1 Kosmas et al. with Population Density (Pd) and the Old Age Index (Oa) the two basic
material ultra basic, (1999)
parameters used to evaluate the SoQI. Pd is closely related to the
Conglomerates,
unconsolidated
Limestone, marble, 1.7
granite, Table 4
Rhyolite, Ignibrite, Description of parameters used to obtain the Management Quality Index (MQI).
gneiss, Indicator Parameter Description Score References
siltstone, sandstone
Marl, Pyroclastics 2 MQI Agricultural 222, 243, 311, 321, 1 Prăvălie et al.
Organic Kadović et al. intensity* 324, 331 (2017)
Matter, (2016) 211, 231, 242 1.5
(Humus >3.0 1 21, 221 2
Content) Policy Complete: >75% of 1 Kosmas et al.
1.0–3.0 1.5 enforcement area under protection (1999)
<1.0 2 Partially: 25–75% of 1.5
Slope (%) <6 1 Kosmas et al. area under protection
6–18 1.2 (1999) Incomplete: <25% of 2
18–35 1.5 area under protection
>35 2
* Codes are nomenclature of the CORINE Land Cover database.
4
intensity of human pressure on natural resources, while Oa emphasises Table 6

the strong imbalance between a large number of elderly people in Multi-model ensemble members, consisting of results from listed regional
relation to the total population (Tavares et al., 2015). In this study, the climate models (RCM).
parameter Pd (the number of people per square kilometre) was taken at RCM GCM Ensemble member
30 arc-second horizontal resolution in accordance with national cen CCLM4-8-17 CNRM-CERFACS-CNRM-CM5 1
suses and population registers (CIESIN, 2018), while data for the CCLM4-8-17 ICHEC-EC-EARTH 1
parameter Oa was taken from the Statistical Office of the Republic of CCLM4-8-17 MOHC-HadGEM2-ES 1
Serbia (www.stat.gov.rs). SoQI (Table 5) was determined as the geo CCLM4-8-17 MPI-M-MPI-ESM-LR 1
HIRHAM5 ICHEC-EC-EARTH 1
metric mean of these two factors using the following formula (Tavares
RACMO22E ICHEC-EC-EARTH 1
et al., 2015): RACMO22E MOHC-HadGEM2-ES 1
REMO2009 MPI-M-MPI-ESM-LR 1
SoQI = (population density × old age index)1/2 (7) REMO2009 MPI-M-MPI-ESM-LR 2
* GCM stands for Global Climate Model and RCM for Regional Climate Model.
2.3. Climate change scenarios
the basis of which a percentage distribution of possible types of degra
The climatological periods used in this study were based on the
dation sources was obtained. Kaiser-Meyer-Olkin (KMO) testing was
recommendations of the IPCC’s Fifth Assessment Report (IPCC AR5),
used to evaluate the quality of the PCA results obtained, thereby
which takes the period between 1986 and 2005 as the base or reference
establishing the compactness and reliability of the factor model (Salvati
period, 2016–2035 as the near future, 2046–2065 as mid-century and
et al., 2014). A complete statistical analysis was achieved using SPSS
2081–2100 as the end of the 21st century. EURO-CORDEX datasets for
software (IBM SPSS Statistics, 2016).
nine different models were used (Jacob et al., 2014), (Table 6), as well as
In addition, differential local Moran’s I (DLM) analysis was used in
two Representative Concentration Pathway (RCP) scenarios for GHG
this study, implemented in GeoDa (version 1.14) software. DLM analysis
emissions: RCP4.5 - a stabilisation scenario with a GHG emission peak in
is based on changes over time, i.e. on the difference between yt and yt− 1.
2040 and then declining, and RCP8.5 - a steady increase scenario.
Spatial and temporal dynamics of change in LDD patterns (ESA values)
EURO-CORDEX scenario simulations use the new Representative Con
were analysed for 1986–2005 (the base or reference period) and
centration Pathways (RCPs), (Moss et al., 2010). Unlike Special Report
2080–2100 (the end of the 21st century). These time series allowed
on Emissions Scenarios (SRES), RCP scenarios do not use socio-economic
change patterns to be identified and the spatial dynamics of LDD to be
scenarios, but assume pathways to various target radiative forcing levels
better interpreted. DLM analysis was tested using 999 permutations with
at the end of the 21st century (Jacob et al., 2014). A comparison be
a level of significance of 0.05. It should be borne in mind that this is the
tween the climate effects of the SRES and RCP scenarios indicates that
actual difference, and not the absolute difference between the two time
the A1B scenario leads to a global mean temperature increase of be
series of data; hence, positive changes will be seen as high and negative
tween 2.8 ◦ C and 4.2 ◦ C, which is close to RCP6 (between RCP4.5 and
ones as low (Anselin, 2018). The formula for DLM is as follows:
RCP8.5), (Rogelj et al., 2012). In this way, the applied methodology
( )∑
allows the comparison of results from other regions in the fight against ID, i = c yi, t − yi, t− 1 Wij(yj, t − yj, t− 1 ) (8)
the negative impacts of climate change (Vukovic et al., 2018). j
1
2.4. Statistical analysis where yi, t and yi, t − are the normalised values of the changes in the two
periods.
Raster data on CQI, SQI, VQI, MQI and SoQI was converted to a
vector format so the major drivers of LDD in Western Serbia could be 3. Results
analysed and estimated. Then, Principal Component Analysis (PCA) was
applied to a matrix composed of these five indicators, thus enabling the 3.1. LDD dynamics from the reference period (1986–2005)
grouping of factors with a similar influence, selected by the percentage
of variance. The absolute principal scores from PCA were further ana Seventeen LDD parameters including five climate parameters (CQI),
lysed using MLRA as independent factors over the quality indicators of four soil quality parameters (SQI), four vegetation quality parameters
the MEDALUS method as dependent factors (Čakmak et al., 2018), on (VQI), two management parameters (MQI) and two socio-economic
parameters (SoQI) were calculated to determine the map of ESAs in
Western Serbia using the geometric mean. The results showed that the
Table 5
majority of the study area can be classed as high quality in terms of CQI
Description of parameters used to obtain the Social Quality Index (SoQI).
(an average value of 1.14; SD 0.34; Fig. 1S, 2S), VQI (an average of 1.25;
Indicator Parameter Range Score References SD 0.31; Fig. 1S, 2S), MQI (an average of 1.2; SD 0.22; Fig. 1S, 2S) and
SoQI Old age index (%) >5 1 Tavares et al. SoQI (an average of 1.13, SD 0.09; Fig. 1S, 2S). In the case of SQI, the
5–10 1.4 (2015) study area was found to be of medium quality, with an average value of
10–20 1.5
1.38 (SD 0.25; Fig. 1S, 2S). ESAs were obtained by integrating all five
>20 2
Population density <4 1 Ferrara et al. quality indicators (Fig. 2). According to the results, those areas falling
(inhabitants/km2) 4 < 30 1.1 (2020) into the categories of C1, C2 and C3 are located in the low-lying,
30 < 80 1.2 northern part of the region, as well as in the hilly and mountainous
80 < 170 1.3 parts, mainly in the southeast and southwest of the study area. These
170 < 300 1.4
300 < 500 1.5
areas, with index values over 1.375, account for 2304.42 km2, repre
500 < 850 1.6 senting 37.53% of the total study area (Table 7). Fragile areas (F1, F2
850 < 1.7 and F3), with index values of 1.225–1.375, cover 2152.34 km2, or
1400 35.05% of the total area, mainly in the central part of the region on the
1400 < 1.8
slopes of mountains and in foothills. Areas at relatively low threat from
2000
2000 < 1.9 LDD (index values of 1.170–1.225) are also found in the central part of
2700 the region, covering 749.13 km2 (12.2%), and are located in wetter
≥ 2700 2 zones where there are large expanses under natural forests (Table 7).
5
Fig. 2. The distribution of ESAs in Western Serbia.
Non-threatened areas (index values < 1.170) account for 15.21% of the (77.49%) and partially by natural ones (21.07%). A similar trend was
total area or 934.11 km2 (Table 7). found for SoQI, where the impact of natural/anthropogenic factors is
Using PCA and MLRA, three components were identified that dominant with 62.64%, while the influence of natural factors is lower at
accounted for 77.9% of the variance (Table 1S). The first component, 20.46% (Fig. 4). The third component (CQI) (loading = 0.992 and 0.775;
accounting for 34.4% of the variance, comprises VQI and MQI (loading Table 2S), accounting for 20.1% of the variance, indicates a natural
= 0.927; Fig. 3, Table 2S), which indicates an anthropogenic influence influence (92.83%), meaning that only 19.8% of the impact on the
on LDD, i.e. it shows that the said component had the greatest impact on current state of LDD can be attributed to climatic parameters (Figs. 3 and
the current processes of LDD. Specifically, the VQI component was found 4). The Kaiser-Meyer-Olkin (KMO) value for the above variables was
to be 92.14% dependent on anthropogenic factors, while MQI depen 0.501 (Table 3S), which represents a value above the recommended cut-
dence on this factor was 93.03% (Fig. 4). Accounting for 23.5% of the off of 0.5, indicating the existence of a compact correlation and showing
variance, the second component, consisting of SQI and SoQI (loading = that PCA provides clear and reliable factors (Field, 2009).
0.755 and 0.775; Table 2S), indicates a natural/anthropogenic influence
on LDD. SQI was most influenced by natural/anthropogenic factors
6
Table 7 3.2. LDD dynamics in the future (2046–2065 and 2081–2100)

Summarised results of ESAs (Ferrara, 2005).
Class Sub-class Score range km2 % In this study, EURO-CORDEX datasets for nine different models and
two representative concentration pathways (RCP4.5 and RCP8.5) were
Non-affected N >=1.00<=1.170 934.11 15.21
Potential P >1.170<=1.225 749.13 12.2 used. Climate change in the near future (2016–2035) did not have a
Fragile (high) F1 >1.225<=1.275 577.13 9.4 significant impact on LDD processes when compared to the reference
Fragile (medium) F2 >1.275<=1.325 1002.27 16.32 period (1986–2005), and hence this period was not taken into further
Fragile (low) F3 >1.325<=1.375 572.94 9.33 consideration in this research. Fig. 5 shows the categorisation of soil
Critical (high) C1 575.05 9.37
sensitivity according to the RCP4.5 scenario for 2046–2065 and
>1.375<=1.425
Critical (medium) C2 >1.425<=1.530 1124.48 18.31
Critical (low) C3 > 1.530 604.89 9.85 2081–2100. It was simulated that climate change would result in a
Total 6140 100 25.6% expansion of critical areas (C1, C2 and C3) by 2065 and a 33.6%
Note: N – Areas which are not at threat or are at virtually no threat from
expansion by 2100 compared to the reference period (Table 8). Research
degradation; P – Areas with low sensitivity to land degradation; F – Areas with based on the RCP8.5 scenario indicates a continuous increase in GHG
medium sensitivity to land degradation; C – Areas with high sensitivity to land emissions. With this in mind, a dramatic increase in critical areas (C1, C2
degradation and C3) was simulated, amounting to an increase of 39.2% by 2065 and
as high as 51.7% by 2100 compared to the reference period (Fig. 5 and
Table 8).
3.2.1. Spatio-temporal cluster analysis

In this study, different types of LDD spatial patterns were identified
for the period 1986–2005 (reference period) and 2080–2100 (the end of
the 21st century) using DLM. Essentially, high positive values indicate a
high level of change (above average), while high negative values point
to a low level of change (below average). In this regard, Fig. 6 illustrates
the spatial and temporal distribution of LDD patterns, with two types of
spatio-temporal clusters predominant: High-High (HH) and Low-Low
(LL). Analysis based on the RCP4.5 scenario shows that the majority
of the HH spatio-temporal clusters are concentrated in the northern part
of the study region (Fig. 6a), where agricultural areas with intensive
production predominate and with a higher population concentration as
well. On the other hand, results obtained on the basis of the RCP8.5
scenario point to an overall increase in the number of HH spatio-
temporal clusters, particularly in the southern part of the study region
(Fig. 6b), which leads us to the conclusion that these regions are sus
ceptible to LDD processes in the reference period and that this trend will
continue up until the end of the century. Generally speaking, the pattern
of HH spatio-temporal clusters in Western Serbia is tied to those areas
with insufficiently good quality indicators like CQI, MQI, SQI and VQI.
Fig. 3. PCA loading plot for MEDALUS quality indicators. Conversely, in the central and northern parts of the region, there are a
significant number of locations with LL spatio-temporal clusters, i.e.
locations with negative changes surrounded by locations with similar
Fig. 4. Percentage distribution of possible types of degradation source.
7
Fig. 5. The spatial distribution of areas sensitive to LDD under the RCP4.5 and RCP8.5 scenarios a) 2046–2065 (RCP4.5); b) 2046–2065 (RCP8.5); c) 2081–2100
(RCP4.5); d) 2081–2100 (RCP8.5).
8
Table 8
Summarised results of LDD sensitive areas under the RCP4.5 and RCP8.5 scenarios.
Sub-class RCP4.5 2046–2065 RCP8.5 2046–2065 RCP4.5 2081–2100 RCP4.5 2081–2100
km2 % km2 % km2 % km2 %
N 553.5 9.01 57.98 0.94 52.69 0.86 28.99 0.47

P 593.67 9.67 409.62 6.67 428.68 6.98 58.3 0.95
F1 515.9 8.4 494.52 8.05 934.22 15.22 265.29 4.32
F2 941.97 15.34 786.17 12.8 750.6 12.22 232.85 3.79
F3 436.29 7.11 601.77 9.8 504.68 8.22 783.55 12.76
C1 1068.74 17.41 1322.85 21.54 1156.91 18.84 955.57 15.56
C2 1303.23 21.23 1606.94 26.17 1402.07 22.84 2395.72 39.02
C3 726.7 11.84 860.16 14.01 910.13 14.82 1419.73 23.12
Total 6140 100 6140 100 6140 100 6140 100
Note: N – Areas which are not at threat or are at virtually no threat from degradation; P – Areas with low sensitivity to land degradation; F – Areas with medium
sensitivity to land degradation; C – Areas with high sensitivity to land degradation.
Fig. 6. LDD differential cluster map: a) for the reference period (1986–2005) and future period (2080–2100) according to the RCP4.5 scenario; b) for the reference
period (1986–2005) and future period (2080–2100) according to the RCP8.5 scenario.
trends. These regions are characterised by significant areas of forest, as temporal evolution of LDD processes in Western Serbia.
well as zones of protected natural resources. Our study demonstrated that the major drivers of LDD in Western
Serbia are primarily related to anthropogenic activities, with VQI and
4. Discussion MQI having a 34.4% influence (PCA 1; see section 3.1.), which fully
supports earlier findings for the Mediterranean region (Feoli et al.,
4.1. Major drivers of LDD processes from the reference period 2003), as well as for the entirety of Europe (Ferrara et al., 2020). In
(1986–2005) Western Serbia, the dominant anthropogenic factors are VQI and MQI
(Figs. 3 and 4), which have been identified as the main drivers of LDD in
The MEDALUS method was designed primarily for the Mediterra many studies in other parts of Serbia (Kadović et al., 2016; Momirović
nean region and validated for the socio-economic conditions in Medi et al., 2019), as well as the Balkan Peninsula (Gül and Erşahin, 2019).
terranean countries and other parts of Europe; however, its conceptual Their negative impact in Western Serbia is related to the improper use
approach can be applied globally (Mairota et al., 1998; Ladisa et al., and management of forests, including deforestation, especially pro
2012). As LDD is a dynamic spatial and temporal process that is multi nounced in the southern part of the region, and also to the conversion of
dimensional in nature and impacts a variety of socio-economic and pastures and semi-natural vegetation into arable land, intensive soil
biophysical factors (Vogt et al., 2011; Kosmas et al., 2016), researchers erosion, and an overall lack of integrated natural resource management.
have devoted a lot of time over the last few decades to analysing LDD Although Western Serbia is rich in forests, its development is hindered
and its links to natural and anthropogenic factors on a global, regional, by various constraints related to the degradation of parts of the forested
national and local scale (Prăvălie et al., 2017; Kadović et al., 2016; area as a factor in diminished productivity and biological stability,
Momirović et al., 2019). With LDD representing a serious problem in insufficient area and age structure of forest vegetation, insufficient ac
Serbia as well, this study sought to assess and quantify the spatial and tivity in accessing appropriate forest development funds and inefficient
9
management of the total forest resource. A good example of well pre poorly bound and erodible. In addition, soils formed on serpentinite are
served natural forests are coniferous forests, mostly pine forests and to a very shallow and impermeable to water, which affects the formation of
lesser extent spruce forests, while other types of forests are devastated defiles, gullies, ravines and various other types of water erosion features
and degraded, having been reduced to smaller open formations (MEP, (Alexander and DuShey, 2011). Earlier research showed that it is pre
2012). Mountainous areas with a slightly wetter climate are charac cisely those areas with a geological substrate predominantly composed
terised by oak forests, which are mainly found on the steep slopes of of shales, flysch sandstones, serpentinites and limestones that are most
valleys and ravines prone to erosion processes. Much of the pasture of vulnerable to erosion processes (Tanasijević et al., 1966) and the study
Western Serbia is subject to the most diverse forms of erosion caused by area is one such area. Specifically, many areas on limestone formations
water, ice and snow, while traces of aeolian erosion are also not un in the Mediterranean region have already been ravaged and eroded,
common (MAFWM, 2018). Surface runoff is a frequent occurrence in the while vegetation cover has been completely devastated (Basso et al.,
sparse grassy areas, causing the shallow layer of soil to disappear from 2012). The accumulation capacity of these soils is insignificant, which is
entire slopes. Also, a lot of the pasture in the south of the study area why large swathes of pasture and meadow in these areas become dry
(Pešter Plateau, Nova Varoš, Prijepolje) has been used for livestock habitats, unsuitable for the good development of grass and other types of
grazing for hundreds of years, and it is only natural that such over vegetation. From the point of view of the spatial distribution and
grazing causes the land to be degraded (Papanastasis, 1998), the grassy organisation of agricultural production, these areas have a negative
surface becomes thinned out, and the floristic composition of the grass character in agro-ecological and socio-economic terms, as well as an
communities is modified. Given the mountainous character of this part underdeveloped infrastructure and higher rates of rural poverty and
of Serbia, the threat from potential natural disasters such as landslides, unemployment. These factors, especially poverty, tend to affect LDD,
fires, and earthquakes is pronounced (Dragićević et al., 2011). Fires are primarily due to the dependence on biomass-focused production (Pra
particularly frequent in areas with pine forests (Aleksić et al., 2009), kash et al., 2016). In addition, since Serbia, like many other countries in
after which soil degradation is accelerated due to changes in the physical transition, is facing a demographic decline, especially in rural areas
properties of the soil (Certini, 2005; Gül and Erşahin, 2019). Fires also (Manojlović et al., 2018), the impact of socio-economic factors on the
change the vegetation cover and thus significantly reduce the protection study area is negligible, primarily due to the low population density,
of soil from the erosion power of rain and, consequently, from soil particularly marked in the central and southern areas of the region. It
erosion (De Paola et al., 2013). Similar and somewhat more intense should be noted that the study area is a significant area in terms of
forms of LDD exist in the Mediterranean, but also in certain parts of tourists, and tourism, although not a direct cause of LDD, can have a
Central and Eastern Europe (EEA, 2012; Çolak and Sunar, 2020). significant impact on the environment (Ladisa et al., 2012), especially in
The influence of the natural/anthropogenic component was noted in relation to land use patterns and the availability of water resources
Western Serbia, consisting of SQI and SoQI, with its impact on LDD (Pérez-Trejo, 1994). For the reasons mentioned above, we can conclude
processes being 23.5% (PCA 2; see Section 3.1.). These are areas with that there has been a disturbance in the balance between anthropogenic
pronounced topography, i.e. steep slopes, with poor SQI and SoQI in and natural/anthropogenic factors in the study area, which has led to
dicators above all, indicating a strong correlation between land quality the intensification of LDD processes.
and demographic and socio-economic divisions (Salvati et al., 2017). The current impact of climate parameters (CQI) on LDD in this part of
The topography of this part of Serbia has a strong impact on land use and Western Serbia is 20.1% (PCA 3; see Section 3.1.), as this region is
factors contributing to soil erosion (Kostadinov et al., 2006). The slope exposed to air currents from the west, making precipitation slightly
of the terrain affects the infiltration rate in particular and accelerates more abundant here than in the rest of Serbia. Since a higher altitude
runoff, with the slope length increasing the transport of sediment region separates Western Serbia from the influence of the Adriatic Sea,
(Kadović, 1999). The steepest slopes often create landslides and ava the climate of this part of Serbia is slightly cooler. However, due to the
lanches, damaging vegetation and increasing soil erosion (Pérez-Trejo, existence of valleys, these influences converge and weaken here. The
1994; Ristić et al., 2011). central area and certain parts in the north and south of the region are in
In addition, in areas where ESA index values fell into the C1, C2 and the low risk zone, primarily because of the lower intensity of agriculture
C3 categories, the SOM content was reduced. It is believed that the low and the larger areas under forests. Specifically, these areas mainly
SOM content, which is frequently ≤2%, is a consequence of high sum extend into zones of protected natural resources, such as the Tara Na
mer temperatures, which stimulates the faster mineralisation of fallen tional Park, ‘Golija’ Nature Park, ‘Ovčar-Kablar Gorge’ (classed as ‘a
leaves, increases the occurrence of fires, and results in the domination of landscape of exceptional features’), ‘Uvac’ Special Nature Reserve, and
open shrubland (García-Ruiz et al., 2013). The replacement of forest ‘Zlatibor’ Nature Park.
vegetation with grassland in many areas of Western Serbia has caused
changes in morphology, which above all has affected the accumulation 4.2. Major drivers of the LDD process in the future (2046–2065 and
of humus and the creation of a humus horizon. As a major indicator 2081–2100)
affecting the physical, chemical and biological properties of soil, the lack
of SOM accelerates soil erosion and leads irreversibly to LDD (Salunkhe The spatial distribution of LDD in the future will depend on the
et al., 2018). Moreover, the deep ploughing of arable land is practised in interaction between various drivers, with climatic aspects having the
many agricultural areas, leading to the rapid mineralisation of labile greatest impact (Lu et al., 2019). Therefore, this study analysed future
components of SOM. Erodible texture classes that are more prevalent in CQI projections (according to the RCP4.5 and RCP8.5 scenarios), as it is
the southern part of the region affect the stability of soil structure, expected that Western Serbia will experience the negative effects that
reducing the rate of infiltration and available water capacity, as well as climate change can cause in the near future, and particularly in the
resistance to water and wind erosion (Le Bissonnais and Arrouays, forthcoming decades (Vukovic et al., 2018). Namely, it is estimated that
1997). For this reason, soil texture is one of the key factors that affect the global temperatures will increase by 1.4 ◦ C according to RCP4.5 and by
risk of LDD (Vieira et al., 2015). Similar to these findings, natural/ 1.8 ◦ C according to the RCP8.5 scenario for 2046–2065 compared to
anthropogenic factors are also among the major indicators of LDD across 1986–2005, while for the period from 2081 to 2100, it is estimated that
Europe (Imbrenda et al., 2013). In general, soils in the highest ESA class this increase will be 2.0 ◦ C and 4.4 ◦ C respectively (IPCC, 2013). In
(C1, C2 and C3) are shallow and disappear due to erosion processes. Serbia, analyses show that by the end of the century, global warming
This is especially the case with soils formed on soft limestones (marly will cause an increase of over 2.5 ◦ C in the mean temperature under the
and soft limestone) and serpentinites, which predominate in this region RCP4.5 stabilisation scenario and of over 5 ◦ C under the constant
(Pavlović et al., 2017). Although soils on limestone and serpentinite are RCP8.5 scenario with a decrease in summer precipitation (Vukovic
not sandy, in most cases they have a silty texture, which makes them et al., 2018). Beyond these findings, there is generally a broad scientific
10
consensus that Southeast Europe, and especially the Mediterranean re areas are expected to expand to about 50% of the global land area,
gion, can be considered one of the most critical areas for climate change whereas under the RCP 8.5 scenario, as much as 56% of the global land
in the 21st century (Giorgi, 2006; Diffenbaugh et al., 2007; Spinoni area will be affected by arid conditions (Huang et al., 2016). Research in
et al., 2017, 2018). In this regard, according to projections for both Europe has indicated an increase in aridity, especially in the Mediter
climate scenarios, changing climate patterns caused by climate change ranean region (Feng and Fu, 2013), with an increase in drought sensi
will intensify the process of LDD in Western Serbia, especially in the C1, tivity in the future, resulting in the extension of aridity zones towards
C2 and C3 categories, as is shown by the spatial grouping patterns of the central parts of Europe (Prăvălie et al., 2019). In this regard, European
HH spatio-temporal clusters following DLM analysis (Fig. 6; Table 8). Commission projections indicate that there is a high risk of LDD by the
Specifically, according to the RCP4.5 and RCP8.5 scenarios, tempera end of the century, especially in Spain, southern Italy, and Portugal, as
tures are expected to increase and precipitation to decrease, especially in well as in Southeast Europe (Spinoni et al., 2018). As the Balkan region
the summer months, while the AI and SPEI indicate an increase in aridity is climatically linked to Southern Europe (Giannakopoulos et al., 2009),
and drought. In addition, a reduction in the annual Rainfall Erosivity it is clear that the changes in the Balkans will follow the same trend as in
Factor is projected by the end of the century, primarily due to a decrease the Mediterranean region, primarily in terms of the expansion of arid
in precipitation. However, of all of the climate parameters mentioned areas (Spinoni et al., 2017; Carrão et al., 2017; Ciscar et al., 2018). At the
previously, the SPEI is the most reliable index for assessing the impact of same time, the spread of arid areas is likely to lead to humanitarian
climate change on the occurrence of drought, i.e. LDD (Manzano et al., crises associated with increased food insecurity, the emergence of new
2019; Spinoni et al., 2017; Lu et al., 2019), thus confirming the use diseases and famines, large population migration, and probably the
fulness of including the SPEI in the climatic parameters of the MEDALUS outbreak of conflicts over the remaining natural resources (Prăvălie
method. This recommendation could improve the reliability assessment et al., 2019; Perović et al., 2019).
of the methodology, primarily when analysing climate parameters. In
this regard, the SPEI in this study was calculated over a six-month period 5. Conclusions
(March-August), which allows for the monitoring of the dynamics of
change during the warmer part of the year, especially for those periods LDD is one of the most serious environmental problems globally,
when, on the one hand, maximum rainfall is expected in Western Serbia regionally, nationally and locally. Given the dynamic nature of this
(May-June-July) and when, on the other, the largest deficits are pro phenomenon, it is crucial to understand the underlying processes pri
jected (Vukovic et al., 2018). Hence, if the SPEI is <0, it indicates pro marily on a spatial and temporal scale.
cesses related to a dry climate, while if it is >0, it indicates wet climate The results of this study showed the spatial distribution of varying
processes (Gao et al., 2017). When looking at the SPEI for 2081–2100 for degrees of LDD susceptibility in the study area obtained through the
both scenarios (Fig. 3S), it is clear that the RCP4.5 scenario indicates MEDALUS method. Primarily, the results obtained for the reference
that much of Western Serbia will be affected by a dry climate (values < period (1986–2005) highlight the significant critical areas in the C1, C2
− 0.50) and moderately severe droughts, while RCP8.5 shows an even and C3 categories, covering 37.53% of the study area, while 27% con
more dramatic forecast, whereby nearly the entire area will be affected stitutes potentially safe areas. In terms of the major drivers of LDD, three
by severe droughts (values < − 1), including zones of extreme drought components were identified, using PCA and MLRA, which explain 77.9%
(values < − 2), especially in those areas that border Montenegro. of the variance. The first component, with a contribution rate of 34.4%,
The spatial distribution of the SPEI in Western Serbia coincides with indicated that anthropogenic factors have the greatest influence on the
the majority of studies which have analysed this index in Europe and occurrence of degradation processes in the study area. The second
which predict episodes of increased risk of fire (Cardil et al., 2019), component, explaining 23.5% of the variance, indicated a natural/
disturbance events in forest ecosystems (Tognetti et al., 2019), patterns anthropogenic influence, while the third component indicated that
of water deficit (Lu et al., 2019) and an overall increase in areas of natural (climate) parameters have a 20.1% impact on the current state of
drought (Carrão et al., 2017; Jacob et al., 2014; IPCC, 2013; Spinoni LDD.
et al., 2017, 2018; Lu et al., 2019). Moreover, it is important to mention However, with increasing temperatures and spatial and temporal
that the highest number of HH spatio-temporal clusters were mapped in changes in the precipitation regime due to the climate changes expected
the southwest and southeast of the region (Fig. 6), which coincides in Western Serbia, and all the other related components, these two
geographically with the lowest SPEI values (values < − 1), thus pointing climate parameters will have a negative impact on LDD. Specifically, by
to a link between ESAs and climate change, but also the impact of other the end of the 21st century, according to the RCP4.5 scenario, critical
factors, such as vegetation, topography and soil. areas (C1, C2 and C3) will expand by 33.6%, while under the RCP8.5
In this regard, it is predicted that projected climate components, scenario these areas will increase by 51.7% compared to the reference
based on the MEDALUS method, may influence LDD in Western Serbia period, with the effects of climate change being particularly marked in
(particularly in the southern part of the region) and beyond through the the southern part of Western Serbia, as indicated by the identification of
occurrence of meteorological drought, caused by the long-term reduc critical areas through DLM analysis. It can be concluded that even those
tion in precipitation that is expected in the study area. Therefore, these areas that might receive more precipitation in the future may become
processes will lead to a prolonged period of reduction in surface runoff drier due to increased evaporation and changes in the seasonal distri
and a lack of groundwater, causing hydrological drought. Furthermore, bution and intensity of precipitation. This is the conclusion most
in this region, there will be the possibility of pedological drought, caused strongly supported by analysis of the SPEI, which has the ability to
by the decrease in available water capacity, which is expected by the end identify the role of evapotranspiration and temperature variability in
of this century. Finally, agricultural or ecological drought, caused by low relation to drought assessment in the context of global climate change.
water availability at critical stages for crop growth, will certainly occur This, in our view, justifies the inclusion of the SPEI in the climatic pa
(Lal, 2012). rameters of the MEDALUS method.
On the other hand, the central part of the region is covered by forests Therefore, one of the main results of this study is a new methodo
and other vegetation with high evapotranspiration rates, which resulted logical approach in characterising and identifying major drivers, as well
in the occurrence of not only a large number of LL spatio-temporal as in isolating those spatio-temporal areas susceptible to LDD. In addi
clusters, but also the greater spatial distribution of Not Significant tion, this study is one of the first in the wider Balkan region that has
areas, which indicates that the changes coincide over time. allowed possible scenarios related to LDD and projected climate change
Although the results obtained in this study are rather worrying, they to be analysed. In this regard, future research should enable a more
are corroborated by several other studies which investigated the issue in detailed analysis of the effects of climate change, above all from the
similar conditions. Namely, under the milder RCP4.5 scenario, arid aspect of assessing the other indicators of the MEDALUS method.
11
Funding CLC, 2018. Corine Land Cover 2018 raster data. accessed on 7 September 2019. https:
//land.copernicus.eu/pan-european/corine-land-cover/clc2018.
This work was supported by the Ministry of Education, Science and
Çolak, E., Sunar, F., 2020. Evaluation of forest fire risk in the Mediterranean Turkish
Technological Development of Serbia, grant no. 451–03-68/2020–14/ forests: A case study of Menderes region, Izmir. Int. J. Disaster Risk Reduct. 45,
200007. 101479. https://doi.org/10.1016/j.ijdrr.2020.101479.
Lavado Contador, J.F., Schnabel, S., Gómez Gutiérrez, A., Pulido Fernández, M., 2009.
Assessing the environmental sensitivity to land degradation. A validation of the
CRediT authorship contribution statement MEDALUS method in SW Spain. EGU General Assembly Conference Abstracts. 11
(10730).
Veljko Perović: Conceptualization, Methodology, Data curation, Čakmak, D., Perović, V., Antić-Mladenović, S., Kresović, M., Saljnikov, E., Mitrović, M.,
Pavlović, P., 2018. Contamination, risk, and source apportionment of potentially
Software, Validation, Writing - original draft. Ratko Kadović: Super toxic microelements in river sediments and soil after extreme flooding in the
vision, Investigation, Conceptualization, Writing - original draft. Vla Kolubara River catchment in Western Serbia. J. Soils Sediments 18 (5), 1981–1993.
dimir Đurđević: Methodology, Software, Writing - original draft. https://doi.org/10.1007/s11368-017-1904-0.
De Paola, F., Ducci, D., Giugni, M., 2013. Desertification and erosion sensitivity. A case
Dragana Pavlović: Writing - review & editing, Visualization. Marija study in southern Italy: the Tusciano River catchment. Environ. Earth Sci. 70 (5),
Pavlović: Writing - review & editing, Visualization. Dragan Čakmak: 2179–2190. https://doi.org/10.1007/s12665-013-2294-2.
Investigation, Data curation, Software, Formal analysis. Miroslava Diffenbaugh, N.S., Pal, J.S., Giorgi, F., Gao, X., 2007. Heat stress intensifcation in the
Mediterranean climate change hotspot. Geophys. Res. Lett. 34 (11), L11706. https://
Mitrović: Writing - review & editing, Visualization. Pavle Pavlović: doi.org/10.1029/2007GL030000.
Supervision, Conceptualization, Funding acquisition. Dragićević, S., Filipović, D., Kostadinov, S., Ristić, R., Novković, I., Živković, N.,
Andjelković, G., Abolmasov, B., Šećerov, V., Djurdjić, S., 2011. Natural hazard
assessment for land-use planning in Serbia. Int J. Environ. Res. 5 (2), 371–380.
Declaration of Competing Interest https://doi.org/10.22059/ijer.2011.322.
Dragović, S., Mihailović, N., Gajić, B., 2008. Heavy metals in soils: distribution,
The authors declare that they have no known competing financial relationship with soil characteristics and radionuclides and multivariate assessment
of contamination sources. Chemosphere 72 (3), 491–549. https://doi.org/10.1016/j.
interests or personal relationships that could have appeared to influence chemosphere.2008.02.063.
the work reported in this paper. EEA, 2012. Climate change, impacts and vulnerability in Europe 2012, an Indicator
Based Report. EEA Report No. 12/2012, European Environment Agency (EEA),
Copenhagen, Denmark, 304 pp. https://www.eea.europa.eu/publications/climate-i
Appendix A. Supplementary data mpacts-and-vulnerability-2012.
FAO-UNEP, 1984. Provisional Methodology for the Assessment and Mapping of
Supplementary data to this article can be found online at https://doi. Desertification. FAO, Rome.
FAO-UNEP, 1997. Land Quality Indicators and Their Use in Sustainable Agriculture and
org/10.1016/j.ecolind.2021.107377. Rural Development. FAO, Rome.
Feng, S., Fu, Q., 2013. Expansion of global drylands under a warming climate. Atmos.
References Chem. Phys. 13, 10081–10094. https://www.atmos-chem-phys.net/13/10081/20
13/.
Feoli, E., Giacomich, P., Mignozzi, K., Oztürk, M., Scimone, M., 2003. Monitoring
Ait Lamqadem, A., Pradhan, B., Saber, H., Rahimi, A., 2018. Desertification sensitivity
desertification risk with an index integrating climatic and remotely-sensed data: an
analysis using MEDALUS model and GIS: A case study of the oases of Middle Draa
example from the coastal area of Turkey. Manage. Environ. Qual. 14 (1), 10–21.
Valley, Morocco. Sensors 18 (7), 2230. https://doi.org/10.3390/s18072230.
https://doi.org/10.1108/14777830310460351.
Aleksić, P., Krstić, M., Jančić, G., 2009. Forest fires – ecological and economic problem in
Ferrara, A., 2005. Expert system for evaluating the Environmental Sensitivity Index (ESI)
Serbia. Botanica Serbica, 33(2), 169–176. UDK 504.2:614.84(497.11).
of a local area. In: “DIS4ME: Desertification Indicator System for Mediterranean
Alexander, E.B., DuShey, J., 2011. Topographic and soil differences from peridotite to
Europe”, Brandt, J. (ed). ISSN: 1749-8996 https://esdac.jrc.ec.europa.eu/p
serpentinite. Geomorphology 135 (3–4), 271–276. https://doi.org/10.1016/j.
ublic_path/shared_folder/projects/DIS4ME/esi_jan_05/esi.htm.
geomorph.2011.02.007.
Ferrara, A., Kosmas, C., Salvati, L., Padula, A., Mancino, G., Nolè, A., 2020. Updating the
Anselin, L., 2018. A local indicator of multivariate spatial association, extending Geary’s
MEDALUS-ESA framework for worldwide land degradation and desertification
c. Geogr. Anal. https://doi.org/10.1111/gean.12164.
asessment. Land Degrad. Dev. 31, 1593–1608. https://doi.org/10.1002/ldr.3559.
Bakr, N., Weindorf, D.C., Bahnassy, M.H., El-Badawi, M.M., 2012. Multi-temporal
Field, A., 2009. Discovering Statistics Using SPSS, kkk SAGE publications, Ltd, London.
assessment of land sensitivity to desertification in a fragile agro-ecosystem:
Gao, X., Zhao, Q., Zhao, X., Wu, P., Pan, W., Gao, X., Sun, M., 2017. Temporal and spatial
environmental indicators. Ecol. Ind. 15 (1), 271–280. https://doi.org/10.1016/j.
evolution of the standardized precipitation evapotranspiration index (SPEI) in the
ecolind.2011.09.034.
Loess Plateau under climate change from 2001 to 2050. Sci. Total Environ. 595,
Basso, B., De Simone, L., Cammarano, D., Martin, E.C., Margiotta, S., Grace, P.R., Yeh, M.
191–200. https://doi.org/10.1016/j.scitotenv.2017.03.226.
L., Chou, T.Y., 2012. Evaluating responses to land degradation mitigation measures
García-Ruiz, J.M., Nadal-Romero, E., Lana-Renault, N., Beguería, S., 2013. Erosion in
in Southern Italy. International. J. Environ. Res. 6 (2), 367–380. https://doi.org/
Mediterranean landscapes: changes and future challenges. Geomorphology 198,
10.22059/ijer.2012.504.
20–36. https://doi.org/10.1016/j.geomorph.2013.05.023.
Boudjemline, F., Semar, A., 2018. Assessment and mapping of desertification sensitivity
Giannakopoulos, C., Le Sager, P., Bindi, M., Moriondo, M., Kostopoulou, E., Goodess, C.
with MEDALUS model and GIS - Case study: Basin of Hodna, Algeria. J. Water Land
M., 2009. Climatic changes and associated impacts in the Mediterranean resulting
Dev. 36 (1), 17–26. https://doi.org/10.2478/jwld-2018-0002.
from a 2◦ C global warming. Global Planet. Change 68 (3), 209–224. https://doi.org/
Budak, M., Günal, H., Çelik, İ., Yıldız, H., Acir, N., Acar, M., 2018. Environmental
10.1016/j.gloplacha.2009.06.001.
sensitivity to desertification in northern Mesopotamia; application of modified
Giorgi, F., 2006. Climate change hot-spots. Geophys. Res. Lett. 33, L08707. https://doi.
MEDALUS by using analytical hierarchy process. Arab. J. Geosci. 11, 481. https://
org/10.1029/2006GL025734.
doi.org/10.1007/s12517-018-3813-y.
GIWA, 2001. Global International Water Assessment. DPSIR Framework for State of
Cardil, A., Vega-García, C., Ascoli, D., Molina-Terrén, D.M., Silva, C.A., Rodrigues, M.,
Environment Reporting (Driving Forces-Pressures-State-Impacts-Responses),
2019. How does drought impact burned area in Mediterranean vegetation
European Environment Agency.
communities? Sci. Total Environ. 693, 133603 https://doi.org/10.1016/j.
Grimm, M., Jones, R.J.A., Rusco, E., Montanarella, L. (2003). Soil erosion risk assessment
scitotenv.2019.133603.
in Italy: a revised USLE approach. European Soil Bureau Research Report No.11,
Carrão, H., Naumann, G., Barbosa, P., 2017. Global projections of drought hazard in a
EUR 20677 EN, (2002), 28pp. Office for Official Publications of the European
warming climate: a prime for disaster risk management. Clim. Dyn. 50 (5–6),
Communities, Luxemburg.
2137–2155. https://doi.org/10.1007/s00382-017-3740-8.
Gül, E., Erşahin, S., 2019. Evaluating the desertification vulnerability of a semiarid
Certini, G., 2005. Effects of fire on properties of forest soils: A review. Oecologia 143 (1),
landscape under different land uses with the environmental sensitivity index. Land
1–10. https://doi.org/10.1007/s00442-004-1788-8.
Degrad. Dev. 30, 811–823. https://doi.org/10.1002/ldr.v30.710.1002/ldr.3269.
Cherlet, M., Hutchinson, C., Reynolds, J., Hill, J., Sommer, S., von Maltitz, G., 2018.
Huang, J., Yu, H., Guan, X., Wang, G., Guo, R., 2016. Accelerated dryland expansion
World Atlas of Desertification. Publication Office of the European Union,
under climate change. Nat. Clim. Change 6, 166–172. https://doi.org/10.1038/
Luxembourg.
nclimate2837.
CIESIN, 2018. Center for International Earth Science Information Network. Columbia
IBM SPSS Statistics for Windows version 21, 2016. SPSS Inc., Chicago, Ill., USA, about
University. Gridded Population of the World, Version 4 (GPWv4): Population
SPSS Inc. http://www.spss.com.hk/corpinfo/history.htm, (accessed on 10 February
Density, Revision 11. Palisades, NY: NASA Socioeconomic Data and Applications
2018).
Center (SEDAC). 10.7927/H49C6VHW.
Imbrenda, V., D’Emilio, M., Lanfredi, M., Ragosta, M., Simoniello, T., 2013. Indicators of
Ciscar, J.C., Iglesias, A., Feyen, L., Goodess, C.M., Szabó, L., Christensen, O.B., Nicholls,
land degradation vulnerability due to anthropic factors: Tools for an efficient
R., Amelung, B., Watkiss, P., Bosello, F., Dankers, R., Garrote, L., Hunt, A., Horrocks,
planning. In: Borruso, G., Bertazzon, S., Favretto, A., Murgante, B., Torre, C.M.
L., Moneo, M., Moreno, A., Pye, S., Quiroga, S., van Regemorter, D., Richards, J.,
(Eds.), Geographic information analysis for sustainable development and economic
Roson, R., Soria, A., 2018. Climate impacts in Europe: Final report of the JRC
planning: New technologies, Hershey: IGI Global. pp. 87–101. 10.4018/978-1-4666-
PESETA III project, EUR 29427 EN, Publications Office of the European Union,
4852-4.ch078.
Luxembourg, 2018. ISBN 978-92-79-97218-8. 10.2760/93257.
12
IPCC, 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Mrvić, V., Antonović, G., Čakmak, D., Perović, V., Maksimović, S., Saljnikov, E.,
Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Nikoloski, M., 2013. Pedological and pedogeochemical map of Serbia. In: The 1st
Change, Cambridge University Press, New York, USA. International Congress on Soil Science XIII National Congress in Soil Science,
Jacob, D., Petersen, J., Eggert, B., Alias, A., Christensen, O.B., Bouwer, L.M., Braun, A., September 23-26th, Belgrade, Serbia. UDC:631.455, pp. 93–105. http://data.sfb.bg.
Colette, A., Déqué, M., Georgievski, G., Georgopoulou, E., Gobiet, A., Menut, L., ac.rs/sftp/olivera.kosanin/book_of_proceedings.pdf.
Nikulin, G., Haensler, A., Hempelmann, N., Jones, C., Keuler, K., Kovats, S., Myoung-Jin, U., Yeonjoo, K., Daeryong, P., Kichul, J., Zhan, W., Mun, M.K.,
Kröner, N., Kotlarski, S., Kriegsmann, A., Martin, E., van Meijgaard, E., Moseley, C., Hongjoon, S., 2020. Impacts of potential evapotranspiration on drought phenomena
Pfeifer, S., Preuschmann, S., Radermacher, C., Radtke, K., Rechid, D., in different regions and climate zones. Sci. Total Environ. 703, 135590 https://doi.
Rounsevell, M., Samuelsson, P., Somot, S., Soussana, J.F., Teichmann, C., org/10.1016/j.scitotenv.2019.135590.
Valentini, R., Vautard, R., Weber, B., Yiou, P., 2014. EURO CORDEX: new high- Papanastasis, V.P., 1998. Livestock grazing in Mediterranean ecosystems: an historical
resolution climate change projections for European impact research. Reg. Environ. and policy perspective. In: Papanastasis, V.P., Peter, D. (Eds.), Ecological Basis of
Change 14, 563–578. https://doi.org/10.1007/s10113-013-0499-2. Livestock Grazing in Mediterranean Ecosystems. European Commission Science,
Kadović, R., 1999. Anti-erosion agrosystems. Faculty of Forestry, Belgrade (in Serbian). Research and Development, Bruxelles, pp. 5–9.
Kadović, R., Bohajar, Y., Perović, V., Simić, S., Todosijević, M., Tošić, S., Anđelić, M., Pavlović, P., Kostić, N., Karadžić, B., Mitrović, M., 2017. The Soils of Serbia. Springer
Mlađan, D., Dovezenski, U., 2016. Land sensitivity analysis of degradation using Science+Business Media, Dordrecht, 10.1007/978-94-017-8660-7.
MEDALUS model: case study of Deliblato Sands, Serbia. Arch. Environ. Prot. 42 (4), Pérez-Trejo, F., 1994. Desertification and land degradation in the European
114–124. https://doi.org/10.1515/aep-2016-0045. Mediterranean. Report EUR 14850 EN, European Commission.
Karavitis, C.A., Tsesmelis, D.E., Oikonomou, P.D., Kairis, O., Kosmas, C., Fassouli, V., Perović, V., Kadović, R., Djurdjević, V., Braunović, S., Čakmak, D., Mitrović, M.,
Ritsema, C., Hessel, R., Jetten, V., Moustakas, N., Todorovic, B., Skondras, N.A., Pavlović, P., 2019. Effects of changes in climate and land use on soil erosion: a case
Vasilakou, C.V., Alexandris, S., Kolokytha, E., Stamatakos, D.V., Stricevic, R., study of the Vranjska Valley, Serbia. Reg. Environ. Change. 19, 1035–1046. https://
Chatzigeorgiadis, E., Brandt, J., Geeson, N., Quaranta, G., 2020. A desertification doi.org/10.1007/s10113-018-1456-x.
risk assessment decision support tool (DRAST). Catena 187, 104413. https://doi. Prakash, S., Sharma, M.C., Kumar, R., Dhinwa, P.S., Sastry, K.L.N., Rajawat, A.S., 2016.
org/10.1016/j.catena.2019.104413. Mapping and assessing land degradation vulnerability in Kangra district using
Kolios, S., Mitrakos, S., Stylios, C., 2018. Detection of areas susceptible to land physical and socio-economic indicators. Spatial Inf. Res. 24, 733–744. https://doi.
degradation in Cyprus using remote sensed data and environmental quality indices. org/10.1007/s41324-016-0071-5.
Land Degrad. Dev. 29, 2338–2350. https://doi.org/10.1002/ldr.v29.810.1002/ Prăvălie, R., Patriche, C., Bandoc, G., 2017. Quantification of land degradation
ldr.3024. sensitivity areas in Southern and Central Southeastern Europe. New results based on
Kosmas, C., Kirkby, M., Geeson, N., 1999. The MEDALUS project: Mediterranean improving DISMED methodology with new climate data. Catena 158, 309–320.
desertification and land use. Manual on Key Indicators of Desertification and https://doi.org/10.1016/j.catena.2017.07.006.
Mapping Environmentally Sensitive Areas to Desertification European Commission, Prăvălie, R., Bandoc, G., Patriche, C., Sternberg, T., 2019. Recent changes in global
Project ENV4 CT 95 0119 (EUR 18882). https://cordis.europa.eu/project/rcn drylands: Evidences from two major aridity databases. Catena 178, 209–231.
/30864/factsheet/en. https://doi.org/10.1016/j.catena.2019.03.016.
Kosmas, C., Karamesouti, M., Kounalaki, K., Detsis, V., Vassiliou, P., Salvati, L., 2016. Prăvălie, R., Patriche, C., Săvulescu, I., Sîrodoev, I., Bandoc, G., Sfîcă, L., 2020. Spatial
Land degradation and long-term changes in agro-pastoral systems: An empirical assessment of land sensitivity to degradation across Romania. A quantitative
analysis of ecological resilience in Asteroussia-Crete (Greece). Catena 147, 196–204. approach based on the modified MEDALUS methodology. Catena 187, 104407.
https://doi.org/10.1016/j.catena.2016.07.018. https://doi.org/10.1016/j.catena.2019.104407.
Kostadinov, S., Zlatić, M., Dragović, N. (2006). Serbia and Montenegro. In: Boardman, J., QGIS Development Team, 2020. QGIS Geographic Information System. Open Source
Poesen, J. (Eds), Soil Erosion in Europe. Chichester, Wiley, pp. 271–277. Geospatial Foundation Project. http://qgis.osgeo.org.
Ladisa, G., Todorović, M., Liuzzi, G.T., 2012. A GIS-based approach for desertifcation risk Ristić, R., Marković, A., Radić, B., Nikić, Z., Vasiljević, N., Živković, N., Dragićević, S.,
assessment in Apulia region, SE Italy. Phys. Chem. Earth. 49, 103–113. https://doi. 2011. Environmental impacts in Serbian ski resorts. Carpathian J. Earth Environ. Sci.
org/10.1016/j.pce.2011.05.007. 6 (2), 125–134.
Lal, R., 2012. Climate change and soil degradation mitigation by sustainable Rogelj, J., Meinshausen, M., Knutti, R., 2012. Global warming under old and new
management of soils and other natural resources. Agric. Res. 1 (3), 199–212. https:// scenarios using IPCC climate sensitivity range estimates. Nat. Clim. Change. 2,
doi.org/10.1007/s40003-012-0031-9. 248–253. https://www.nature.com/articles/nclimate1385.
Le Bissonnais, Y., Arrouays, D., 1997. Aggregate stability and assessment of soil Salunkhe, S.S., Bera, A.K., Rao, S.S., Venkataraman, V.R., Raj, U., Murthy, Y.V.N.K.,
crustability and erodibility: II. Application to humic loamy soils with various organic 2018. Evaluation of indicators for desertification risk assessment in part of Thar
carbon contents. Eur. J. Soil Sci. 48, 39–48. https://doi.org/10.1111/j.1365- Desert Region of Rajasthan using geospatial techniques. J. Earth Syst. Sci. 127, 116.
2389.1997.tb00183.x. https://doi.org/10.1007/s12040-018-1016-2.
Liniger, H., van Lynden, G., Nachtergaele, F., Schwilch, G., 2008. A Questionnaire for Salvati, L., Bajocco, S., 2011. Land sensitivity to desertification across Italy: past, present,
Mapping Land Degradation and Sustainable Land Management. CDE/WOCAT, FAO/ and future. Appl. Geogr. 31, 223–231. https://www.sciencedirect.com/science/artic
LADA, ISRIC. file:///C:/Users/KORISN~1/AppData/Local/Temp/MapQuest_versio le/pii/S0143622810000524.
n_1-1.pdf. Salvati, L., Zitti, M., Carlucci, M., 2014. Territorial systems, regional disparities and
Lu, Y., Cai, H., Jiang, T., Sun, S., Wang, Y., Zhao, J., Yu, X., Sun, J., 2019. Assessment of sustainability: economic structure and soil degradation in Italy. Sustainability 6 (5),
global drought propensity and its impacts on agricultural water use in future climate 3086–3104. https://doi.org/10.3390/su6053086.
scenarios. Agric. For. Meteorol. 278, 107623. https://doi.org/10.1016/j. Salvati, L., Ferrara, C., Corona, P., 2015. Indirect validation of the environmental
agrformet.2019.107623. sensitive area index using soil degradation indicators: a country-scale approach.
MAFWM, 2018. Ministry of Agriculture, Forestry and Water Management of Republic of Ecol. Ind. 57, 360–365. https://www.sciencedirect.com/science/article/pii/S1470
Serbia. Degree of threat from erosion to agricultural land in the Zlatibor 160X15001685.
Administrative District. University of Belgrade, Faculty of Agriculture; University of Salvati, L., Zitti, M., Perini, L., 2016. Fifty years on: long-term patterns ofland sensitivity
Belgrade, Institute of Biological Research ‘Siniša Stanković’, Department of Ecology. to desertification in Italy. Land Degrad. Dev. 27 (2), 97–107. https://doi.org/
(in Serbian). 10.1002/ldr.2226.
Manojlović, S., Antić, M., Šantić, D., Sibinović, M., Carević, I., Srejić, T., 2018. Salvati, L., Tombolini, I., Gemmiti, R., Carlucci, M., Bajocco, S., Perini, L., Ferrara, A.,
Anthropogenic impact on erosion intensity: case study of rural areas of pirot and Colantoni, A. (2017). Complexity in action: Untangling latent relationships between
dimitrovgrad municipalities, Serbia. Sustainability 10 (3), 826. https://doi.org/ land quality, economic structures and socio-spatial patterns in Italy. PLOS ONE 12
10.3390/su10030826. (6), e0177853. 10.1371/journal.pone.0177853.
Manzano, A., Clemente, M.A., Morata, A., Luna, M.M., Beguería, S., Vicente-Serrano, S. Sepehr, A., Hassanli, A.M., Ekhtesasi, M.R., Jamali, J.B., 2007. Quantitative assessment
M., Martín, M.L., 2019. Analysis of the atmospheric circulation pattern effects over of desertification in south of Iran using MEDALUS method. Environ. Monit. Assess.
SPEI drought index in Spain. Atmos. Res. 230, 104630 https://doi.org/10.1016/j. 134 (1–3), 243–254. https://doi.org/10.1007/s10661-007-9613-6.
atmosres.2019.104630. Smiraglia, D., Tombolini, I., Canfora, L., Bajocco, S., Perini, L., Salvati, L., 2019. The
Mairota, P., Thornes, J.B., Geeson, N., 1998. Atlas of Mediterranean Environments in latent relationship between soil vulnerability to degradation and land fragmentation:
Europe: The Desertification Context. John Wiley & Sons, Chichester, UK, p. 205. A statistical analysis of landscape metrics in Italy, 1960–2010. Environ. Manage. 64,
MEP, 2012. Ministry of Environmental Protection of Republic of Serbia. Regional Spatial 154–165. https://doi.org/10.1007/s00267-019-01175-6.
Plan for the Zlatibor and Morava Administrative Districts. (in Serbian). Spinoni, J., Naumann, G., Vogt, J., Barbosa, P., Dosio, A., 2017. Will drought events
MEP, 2015. Ministry of Environmental Protection of Republic of Serbia. National Action become more frequent and severe in Europe? Int. J. Climatol. 38, 1718–1736. http
Plan to Ease the Consequences of Drought and Land Degradation. (in Serbian) s://rmets.onlinelibrary.wiley.com/doi/epdf/10.1002/joc.5291.
https://www.ekologija.gov.rs. Spinoni, J., Barbosa, P., Dosio, A., McCormick, N., Vogt, J., 2018. Is Europe at risk of
Momirović, N., Kadović, R., Perović, V., Marjanović, M., Baumgertel, A., 2019. Spatial desertification due to climate change? Geophysical Research Abstracts. 20:
assessment of the areas sensitive to degradation in the rural area of the municipality EGU2018-9557. 2018 EGU General Assembly.
Čukarica. Int. Soil Water Conserv. Res. 7 (1), 71–80. https://doi.org/10.1016/j. Symeonakis, E., Karathanasis, N., Koukoulas, S., Panagopoulos, G., 2016. Monitoring
iswcr.2018.12.004. sensitivity to land degradation and desertification with the environmentally
Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., van Vuuren, D.P., sensitive area index: the case of Lesvos Island. Land Degrad. Dev. 27 (6), 1562–1573.
Carter, T.R., Emori, S., Kainuma, M., Kram, T., Meehl, G.A., Mitchell, J.F.B., https://doi.org/10.1002/ldr.2285.
Nakicenovic, N., Riahi, K., Smith, S.J., Stouffer, R.J., Thomson, A.M., Weyant, J.P., Tanasijević, M.Đ., Antonović, A., Aleksić, Ž., Pavićević, N., Filipović, Đ., Spasojević, M.,
Wilbanks, T.J., 2010. The next generation of scenarios for climate change research 1966. Pedological cover of Western and North-western Serbia. Soil Research
and assessment. Nature 463, 747–756. https://doi.org/10.1038/nature08823. Institute in Topčider, Serbia.
13
Tavares, J.D.P., Baptista, I., Ferreira, A.J.D., Amiotte-Suchet, P., Coelho, C., Gomes, S., Vieira, R.M.S.P., Tomasella, J., Alvalá, R.C.S., Sestini, M.F., Affonso, A.G., Rodriguez, D.
Amoros, R., Dos Reis, E.A., Mendes, A.F., Costa, L., Bentub, J., Varela, L., 2015. A., Barbosa, A.A., Cunha, A.P.M.A., Valles, G., Crepani, E., de Oliveira, S.B.P., de
Assessment and mapping the sensitive areas to desertification in an insular Sahelian Souza, M.S.B., Calil, P.M., de Carvalho, M.A., Valeriano, D.M., Campello, F.C.B.,
mountain region case study of the Ribeira Seca watershed, Santiago Island, Cabo Santana, M.O., 2015. Identifying areas susceptible to desertification in the Brazilian
Verde. Catena 128, 214–223. https://doi.org/10.1016/j.catena.2014.10.005. northeast. Solid Earth 6, 347–360. https://doi.org/10.5194/se-6-347-2015, 2015.
Tognetti, R., Lasserre, B., Di Febbraro, M., Marchetti, M., 2019. Modeling regional Vogt, J.V., Safriel, U., Von Maltitz, G., Sokona, Y., Zougmore, R., Bastin, G., Hill, J.,
drought-stress indices for beech forests in Mediterranean mountains based on tree- 2011. Special issue on understanding dryland degradation trends. Land Degrad. Dev.
ring data. Agric. For. Meteorol. 265, 110–120. https://doi.org/10.1016/j. 22, 150–165. https://doi.org/10.1002/ldr.1075.
agrformet.2018.11.015. Vukovic, A., Vujadinovic, M., Rendulic, S., Djurdjevic, V., Ruml, M., Babic, V.,
UNCCD, 1994. United Nations Convention To Combat Desertification. Elaboration of an Popovic, D., 2018. Global warming impact on climate change in Serbia for the period
International Convention to combat desertification in countries experiencing serious 1961–2100. Thermal Science. 22 (6A), 2267–2280. https://doi.org/10.2298/
drought and/or desertification, particularly in Africa. United Nations. TSCI180411168V.
Uzuner, Ç., Dengiz, O., 2020. Desertification risk assessment in Turkey based on Yao, X., Yu, K., Deng, Y., Liu, J., Lai, Z., 2019. Spatial variability of soil organic carbon
environmentally sensitive areas. Ecol. Ind. 114, 106295. https://doi.org/10.1016/j. and total nitrogen in the hilly red soil region of Southern China. J. For. Res. https://
ecolind.2020.106295. doi.org/10.1007/s11676-019-01014-8.
Van der Knijff, J.M., Jones, R.J.A., Montanarella, L., 1999. Soil erosion risk assessment in Zhang, C., Wang, X., Li, J., Hua, T., 2020. Identifying the effect of climate change on
Italy, Luxemburg: European Commission Directorate. General JRC Joint Research desertification in northern China via trend analysis of potential evapotranspiration
Centre, Space Applications Institute. European Soil Bureau. and precipitation. Ecol. Ind. 112, 106141. https://doi.org/10.1016/j.
Vicente-Serrano, S.M., Beguería, S., López-Moreno, J.I., 2010. A multi-scalar drought ecolind.2020.106141.
index sensitive to global warming: the standardized precipitation evapotranspiration
index – SPEI. J. Clim. 23, 1696–1718. https://doi.org/10.1175/2009JCLI2909.1.
14

1 s2.0 S1470160X2100042X Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S1470160X2100042X Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S1470160X2100042X Main

Uploaded by

Copyright:

Available Formats

Ecological Indicators 123 (2021) 107377

Contents lists available at ScienceDirect

Major drivers of land degradation risk in Western Serbia: Current trends

1. Introduction coexistence, political stability, social equality, etc. (Kadović et al.,

Fig. 1. The location of the study area.

formula (Kosmas et al., 1999): the following formula:

ESAs = (CQI × SQI × VQI × MQI × SoQI) 1/5

where P is precipitation and PET the potential evapotranspiration, ob

Table 2 MQI = (intensity of land use × protection policies)1/2 (6)

intensity of human pressure on natural resources, while Oa emphasises Table 6

Fig. 2. The distribution of ESAs in Western Serbia.

Table 7 3.2. LDD dynamics in the future (2046–2065 and 2081–2100)

3.2.1. Spatio-temporal cluster analysis

Fig. 4. Percentage distribution of possible types of degradation source.

N 553.5 9.01 57.98 0.94 52.69 0.86 28.99 0.47

You might also like

1 s2.0 S1470160X2100042X Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S1470160X2100042X Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S1470160X2100042X Main

Uploaded by

Copyright:

Available Formats

Ecological Indicators 123 (2021) 107377

Contents lists available at ScienceDirect

Major drivers of land degradation risk in Western Serbia: Current trends

1. Introduction coexistence, political stability, social equality, etc. (Kadović et al.,

Fig. 1. The location of the study area.

formula (Kosmas et al., 1999): the following formula:

ESAs = (CQI × SQI × VQI × MQI × SoQI) 1/5

where P is precipitation and PET the potential evapotranspiration, ob­

Table 2 MQI = (intensity of land use × protection policies)1/2 (6)

intensity of human pressure on natural resources, while Oa emphasises Table 6

Fig. 2. The distribution of ESAs in Western Serbia.

Table 7 3.2. LDD dynamics in the future (2046–2065 and 2081–2100)

3.2.1. Spatio-temporal cluster analysis

Fig. 4. Percentage distribution of possible types of degradation source.

N 553.5 9.01 57.98 0.94 52.69 0.86 28.99 0.47

You might also like

where P is precipitation and PET the potential evapotranspiration, ob