Science of the Total Environment 485–486 (2014) 110–120

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Soil losses in rural watersheds with environmental land use conflicts F.A.L. Pacheco a,⁎, S.G.P. Varandas b, L.F. Sanches Fernandes b, R.F. Valle Junior b a b

Chemistry Research Centre, University of Trás-os-Montes and Alto Douro, Vila Real, Portugal Centre for Research and Technology of Agro-Environment and Biological Sciences, University of Trás-os-Montes and Alto Douro, Vila Real, Portugal

H I G H L I G H T S • Conceive environmental land use conflicts (LUC) in rural watersheds. • Investigate soil erosion in watersheds with LUC. • Predict soil erosion in the absence of LUC.

a r t i c l e

i n f o

Article history: Received 1 January 2014 Received in revised form 10 March 2014 Accepted 16 March 2014 Available online 3 April 2014 Editor: Simon Pollard Keywords: Soil loss Hydric erosion Environmental land use conflict USLE GIS

a b s t r a c t Soil losses were calculated in a rural watershed where environmental land use conflicts developed in the course of a progressive invasion of forest and pasture/forest lands by agriculture, especially vineyards. The hydrographic basin is located in the Douro region where the famous Port wine is produced (northern Portugal) and the soil losses were estimated by the Universal Soil Loss Equation (USLE) in combination with a Geographic Information System (GIS). Environmental land use conflicts were set up on the basis of land use and land capability maps, coded as follows: 1—agriculture, 2—pasture, 3—pasture/forest, and 4—forest. The difference between the codes of capability and use defines a conflict class, where a negative or nil value means no conflict and a positive i value means class i conflict. The reliability of soil loss estimates was tested by a check of these values against the frequency of stone wall instabilities in vineyard terraces, with good results. Using the USLE, the average soil loss (A) was estimated in A = 12.2 t · ha−1 · yr−1 and potential erosion risk areas were found to occupy 28.3% of the basin, defined where soil losses are larger than soil loss tolerances. Soil losses in no conflict regions (11.2 t · ha−1 · yr−1) were significantly different from those in class 2 (6.8 t · ha−1 · yr−1) and class 3 regions (21.3 t · ha−1 · yr−1) that in total occupy 2.62 km2 (14.3% of the basin). When simulating a scenario of no conflict across the entire basin, whereby land use in class 2 conflict regions is set up to permanent pastures and in class 3 conflict regions to pine forests, it was concluded that A = 0.95 t · ha−1 · yr−1 (class 2) or A = 9.8 t · ha−1 · yr−1 (class 3), which correspond to drops of 86% and 54% in soil loss relative to the actual values. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Among the factors explaining the intensity of soil erosion, plant cover and land uses are considered the most important, exceeding the influence of rainfall intensity and slope gradient (García-Ruiz, 2010; Kosmas et al., 1997; Thornes, 1990; Wainwright and Thornes, 2004). Estimates of soil losses under various plant cover and land use settings are reported in quite a number of studies (Cerdan et al., 2010; Durán-Zuazo et al., 2013; López-Vicente et al., 2013; Nunes et al., 2011; Tefera and Sterk, 2010; Vacca et al., 2000). Given the disparity of erosion rates among the different settings, it becomes evident that a change in the plant cover or land use of a region will inevitably lead to a soil loss increment or decrement in that region. ⁎ Corresponding author. Fax: +351 259 350480. E-mail address: [email protected] (F.A.L. Pacheco).

http://dx.doi.org/10.1016/j.scitotenv.2014.03.069 0048-9697/© 2014 Elsevier B.V. All rights reserved.

Cases of soil loss increment are frequently related to deforestation and substitution of forests by crops, meadows or permanent cultures such as orchards or vineyards. Usually, these land use changes result in gully development and shallow landslides that can increase the sediment load in rivers and ultimately contribute to the formation of new sedimentary structures including fluvial terraces, alluvial fans and deltas (Beguería et al., 2006; García-Ruiz and Valero-Garcés, 1998; Martínez-Casasnovas and Sánchez-Bosch, 2000). Cases of soil loss decrement are often related to farmland abandonment (Bellin et al., 2011; Ruiz-Flaño, 1993; Ruiz-Flaño et al., 1992), although abandoned farms have often been associated with important erosion processes shortly after desertion (Bellin et al., 2009; Lesschen et al., 2007). The decrement occurs in the long term and is associated with the recovery of the vegetation cover (dense forest or shrub) and the improvement of the chemical, physical and hydrological properties of soils (Bonet, 2004; García-Ruiz and Lana-Renault, 2011). Other cases

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

of soil loss decrement are related to reforestation of croplands (Wang et al., 2012). In rural watersheds, land uses are usually characterized by farmlands in the lowland valleys, where soils are thicker and more fertile; pastures for livestock production, vineyards and orchards in the mid-altitude valleys, with small forest spots in the adjacent hillsides; and continuous forests in the highlands. In traditional agrarian systems, these land uses are conforming to the land capability determined by an evaluation of soil characteristics such as depth or fertility and local environmental conditions such as topographic slope or water availability (Agroconsultores, Ltd., Coba, Ltd., 1991). But there are rural watersheds where the actual land use deviates from the most capable use, in which case an environmental land use conflict is generated with consequences on soil erosion intensity (Mello Filho, 1992; Valle Junior, 2008; Valle Junior et al., 2013, 2014). Notwithstanding the literature about the impact of land use changes on soil erosion is vast (Alkharabsheh et al., 2013; Ciampalini et al., 2012; Cotler and Ortega-Larrocea, 2006; Heckmann, 2014; Szilassi et al., 2006; Wijitkosum, 2012; Zokaib and Naser, 2011), only a few papers were specifically dedicated to the analysis of soil losses in places where environmental land use conflicts were developed (Haygarth and Ritz, 2009; Olarieta et al., 2008; Zucca et al., 2010). The purpose of this paper is to contribute with some new insights about this topic. Firstly, land use conflicts are investigated in a rural watershed, by comparing the maps of land use and land capability using the approach of Valle Junior (2008). Secondly, soil losses determined by the Universal Soil Loss Equation (Wischmeier and Smith, 1978) are compared among places that are inside or outside the conflict areas.

2. Study area The Meia Légua stream is a right margin tributary of the Douro River located in the southern limit of the Vila Real district, Trás-os-Montes and Alto Douro province, north of Portugal (Fig. 1). The stream is a 6.5 km long 9.2% inclined water course following NW–SE and NNE– SSW directions determined by the local fracture network. The hydrographic basin covers an area of approximately 18.3 km2, being

111

asymmetrical to a NNE–SSW longitudinal axis with the wider hillsides located at the right margin. Altitudes in the basin of Meia Légua stream range from 50 to 650 m and annual precipitation from 1240 to 1540 mm. As is typical for the SW European countries, precipitation regime in the area is characterized by long dry periods followed by heavy rain bursts. According to Brandão et al. (2001), the maximum precipitation (Pmax, in mm) in a period comprehended between 10 min and 24 h (D, in minutes) is described by the relationship Pmax = a × ln(D) − b, with a = 21.484 and b = −6.997 around the study area. This relationship provides an estimate of 163.2 mm for the local maximum precipitation in 24 h. The watershed is entirely shaped on Cambrian schists and graywackes, the alteration of which produced leptosols along the western boundary of the catchment and the NW–SE branch of the main valley, fluvisols along the NNE–SSW branch of the main valley, and anthrosols elsewhere (Fig. 2). Using a method by Smith and Stamey (1964), Catalão (2009) estimated the following erosion tolerances for these soil types (values in t · ha− 1 · yr− 1): 3.83 for leptosols, 12.93 for the fluvisols and 14.81 for the anthrosols. Based on soil characteristics such as depth or fertility as well as on environmental conditions such as topographic slope or water availability, land capability outside the main urban areas was assessed and defined by the codes AiPjFk, where A, P and F mean agriculture, pasturing and forestry, respectively, and the subscripts mean not adapted (0), well adapted (1), moderately adapted (2), poorly adapted (3) or conditionally adapted (4). According to this nomenclature, a parcel coded as A0P3F1 is not adapted to agriculture, is poorly adapted to pasturing and is well adapted to forestry. The most capable use has the code closer to 1. In the basin of Meia Légua stream, because specific soil types developed under specific environmental conditions—leptosols in the hilly regions, fluvisols along the valleys and anthrosols in the vineyards, the most capable uses (Fig. 2) reflect the soil characteristics, as follows: agriculture in the anthrosols, forestry in the leptosols and pasturing for livestock production mixed with forestry in the fluvisols (Agroconsultores, Ltd., Coba, Ltd., 1991). Land is mostly (75%) occupied by vineyards, the reminder being used for orchards, oliveyards, cropland, small spots of pine, eucalyptus and various deciduous forests; or taken by urban areas, roads and

Fig. 1. Geographic location, digital elevation model and precipitation contours in the hydrographic basin of Meia Légua stream.

112

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

Fig. 2. Soil and land suitability maps of the Meia Légua stream watershed.

other infrastructures, small ponds, and bare land (Fig. 3). Given the rugged topography, most vineyard plantations were accompanied by the construction of terraces supported by stone walls.

evaluates the long term average annual soil loss (A) by sheet and rill erosion and is defined by: A¼R·K ·L·S·C·P

3. Materials and methods 3.1. Software and digital database The modeling of soil losses and environmental land use conflicts at the scale of a stream watershed is facilitated through the use of a geographic information system (GIS), the reason why the ArcGIS version 10 software (ESRI, 2010) was used in this study. The cartographic and alphanumeric data required for running the soil loss and environmental land use conflict models were obtained from the information sources listed in Appendix A. The ArcGIS tools used to calculate the soil losses and delineate the conflict areas are enumerated in Appendix B. 3.2. Soil losses Soil losses were calculated by the Universal Soil Loss Equation (USLE) introduced by Wischmeier and Smith (1978), revised by Renard et al. (1997) and adopted in numerous studies (Wijitkosum, 2012; Martín-Fernández and Martínez-Núñez, 2011). The USLE

ð1Þ

where A R K L S C P

Soil loss per unit of area per unit of time (t ha−1 yr−1); Rainfall–runoff factor (MJ · mm · h−1 · ha−1 · yr−1); Soil erodibility factor (t · h · MJ−1 · mm−1); Slope length factor (dimensionless); Slope steepness factor (dimensionless); Cover-management factor (dimensionless); Support practice factor (dimensionless).

The rainfall–runoff factor is related to the kinetic energy of a storm and its maximum 30-minute intensity. Using probabilistic models of rainfall distribution, Ferro et al. (1991) showed that R can be approached by a power function:  b R ¼ a P 6;2

ð2Þ

where P6,2 is the average rainfall of 2-year return period 6-hour duration precipitation events and a and b are spatially dependent fitting

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

113

Fig. 3. Simplified land use map of the Meia Légua stream watershed.

parameters. For continental Portugal, Brandão et al. (2001) estimated a = 0.1437 and b = 2.2. The soil erodibility factor quantifies the inherent erodibility of a particular soil. It is a measure of the soil particles' susceptibility to detachment and transport by rainfall and runoff. Determination of erodibility is based on a diversity of data including the percentages in sand, silt, clay (texture) and organic matter, as well as the permeability and structure of the top soil layer. Based on these types of data, Pimenta (1998) estimated K for a number of soil types including those formed in the study area (values in t · h · MJ−1 · mm−1): 0.035 for the anthrosols, 0.027 for the fluvisols, and 0.040 for the leptosols and urban soils. The slope length factor is the ratio of expected soil loss to that observed for a field of 22.1 m in length, while the slope steepness factor is the ratio of expected soil loss to that observed for a field of 9% slope (inclination of 0.09 rad). In a formulation by Moore and Burch (1986) factors L and S in a point of a field are combined into a single parameter (LS) and calculated by:  LS ¼

0:4

λ 22:1

 

1:3

sinðβÞ 0:0899

ð3Þ

where λ is the distance from the onset of overland flow to the location where deposition occurs or when runoff enters a channel that is bigger than a rill, and β is the field inclination in radians; the numbers 0.4 and 0.3 are fitting parameters whereas 0.0899 is the sine value of 0.09 rad. The land cover/management factor is an index for the protective coverage of canopy and organic material in direct contact with the

ground. It is measured as the ratio of soil loss from land cropped under specific conditions to the corresponding loss from tilled land under clean-tilled continuous fallow conditions. High values of C factor occur on bare land while low values are found in the areas of dense forest or grain cover (Park et al., 2011). Based on the cartography of land occupation of continental Portugal Pimenta (1998) estimated values for the C factor as depicted in Table 1. The support practice factor is the ratio of expected soil loss to that observed for a field where soil conservation practices have been implemented. Wischmeier and Smith (1978) assigned values of P to various conservation practices, such as terraces or cultures along topographic contours, making them vary depending on the slope of Table 1 C factors attributed to different land uses and occupations. Adapted from Pimenta (1998). Land use/occupation

Factor C

Small ponds Roads and other infrastructures Urban areas Orchards Pine forests Oliveyards Other deciduous forests Eucalyptus forests Vineyards Cropland Bare land

0.005 0.01 0.01 0.05 0.05 0.1 0.1 0.2 0.2 0.3 0.3

114

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

the field. The conservation practices implemented in the study area are mostly the vineyard terraces (Fig. 3), the reason why the following P values were used in keeping with the hillside slope: class 2–7%, P = 0.1; class 8–12%, P = 0.12; class 13–18%, P = 0.16; class 19–24%, P = 0.18. All factors required for the USLE were prepared as raster layers with a cell size of 25 m. Then these layers were overlaid in ArcGIS and processed for soil loss calculation according to Eq. (1). 3.3. Environmental land use conflicts and soil losses In a wide range of environments, many authors confirmed that soil losses decrease exponentially as the percentage of vegetation cover increases (Nunes et al., 2011; and references therein). The percentage of vegetation cover is inherently related to land use, as clearly demonstrated by Cerdan et al. (2010) who quantified average soil losses for a number of land use types (values in t ha−1 year− 1): bare land (15.1), vineyards (12.2), orchards (11.8), arable land (4.4), shrubland (0.5), grassland (0.3), and forests (0.1). As soil loss is specifically associated to land use, any change to the use feeds back to the loss. In general, a change caused by deforestation for agricultural needs or grazing will intensify erosion (Szilassi et al., 2006; Wijitkosum, 2012; Zucca et al., 2010), while a change related to afforestation or conversion of crop lands to grass or shrubland will attenuate erosion (Alkharabsheh et al., 2013; Wijitkosum, 2012). However, the extent of soil degradation may depend on whether deforestation occurs in areas that are considered proper for agriculture or grazing, or away from them. In case the land capability evaluation of a forest spot is in favor of its use for agriculture or grazing, deforestation is considered the natural expansion of farming, or livestock production, with presumed modest consequences on soil erosion. Otherwise, deforestation will be viewed as the invasion of forests by farms with assumed significant impacts on soil loss. In this case, the land use change is also said to develop an environmental land use conflict. Because conflict areas may be critical as regards soil degradation, experts and policy makers should give preference to these areas as study sites in projects that aim to reduce efficiently soil loss by adequate land use planning. The concept of environmental land use conflict was introduced by Mello Filho (1992) and developed by Valle Junior (2008) and Valle Junior et al. (2013, 2014). According to these authors, a conflict may exist if the actual use deviates from a use standing on a capability evaluation (called natural use). To codify the natural (Code N) as well as the actual (Code A) land uses, Valle Junior et al. (2013) defined four general classes (Table 2): Code N = Code A = 1 for cropping agriculture, Code N = Code A = 2 for pasturing livestock, Code N = Code A = 3 for mosaic of natural pastures and forest spots, and Code N = Code A = 4 for forestry. Then, the environmental land use conflict was estimated by the equation:

Aj ≤ 0, with the negative values representing the areas with potential for a sustainable expansion of agriculture or grazing; b) areas suited for pasturing livestock (Code N = 2) but actually used for cropping agriculture (Code A = 1) are classified as Class 1 (minor) conflict areas; c) areas with potential for forestry (Code N = 4) or a mixed occupation by forests and pastures (Code N = 3) but occupied with a farm (Code A = 1) are referred to as Class 3 (major) or Class 2 (moderate) conflict areas, respectively. The approach of Valle Junior et al. (2013, 2014) was used in the present study, with the natural and actual land uses being represented by Figs. 2 and 3, respectively. As with the USLE, factors required for the conflict analysis were prepared as raster layers with a cell size of 25 m. Then these layers were overlaid in ArcGIS and used in Eq. (4) to calculate the environmental land use conflict. 4. Results and discussion 4.1. Soil losses and erosion risk The spatial distribution of USLE factors is illustrated in Fig. 4 and the map of soil losses is shown in Fig. 5. The average soil loss is 12.2 t · ha−1 · yr−1, which is a relatively high value but acceptable considering the dominance of the land use by vineyards. Besides, the losses are equivalent to regional-scale multi-decennial erosion rates (10.5 t · ha−1 · yr−1) estimated by stock unearthing–burying measurements in a vineyard of Languedoc, France (Paroissien et al., 2010). They are also within the range of decennial erosion rates estimated in vineyard hill slopes of Burgundy (2.6−12.3 t · ha−1 · yr−1), Monthélie, France, also based on vine-stock bio-markers (Brenot et al., 2008). In a region of craggy topography also dominated by vineyards but located in Spain, average erosion rates determined by Usón (1998) were larger approaching 22 t · ha−1 · yr−1. In approximately one third of the basin (34.6%) soil losses by hydric erosion are b 2 t · ha−1 · yr−1. These losses may be considered very low as they are smaller than the erosion tolerances calculated for the most sensible soil type (leptosols, with 3.83 t · ha−1 · yr− 1). However, there are sectors of the basin (23.6%) where the soil losses are N15 t · ha−1 · yr−1. In these cases the losses may be considered excessive as they are larger than the erosion tolerances calculated for the least sensible soil type (anthrosols, with 14.81 t · ha−1 · yr−1). By cross tabulating the spatial distribution of soil losses (Fig. 5) and soil types (Fig. 2), one finds that losses are larger than tolerances in 28.3% of the basin, and hence that soil is at risk of erosion in that area. Cases of erosion risk in vineyard fields were also reported in Navarre, Spain, where long-term erosion rates estimated by botanical benchmarks (30 t · ha−1 · yr−1) greatly exceeded even the most conservative soil loss tolerance thresholds, locally assumed to be 5–11 t · ha−1 · yr−1 (Casalí et al., 2009). 4.2. Soil erosion and hillside instability

Conflict Classi ¼ Code Ni −Code A j with 1≤i ≤n and 1≤ j ≤n

ð4Þ

where n is the number of classes (4). According to Eq. (4), a) the no conflict areas are represented by regions where Code Ni − Code

Table 2 Classification codes of natural (N) or actual (A) land uses selected for the analysis of conflicts. Land use

Classification code (A, N)

Agriculture Pastures for livestock production Pastures for livestock production/forestry Forestry

1 2 3 4

The conflict is calculated by Eq. (4).

A common consequence of soil erosion is hillside instability. This is especially evident in vineyard regions, where the necessity of increasing incomes, reducing manpower and increasing mechanization usually results in the substitution of traditional vineyards located on small terraces with stone walls by land leveling and the construction of new terraces. The new terraces are wide enough to accommodate machinery, although this requires an increase in the height between terraces, with consequent instability (Garcia-Ruíz, 2010). In the Penedès-Anoia vineyard region, where these new types of terraces were built, a rainstorm of about 90 mm in 24 h was enough to cause landslides, particularly on the lower third of the hill slopes. In some cases the upper soil levels were not preserved because they had to be dug out, mixing the fertile topsoil with the subsoil (Ramos and Martínez-Casasnovas, 2006a). The damage caused by landslides has an important economic cost. Martínez Casasnovas

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

115

Fig. 4. Spatial distribution of USLE factors (Eq. (1)) within the Meia Légua stream watershed.

and Ramos (2006) calculated that the costs of nutrient losses and damage to infrastructure caused by landslides and soil erosion in new terraced areas represent about 14% of annual incomes from the vineyards. In areas occupied by vineyards where terraces were built and supported by stone walls, signs of damage in the walls (e.g. deformation, repairs) can be used as an indication of hillside instability. In the attempt to check the estimates of soil loss presented in this study against an independent variable, a cross tabulation was made between the A values (Fig. 5) and the frequency of stone wall instabilities reported in Seixas et al. (2006) and also illustrated in Fig. 5. Firstly, an area (At) was computed that gathers all sectors of the basin where stone wall instabilities were observed (At = 5.6 km2). Secondly, the full range of soil loss values within At was assembled to form 9 categories, as illustrated in the X-axis of Fig. 6. Thirdly, the spatial coverage of each category (Ac) was calculated and the results were converted into percentages (Ac / At × 100). Fourthly, the number of stone wall instabilities (Ic) were computed within each soil loss category and the

results were converted into frequencies (Fi = Ic / It × 100, where It = 696 is the total number of instabilities reported in Seixas et al., 2006). Under the hypothesis of independence between soil erosion and hillside instability, the frequency of stone wall instabilities assigned to a class i of soil loss (Fi, in %) is expected to match the spatial coverage of that class (Ai, in %), i.e. τ = Fi / Ai ≈ 1; otherwise, τ N 1. The scenario of independence is ought to occur where soil losses are small and that of the scenario of explicit dependence is where the losses are large. The relationship between τ and A in the study area is shown in Fig. 6 and visibly confirms the expectations: a) for soil losses ≤ 10 t · ha− 1 · yr− 1, τ ≈ 1 and hence hillside instability is not explicitly linked to erosion at these levels of soil loss; b) however, for A N 10 t · ha−1 · yr−1, 1.5 ≤ τ ≤ 5 meaning that stone wall instabilities are 1.5 to 5 times larger than expected and that erosion might be the specific cause of hillside instability at these levels of soil loss. The consistency between τ and A values is supportive of a reliable estimation of soil losses in the hydrographic basin of Meia Légua stream using the USLE equation.

116

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

Fig. 5. Map of soil losses and of environmental land use conflicts.

4.3. Impact of environmental land use conflicts on soil losses The impact of land use changes on the erosion of the Penedès-Anoia (Catalonia, Spain) vineyard soils has been studied extensively by Martínez-Casasnovas and Sánchez-Bosch (2000). The analysis was

Fig. 6. Relationship between soil loss and stone wall instability. Additional information in the text.

based on the comparison of the estimated soil loss rates in the period just before the mechanization (1950s) and in the most recent past (1990s). Multi temporal data such as aerial photographs and digital terrain models, the revised universal soil loss equation and GIS analysis were used for that purpose. The results show a clear negative soil loss balance, with 12.6% of the agricultural land having experienced major negative changes. This negative balance was associated with the increase of the area dedicated to vineyards, the transformation of old traditional vineyard plantations to modern trained plantations and to the removing of conservation practices to adapt plots to crop mechanization. In this study, the purpose is to assess the impact of environmental land use conflicts on soil erosion. The areas of environmental land use conflict are represented by the hatched regions in Fig. 5, which occupy 2.62 km2 (14.3% of the basin). In these regions land use deviates from land capability and the question to pose is if this circumstance modifies soil erosion intensity. Likewise the analysis of soil loss in relation to stone wall instability (Section 4.2), under the hypothesis of independence between soil loss and land use conflict the average soil loss as well as the spatial coverage of soil loss classes should be similar regardless the region (no conflict, class 2 or class 3), otherwise each region should be characterized by a specific average soil loss and spatial coverage. In the hydrographic basin of Meia Légua stream, the average soil loss differs significantly among no conflict (11.2 t · ha−1 · yr−1), class 2 (6.8 t · ha−1 · yr−1) and class 3 (21.3 t · ha−1 · yr−1) conflict regions. The spatial coverage of soil loss classes is illustrated in Fig. 7. No conflict

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

Fig. 7. Spatial coverage of soil loss classes within the Meia Légua stream watershed. Additional information in the text.

and class 2 conflict regions show a similar coverage with the small-loss classes occupying larger areas of the basin and the large-loss classes those of smaller areas. These distributions are similar to the spatial coverage of total area (dash–dot line), which represents the average pattern of spatial coverage of each soil loss class. At odds with these distributions, the class 3 conflict regions show a relatively uniform spatial coverage of soil loss classes (≈ 0.5 km2 per class). In view of these results, the hypothesis of independence may be confirmed for the no conflict and class 2 conflict regions but certainly cannot extend to the class 3 conflict regions. Overall, Figs. 5 and 7 demonstrate that the class 3 conflict regions are homogeneous-coverage intensively eroded environments whereas the no conflict and class 2 conflict regions are heterogeneous-coverage moderately to gently eroded environments, respectively. The next step is to investigate the factors controlling erosion in the different regions. The average USLE factors in the no conflict, class 2 and class 3 conflict regions are depicted in Table 3. Factor R shows very little variation across the three regions (570 ≤ R ≤ 573 MJ · mm · h−1 · ha−1 · yr−1) meaning that soil loss estimates are not influenced by the climatic factor. In the case of factor K, it is clear that values are smaller in the class 2 conflict regions (0.02 t · h · MJ−1 · mm−1) than in the no conflict or class 3 conflict regions (0.03−0.04 t · h · MJ−1 · mm−1). The reason for this discrepancy is simple: the class 2 conflict regions are distributed along the southern branch of the main valley where the soil cover is composed of fluvisols, the soil type with the lowest K value (compare Figs. 2, 4 and 5). Likewise K, the LS values are also smaller in class 2 than in the other conflict regions because the class 2 regions are restricted to the southern branch of the main valley, where the average hillside gradient is gentle (7.9%). On the other hand, no conflict regions are characterized by smaller LS values (7.6) than the class 3 conflict regions (10.4), mostly because the latter regions tended to implement agricultural activities

117

along the NW–SE trending branch of Meia Légua stream where hillside slopes may exceed 40%. Likewise R, the C values also shows limited variation across the regions (0.16 ≤ C ≤ 20). This is because the entire basin was converted into a large agricultural field, essentially planted with vineyards and orchards, even in the areas where land capability set up for forestry or pasturing mixed with forestry is the most adequate land use. In case land capability had been respected soil losses could have been much smaller. For example, in the class 3 conflict regions the actual C value is on average 0.16 but could have been 0.05 if land had been occupied by pine forests (Table 1). In this case the average soil loss would drop from 21.3 t · ha − 1 · yr − 1 (the actual value) to 0.05 / 0.16 × 21.3 = 6.7 t · ha− 1 · yr− 1. Even in the class 2 conflict regions, where erosion is weak because terrain slopes are gentle and the bedrock is covered by fluvisols, soil losses could have been even smaller if the region had been occupied by permanent pastures, as determined by land capability. In this case C = 0.02 (Table 1), instead of the actual 0.2, and A = 0.02 / 0.2 × 6.8 = 0.68 t · ha− 1 · yr− 1. The support practices in the no conflict regions are characterized by an average P value of 0.5 which is smaller than the values in the class 2 (P = 0.7) or class 3 (P = 0.7) conflict regions. Apparently, the invasion of areas suited for forests or permanent pastures by agriculture was not accompanied by the proper soil conservation measures. If protection practices in the conflict regions were implemented as in the no conflict regions the soil losses would drop to A = 0.5 / 0.7 × 6.8 = 4.9 t · ha−1 · yr−1 in class 2 and to A = 0.5 / 0.7 × 21.3 = 15.2 t · ha− 1 · yr− 1 in class 3 conflict regions. It is hard to predict the overall impact of environmental land use conflicts on the intensity of soil erosion, given the lack of some data. For example, there is a limited possibility of verifying the impact of factor K because this would require specific information on texture, organic matter content, permeability and structure of the top soil layers across the hydrographic basin of Meia Légua stream, before and after the creation of land use conflicts, which is not available. It is also nearly impossible to check the impact of factor LS unless the values of λ and β in Eq. (3) could be estimated from a digital elevation model older than the period of vineyard and orchard plantations in the conflict areas. The consequences for soil loss of changing C and P in the course of a conflict creation were forecasted in the previous paragraph. Now, the attempt is to predict the return of conflict regions to an undisturbed condition, where: a) class 2 conflict regions are used for permanent pastures (C = 0.02) and class 3 conflict regions for pine forests (C = 0.05), respecting land capability; b) class 2 and class 3 conflict regions are set pristine and hence not influenced by support practices (P = 1). In this case, A = 0.95 t · ha−1 · yr−1 (class 2) or A = 9.8 t · ha−1 · yr− 1 (class 3), which correspond to drops of 86% and 54% in soil loss relative to present day values, respectively. In the first case soil losses would become rather low, which is comprehensible as class 2 conflict regions occupy watershed sectors where terrain slopes are minimal and soil types are relatively insensible to erosion. In the second case soil losses would become close to the average value of no conflict regions (A = 11.2 t · ha−1 · yr−1), as expected. 4.4. Causes and controls of soil losses in vineyards of SW Europe

Table 3 Average USLE factors in no conflict, class 2 conflict and class 3 conflict regions. Factor

R K LS C P

Unit

Region

−1

−1

−1

MJ · mm · h · ha · yr ton · h · MJ−1 · mm−1

No conflict

Class 2

Class 3

570 0.03 7.60 0.18 0.51

573 0.02 4.70 0.20 0.73

571 0.04 10.38 0.16 0.68

USLE factors were determined by Eq. (1), conflict regions by Eq. (4).

Soil loss is a significant environmental problem in semi-arid agricultural environments of SW Europe, especially in vineyards. The main reason for the high erosion rates in these vineyards is simple: the soil is almost bare for a large part of the year (Lasanta and Sobrón, 1988). Between November and April the plants lack leaves, and in May the foliage is still moderate. Even in summer, when the plants have reached maximum development, part of the soil is unprotected, unless straw has been added between the vine rows or herbicide has been applied without plowing. For this reason, vineyards provide little

118

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

protection for the soil under the SW Europe precipitation regime, since the autumn and spring rainfall occurs when the soil is almost bare. Although the average soil loss can be explained essentially by land use, the full range of values estimated for the basin of Meia Légua stream can only be interpreted if other factors are accounted for. According to García-Ruiz (2010) the other factors are rainfall intensity and slope gradient. Meyer and Martínez-Casasnovas (1999) estimated the probability of existence of gully erosion in vineyards of the PenedèsAnoia region, using the value of slope degree in a logistic regression model that yielded an overall accuracy of 84.6%. In La Rioja region, Spain, Arnáez et al. (2007) developed a linear equation involving rainfall intensity, soil resistance to drop detachment, slope gradient and gravel cover, which explained 74% of the measured soil loss. In the same region, Lasanta and Sobrón (1988) concluded that under similar gradient conditions, soil erosion in vineyards is controlled by land management practices and the grain size distribution of the soil. The former regulates runoff rates, and the latter explains the infiltration capacity and soil particle cohesion. In the basin of Meia Légua stream (Fig. 4a) it seems that terrain slope can help justify the noteworthiness of the variability of soil losses shown in Fig. 5. Since the A values are generally larger across the western margin of the basin where relief is more pronounced and terrain slopes are larger (compare Fig. 1 with Fig. 5), it can be postulated that soil losses in the hydrographic basin of Meia Légua stream are essentially determined by the local land use (mostly vineyards) being intensified or attenuated where hillside slopes are steep or gentle, respectively. Soil losses can be reduced through implementation of management practices. The following measures are proposed for the basin of Meia Légua stream. The first option is the superficial tillage using a rotary hoe. This can reduce significantly (4.5 times) total soil loss, as compared to no-tillage associated with herbicide application and leading to bare soil (Raclot et al., 2009). In vineyards subject to land preparation for mechanization, major soil movements are required. This rearrangement has enormous environmental implications not only due to changes in the landscape morphology but also due to soil degradation. The resulting cultivated soils are very poor in organic matter and highly susceptible to erosion, which reduces the possibilities of water intake as most of the rain is lost as runoff. In these cases, reduction of soil loss may be accomplished by a massive addition of organic wastes to promote aggregate formation increasing porosity and infiltration. This management practice has been implemented in vineyards of the Penedès-Anoia region, with a 26% raise in infiltration and a reduction of 20–43% in the concentration of sediment in runoff (Ramos and Martínez-Casasnovas, 2006b). However, the addition of large quantities of cattle manure to the Penedès-Anoia vineyards led to an increase in the nitrate concentrations of downstream surface waters. Eventually, the most effective management practice is the replacement of conventional tillage by soil treatments with cover crops. In a vineyard located in the Henares River basin southeast of Madrid, Spain, Ruiz-Colmenero et al. (2013) demonstrated that erosion plots under traditional tillage yielded substantially more erosion (5.88 t · ha− 1 · yr− 1) than when subject to treatments with cover crops of Brachypodium (0.78 t · ha− 1 · yr− 1) or Secale (1.27 t · ha−1 · yr− 1). Equivalent results were obtained by Novara et al. (2011) in a typical blanc wine grape irrigated vineyard located in southwestern Sicily, with different cover crops. 5. Conclusions Soil losses and environmental land use conflicts were assessed in a small watershed located in northern Portugal, called hydrographic basin of Meia Légua stream. Soil losses were calculated by the Universal Soil Loss Equation (USLE). Land use conflicts were set up on the basis of discrepancies between land use and land capability. The land use is dominated by vineyards. The average soil loss in the watershed is 12.2 t · ha−1 · yr− 1. This is comparable to regional-scale multi-

decennial erosion rates estimated in other SW European vineyard regions. In 28.3% of the basin soil losses exceed the tolerance thresholds of soil types, meaning that the soil is at risk of erosion in those areas. Water erosion is also the cause of hillside instability in the areas where soil loss N 10 t · ha−1 · yr−1. Environmental land use conflicts proved to be an important source of soil loss increments, especially when actual values were compared with counterparts predicted for a scenario of no conflict. Under a scenario of no conflict, soil losses were found to be 54% to 86% smaller than when land is used for purposes other than those determined by capability. Soil losses could be reduced through implementation of management practices, including the replacement of conventional tillage by soil treatments with cover crops.

Conflict of interest The authors have no conflicts of interest.

Acknowledgments The authors would like to thank the Coordination of Improvement of Higher Education Personnel (CAPES) for the scholarship Proc. no. 10297/12-0, the University of Trás-os-Montes and Alto Douro (UTAD) and the Center for the Research and Technology of AgroEnvironmental and Biological Sciences (CITAB) for technical support, and the Federal Institute of Triângulo Mineiro (IFTM), Brazil. As regards the first author, the research was funded by the strategic project of the Vila Real Chemistry Research Center (PEst-OE/QUI/UI0616/2014). As regards the other authors, the research was supported by European Union Funds (FEDER/COMPETE - Operational Competitiveness Programme) and by national funds (FCT - Portuguese Foundation for Science and Technology) under the project FCOMP-01-0124-FEDER022692.

Appendix A. Information sources The characterization of topographic relief was based on altimetry data produced at the scale of 1:50,000 and acquired from the Portuguese Geographic Institute (http://www.igeo.pt), more precisely topographic contours with a vertical equidistance of 25E m, randomly distributed altimetric points and geodetic vertices. These data are referenced in the Datum 73 coordinate system, Altimetric Datum of Cascais, Hayford ellipsoid and rectangular coordinates with Gauss projection. The mapping of soils was based on the 1:100,000 scale cartography produced in 1990 by the Trás-os-Montes and Alto Douro University in a partnership with the Agroconsultores and Coba Company. This cartography is available in digital format at the website of the Information Network of Emergency Systems (http://scrif.igeo.pt), under the heading “Project for the digital conversion of the soils, land use and land capability maps of northeast Trás-os-Montes”. The dataset is referenced in the 1924 International Datum coordinate system, Transverse Mercator projection, and rectangular coordinates with Gauss projection. The mapping of use and occupation of land was based on the 1:25,000 scale cartography produced in 2000 by the National Center for Geographic Information, resulting from the interpretation of aerial photographs, available at http://www.cnig.pt. The dataset is referenced in the same system as the altimetry data. The climate data (precipitation) was obtained in the form of listings from the Water Resources National Information System, being available at http://snirh.pt/. These listings include information on the location of the climatic stations, referenced in the Lisbon Datum coordinate system, ellipsoid of Hayford, Transverse Mercator projection, and rectangular coordinates with Gauss projection.

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

Appendix B. ArcGIS tools The map of soil losses (Fig. 5) was based on the production of raster maps for the USLE factors (Eq. (1)) succeeded by their spatial multiplication using the ArcGIS tool Spatial Analyst Tools N Map Algebra N Raster Calculator. The raster map of factor R (Fig. 4a) was based on discrete values of P6,2(Brandão et al., 2001) measured at several udometric stations located around the watershed, being predominantly influenced by two of them: the stations 10H/01 and 3 M/01 of the Water Resources National Information System (http://snirh.pt/), located to the SW and NE of the basin, respectively. The P6,2 values were used in Eq. (2) to calculate the R values, which were then interpolated over the entire basin using the tool Spatial Analyst Tools N Interpolation N IDW. Given the geographic position of the 10H/01 and 3 M/01 udometric stations relative to the basin, the spatial distribution of R resulted in a NW–SE trend. The mapping of factor K was based on the cartography of soils illustrated in Fig. 2. This cartographic data were contained in a polygon shapefile of ArcGIS, which was linked to an attribute table where factor K values (Section 3.2) were indicated. To produce the raster map of factor K (Fig. 4a) this shapefile was used as feature in the tool Conversion Tools N To Raster N Feature to Raster while the K values were used as conversion attribute. The raster map of factor LS (Fig. 4b) was produced by application of Eq. (3) to every pixel in the raster files of parameters λ and β. Parameter λ was equated to the number of pixels accumulating flow into pixel j (λj) multiplied by the pixel size. The raster file of λj resulted from a sequential application of two hydrologic tools to a Digital Elevation Model (DEM) of the basin (Fig. 1), namely the tool Spatial Analyst Tools N Hydrology N Flow Direction and the tool Spatial Analyst Tools N Hydrology N Flow Accumulation. The raster file β = sin(tan(d / 100) was computed by the raster calculator, where sin and tan are trigonometric functions and d is a terrain slope raster map derived from the DEM and computed by the tool Spatial Analyst Tools N Surface N Slope. To produce the raster map of factor C (Fig. 4c) one followed the approach used for factor K. In this case, the shapefile contained the cartography of land use and occupation (Fig. 3) and the attribute table enumerated the factor C values (Table 1). Finally, the map of factor P (Fig. 4d) was produced in four consecutive steps: a) the attributes called flag and P1 were added to the polygon shapefile containing the vineyard terraces (Fig. 4). The attribute flag distinguishes the areas covered by conservation practices (flag = 1) from the other areas (flag = 0). The attribute P1 is the support practice factor in areas not covered by conservation practices, i.e. P1 = 1 where flag = 0; b) two raster files were produced, one for flag and the other for P1, using the conversion to raster tool; c) the slope raster map (d) was reclassified into a P2 raster file considering the relationship set up between slope class and support practice factor (Section 3.2), using the tool Spatial Analyst Tools N Reclass N Reclassify; and d) the map of factor P was calculated as P = P1 + flag × P2 using the raster calculator. The map of environmental land use conflicts (Fig. 5) was based on the production of raster maps for natural (Fig. 2) and actual (Fig. 3) land uses, with the uses reclassified according to the codes listed in Table 2, succeeded by their subtraction (Eq. (4)).

References Agroconsultores, Ltd., Coba, Ltd. Carta de solos, carta do uso actual da terra e carta de aptidão da terra do nordeste de Portugal. Projecto de desenvolvimento rural integrado de Trás-os-Montes e Alto Douro (PDRITM). Universidade de Trás-osMontes e Alto Douro; 1991. Alkharabsheh MM, Alexandridis TK, Bilas G, Misopolinos N, Silleos N. Impact of land cover change on soil erosion hazard in northern Jordan using remote sensing and GIS. Procedia Environ Sci 2013;19:912–21. Arnáez J, Lasanta T, Ruiz Flaño P, Ortigosa L. Factors affecting runoff and erosion under simulated rainfall in Mediterranean vineyards. Soil Tillage Res 2007;93: 324–34. Beguería S, López-Moreno JI, Gómez-Villar A, Rubio V, Lana-Renault N, García-Ruiz JM. Fluvial adjustments to soil erosion and plant cover changes in the Central Spanish Pyrenees. Geogr Ann A 2006;88(3):177–86.

119

Bellin N, VanWesemael B, Meerkerk A, Vanacker V, Barberá GG. Abandonment of soil and water conservation structures in Mediterranean ecosystems: a case study from southeast Spain. Catena 2009;76:114–21. Bellin N, Vanacker V, van Wesemael B, Solé-Benet A, Bakker MM. Natural and anthropogenic controls on soil erosion in the Internal Betic Cordillera (southeast Spain). Catena 2011;87:190–200. Bonet A. Secondary succession of semi-arid Mediterranean old-fields in southeastern Spain: insights for conservation and restoration of degraded lands. J Arid Environ 2004;56:213–33. Brandão C, Rodrigues R, Costa J. Análise de fenómenos extremos, precipitações intensas em Portugal Continental. Lisboa: Direcção de Serviços de Recursos Hídricos do Instituto da Água; 2001. Brenot J, Quiquerez A, Petit C, Garcia J-P. Erosion rates and sediment budgets in vineyards at 1-m resolution based on stock unearthing (Burgundy, France). Geomorphology 2008;100:345–55. Casalí J, Giménez R, De Santisteban L, Álvarez-Mozos J, Mena J, Del Valle de Lersundi J. Determination of long-term erosion rates in vineyards of Navarre (Spain) using botanical benchmarks. Catena 2009;78(1):12–9. Catalão MEC. Erosão hídrica na bacia hidrográfica da Ribeira da Meia Légua. MSc Thesis. Vila Real, Portugal: Trás-os-Montes and Alto Douro University; 2009. Cerdan O, Govers G, Le Bissonnais Y, Van Oost K, Poesen J, Saby N, et al. Rates and spatial variations of soil erosion in Europe: a study based on erosion plot data. Geomorphology 2010;122:167–77. Ciampalini R, Billi P, Ferrari G, Borselli L, Follain S. Soil erosion induced by land use changes as determined by plough marks and field evidence in the Aksum area (Ethiopia). Agric Ecosyst Environ 2012;146:197–208. Cotler H, Ortega-Larrocea MP. Effects of land use on soil erosion in a tropical dry forest ecosystem, Chamela watershed, Mexico. Catena 2006;65:107–17. Durán-Zuazo VH, Francia-Martínez JR, García-Tejero I, Tavira SC. Implications of landcover types for soil erosion on semiarid mountain slopes: towards sustainable land use in problematic landscapes. Acta Ecol Sin 2013;33:272–81. ESRI, ArcMap (version 10). 380 New York Street. USA: Redlands; 2010. Ferro V, Giordano G, Iovino M. Isoerosivity and erosion risk map for Sicily. Bull Hydrol-J Sci Hydrol 1991;36(6):549–64. García-Ruiz JM. The effects of land uses on soil erosion in Spain: a review. Catena 2010; 81:1–11. García-Ruiz JM, Lana-Renault N. Hydrological and erosive consequences of farmland abandonment in Europe, with special reference to the Mediterranean region—a review. Agric Ecosyst Environ 2011;140:317–38. García-Ruiz JM, Valero-Garcés B. Historical geomorphic processes and human activities in the Central Spanish Pyrenees. Mt Res Dev 1998;18(4):309–20. Haygarth PM, Ritz K. The future of soils and land use in the UK: soil systems for the provision of land-based ecosystem services. Land Use Policy 2009;26S:S187–97. Heckmann M. Farmers, smelters and caravans: two thousand years of land use and soil erosion in North Pare, NE Tanzania. Catena 2014;113:187–201. Kosmas C, Danalatos N, Cammeraat LH, Chabart M, Diamantopoulos J, Farand R, et al. The effect of land use on runoff and soil erosion rates under Mediterranean conditions. Catena 1997;29:45–59. Lasanta T, Sobrón I. Influencia de las prácticas de laboreo en la evolución hidromorfológica de suelos cultivados con viñedo. Cuad Investig Geográfica 1988;14:81–97. Lesschen JP, Kok K, Verburg PH, Cammeraat LH. Identification of vulnerable areas for gully erosion under different scenarios of land abandonment in Southeast Spain. Catena 2007;71:110–21. López-Vicente M, Poesen J, Navas A, Gaspar L. Predicting runoff and sediment connectivity and soil erosion by water for different land use scenarios in the Spanish Pre-Pyrenees. Catena 2013;102:62–73. Martínez-Casasnovas JA, Ramos MC. The cost of soil erosion in vineyard fields in the Penedès-Anoia Region (NE Spain). Catena 2006;68:194–9. Martínez-Casasnovas JA, Sánchez-Bosch I. Impact assessment of changes in land use/ conservation practices on soil erosion in the Penedès-Anoia vineyard region (NE Spain). Soil Tillage Res 2000;57:101–6. Martín-Fernández L, Martínez-Núñez M. An empirical approach to estimate soil erosion risk in Spain. Sci Total Environ 2011;409:3114–23. Mello Filho JA. Direcionamento da ocupação da terra, pelo diagnóstico físicoconservacionista, das micro-bacias hidrográficas dos rios Alambari e Sesmaria, em Resende, RJ. Santa Maria: Federal University of Santa Maria; 1992 [Graduation thesis, 50 pp.]. Meyer A, Martínez-Casasnovas JA. Prediction of existing gully erosion in vineyard parcels of the NE Spain: a logistic modeling approach. Soil Tillage Res 1999;50:319–31. Moore ID, Burch GJ. Physical basis of the length-slope factor in the universal soil loss equation. Soil Sci Am J 1986;50:1294–8. Novara A, Gristina L, Saladino SS, Santoro A, Cerdà A. Soil erosion assessment on tillage and alternative soil managements in a Sicilian vineyard. Soil Tillage Res 2011;117:140–7. Nunes AN, Almeida AC, Coelho COA. Impacts of land use and cover type on runoff and soil erosion in a marginal area of Portugal. Appl Geogr 2011;31:687–99. Olarieta JR, Rodríguez-Valle FL, Tello E. Preserving and destroying soils, transforming landscapes: soils and land-use changes in the Valle's County (Catalunya, Spain) 1853–2004. Land Use Policy 2008;25:474–84. Park S, Oh C, Jeon S, Jung H, Choi C. Soil erosion risk in Korean watersheds, assessed using the revised universal soil loss equation. J Hydrol 2011;399:263–73. Paroissien J-B, Lagacherie P, Bissonnais Y. A regional-scale study of multi-decennial erosion of vineyard fields using vine-stock unearthing–burying measurements. Catena 2010;82:159–68. Pimenta MT. Directrizes para a aplicação da equação universal de perda dos solos em SIG: factor de cultura C e factor de erodibilidade do Solo K. Lisboa: Direcção de Serviços de Recursos Hídricos do Instituo da Água; 1998.

120

F.A.L. Pacheco et al. / Science of the Total Environment 485–486 (2014) 110–120

Raclot D, Bissonnais Y, Louchart X, Andrieux P, Moussa R, Voltz M. Soil tillage and scale effects on erosion from fields to catchment in a Mediterranean vineyard area. Agr Ecosyst Environ 2009;134:201–10. Ramos MC, Martínez-Casasnovas JA. Impact of land levelling on soil moisture and runoff variability in vineyards under different rainfall distributions in a Mediterranean climate and its influence on crop productivity. J Hydrol 2006a;321:131–46. Ramos MC, Martínez-Casasnovas JA. Nutrient losses by runoff in vineyards of the Mediterranean Alt Penedès region (NE Spain). Agr Ecosyst Environ 2006b; 113(1–4):356–63. Renard KG, Foster GR, Weesies GA, McCool DK, Yoder DC. Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). Agriculture Handbook. U.S. Department of Agriculture. Agricultural Research Service; 1997. p. 703. Ruiz-Colmenero M, Bienes R, Eldridge DJ, Marques MJ. Vegetation cover reduces erosion and enhances soil organic carbon in a vineyard in the central Spain. Catena 2013; 104:153–60. Ruiz-Flaño P. Procesos de erosión en campos abandonados del Pirineo. Logroño: Geoforma Ediciones; 1993. Ruiz-Flaño P, García-Ruiz JM, Ortigosa L. Geomorphological evolution of abandoned fields: a case study in the Central Pyrenees. Catena 1992;19:301–8. Seixas A, Bateira C, Hermenegildo C, Soares L, Pereira S. Definição de critérios de risco de ocorrência de movimentos de vertente na bacia hidrográfica da ribeira da Meia Légua. Jornadas sobre Terraços e Prevenção de Riscos Naturais, Parque de Tecnologias Ambientais de Maiorca, Palma de Maiorca, 14−16 September; 2006; 2006. p. 95–104. Smith RM, Stamey WL. How to establish erosion tolerances. J Soil Water Conserv 1964; 19(3):110–1. Szilassi P, Jordan G, van Rompaey A, Csillag G. Impacts of historical land use changes on erosion and agricultural soil properties in the Kali Basin at Lake Balaton, Hungary. Catena 2006;68:96–108. Tefera B, Sterk G. Land management, erosion problems and soil and water conservation in Fincha'a watershed, western Ethiopia. Land Use Policy 2010;27:1027–37.

Thornes JB. The interaction of erosional and vegetational dynamics in land degradation: spatial outcomes. In: Thornes JB, editor. Vegetation and erosion. Chichester: Processes and Environments; 1990. p. 41–53. Usón A. Medidas de control de la erosión en suelos de viña de las comarcas Anoia-Alt Penedés (Barcelona): efectividad y viabilidad (PhD Thesis) Spain: University of Lleida; 1998. Vacca A, Loddo S, Ollesch G, Puddu R, Serra G, Tomasi D, et al. Measurement of runoff and soil erosion in three areas under different land use in Sardinia (Italy). Catena 2000; 40:69–92. Valle Junior RF. Diagnóstico de áreas de risco de erosão e conflito de uso dos solos na bacia do rio Uberaba. PhD thesis in Agronomy. Jaboticabal: State University of São Paulo, Faculty of Agrarian and Veterinary Sciences; 2008 (222 pp.). Valle Junior RF, Galbiatti JA, Pissarra TCT, Martins Filho MV. Diagnóstico do conflito de uso e ocupação do solo na Bacia do Rio Uberaba. Glob Sci Technol 2013;6(1):40–52. Valle Junior RF, Varandas SGP, Sanches Fernandes LF, Pacheco FAL. Environmental land use conflicts: a threat to soil conservation. Land Use Policy 2014. [in revision]. Wainwright J, Thornes JB. Environmental issues in the Mediterranean. Processes and perspectives from the past and present. London: Routledge; 2004. Wang J, Chen Y, Shao X, Zhang Y, Cao Y. Land-use changes and policy dimension driving forces in China: present, trend and future. Land Use Policy 2012;29:737–49. Wijitkosum S. Impacts of land use changes on soil erosion in Pa Deng Sub-district, adjacent area of Kaeng Krachan National Park Thailand. Soil Water Res 2012;7(1): 10–7. Wischmeier WH, Smith DD. Predicting rainfall erosion losses—a guide to conservation planning. USDA agriculture handbook no. 537. Washington, DC: US Government Printing Office; 1978. Zokaib S, Naser GH. Impacts of land uses on runoff and soil erosion: a case study in Hilkot watershed, Pakistan. Int J Sediment Res 2011;26:343–52. Zucca C, Canu A, Previtali F. Soil degradation by land use change in an agropastoral area in Sardinia (Italy). Catena 2010;83:46–54.