Open access peer-reviewed chapter

Vulnerability of Soil and Water in Mediterranean Agro-Forestry Systems

Written By

António Canatário Duarte

Submitted: November 29th, 2016 Reviewed: June 9th, 2017 Published: November 8th, 2017

DOI: 10.5772/intechopen.70094

Chapter metrics overview

1,185 Chapter Downloads

View Full Metrics


The study watershed is located within the Idanha Irrigation Scheme, Portugal. A hydrological and water quality station was installed at the outlet of the catchment. The AnnAGNPS model was applied in this study, and afterwards it was calibrated and validated to the conditions in the study catchment. The antecedent soil moisture conditions play an important role for rapid runoff and flash flooding. Relative Water Supply (RWS) is always bellow the unity value and sometimes below the value 0.5. Sometimes in very dry years, like the year 2004-2005 (302 mm), a runoff coefficient is equal to 0.40. Spatial distribution of runoff was primarily influenced by topography and soil management, which is common in Mediterranean agricultural systems, namely in grain crop systems such as oats and wheat. The simulation of spatial distribution of nitrate load shows a dependence of the spatial distribution of runoff, due to its high solubility. Spatial distribution of soil erosion by water indicates that the process does not depend directly on the runoff distribution in the catchment. Therefore, soil erosion is greatly influenced by deficient land cover whenever erosivity of rainfall is strong. Phosphorus losses were less than nitrate losses, due to their lower water solubility and mobility in soil.


  • Mediterranean environment
  • agro-forestry systems
  • soil and water conservation
  • soil erosion
  • water scarcity
  • water pollution
  • Portugal

1. Introduction

Farming activities, as a part of natural resource management practice, impact both soil and water quality at the watershed level. The nonpoint source (NPS) pollution with nitrates and phosphates from agricultural areas, sometimes associated to soil erosion by water, is an important environmental problem [1]. For instance, the source of nitrates and phosphates from agriculture activities is reported as responsible for 46–87 and 20–40%, respectively, into European continental waters [2]. Soil and water conservation practices also help to reduce the loss of chemicals in runoff and maintain water quality [3]. Increases in nutrient losses and riverine nutrient loads have caused eutrophication of many coastal and freshwater ecosystems in many Mediterranean regions (Figure 1). Nonpoint source (NPS) pollution is an important environmental and water quality problem closely related to hydrologic behavior of watersheds as a basic management territorial unit [4]. However, watershed being a geographically dynamic unit, its behavior varies both spatially and temporarily. Nonpoint source pollution has become a global environmental issue and it has been a critical issue concerning environmental degradation caused in recent years.

Figure 1.

Soil erosion by water and consequences downstream, namely eutrophication of the water bodies.

The Mediterranean climate plays an important role in these environmental issues, like soil erosion and degradation of quality water bodies, given the irregularity and uncertainty of climatic patterns. The Mediterranean climate is characterized by irregular inter-annual and intra-annual precipitations (wet years mixed with recurrent droughts), and high concentrations of rainfall over a few months, many times over only a few days, and low rainfall during the summer [5, 6]. On the other hand, many of the conventional Mediterranean arable land cropping systems have a soil surface, which remains uncovered during long periods of time, corresponding to the soil preparation and crop establishment phases, which frequently occur during the periods of high rainfall concentration (Figure 1) [7, 8]. The runoff regime associated to this climatic condition determines in many years a low capacity for harvesting a convenient volume of water for agricultural activity and the implementation of some strategies for coping with water scarcity [9]. Both water scarcity and Mediterranean forestry systems create preferential conditions for wildfires, which can considerably change the hydrological processes and the landscape’s vulnerability in major peak flow and erosion events [10]. Thus, to have a better understanding of the processes of soil erosion and nonpoint source pollution at the basin scale, to better actuate on their prevention, monitoring studies conducted at a watershed scale are difficult to replicate in the way that traditional plot-scale research is designed, in order to compare responses of alternative management practices using only field observations [11]. At plot scale, it is easier to conduct research actions to study those processes, but it is impossible to obtain accurate valid information for large areas (Figure 2). So, computer simulation models provide an efficient and effective alternative for evaluating the effects of agricultural practices on soil and water quality at the watershed level [12].

Figure 2.

Plots to evaluate soil erosion by water under some crop rotations (Experimental Station in Polytechnic Institute of Castelo Branco/School of Agriculture).

Erosion is mainly due to climatic conditions, namely the precipitation patterns, and the inadequate use of soil by agriculture and forestry, but also through building constructions and uncontrolled water runoff from roads and other sealed surfaces. According to this, the regions of the Mediterranean basin where the higher rates of soil erosion are mapped are coincident with the regions where the highest values of erosivity density are verified (Figure 3). In more than one-third of the total land of the Mediterranean basin, average yearly soil losses can exceed 15 tons ha−1. On the other hand, loss of organic matter is mainly due to intensive use of the land by agriculture, especially when organic residues are not sufficiently produced or recycled to soil. Agronomists consider soil with less than 1.7% organic matter to be in predesertification stage [13].

Figure 3.

Erosivity density (erosivity factor per precipitation unit) and soil erosion by water in Europe [13].

The main objective of this study is to understand the dynamics of pollutant sediments, nitrates, and phosphorus in close relation to hydrologic behavior of a small basin in typical Mediterranean environmental conditions.


2. Methodology

The study watershed is located within the Idanha Irrigation Scheme, Idanha-a-Nova County, Portugal, near the border with Spain and just north of the Tagus River (Figure 4). The study area is close to the International Tagus Natural Park, where we can find many terrestrial and aquatic unique ecosystems, like a Mediterranean forest very well adapted to these conditions (holm oak forest—Quercus ilex L.), and some communities of rare birds of prey (imperial eagle) and black stork. Therefore, this region is very sensitive from an environmental point of view. It covers an area of 189 ha and a perimeter of 6510 m, and presents a third-order hierarchy stream. The main natural stream is 2300 m long and runs north-southwest. The drainage density of the perennial streams is 12.2 m ha−1 [14].

Figure 4.

Location of the study catchment in Portugal.

Altitude varies from 212 m at the outlet of the basin to 248 m, and the slopes range from 0 to 4%; thus, the topography is flat to gently undulating (Figure 5). The limits of the study catchment are well defined.

Figure 5.

Topography of the study catchment and a downstream and upstream view.

The climate is typically Mediterranean continental. Average annual rainfall is 638 mm, with a rainless summer. The average temperature varies from 8.1°C in January to 25.3°C in August; the average reference evapotranspiration (ETo) ranges from 0.5 mm day−1 in January to 9.0 mm day−1 in July (Figure 6) [14]. According to the Köppen climate classification, the study catchment enters a Mediterranean Csa type climate (humid temperate with a dry and hot summer).

Figure 6.

Characterization of climate in the study catchment through parameters of rainfall, temperature, and evapotranspiration.

The study catchment covers an area of 189 ha and it is subdivided into 18 fields. About one-third (31%) of the catchment is not irrigable and is now devoted to a young cork (Quercus suber L.) and oak tree (Quercus rotundifolia Lam.) forest that was planted in 2001. Three distinct areas can be identified in terms of agricultural practices: uncultivated zone (area with cork and oak young forest), an area with intensive agricultural use in monoculture (maize and tobacco), that in some recent years has remained fallow, and an area with diversity of agricultural uses with smaller fields. The area with young cork and oak forest, located in the eastern part of the basin, can be seen in long fallow as it has not been cultivated for several years [14]. The irrigated fields use sprinklers with either center pivot units or stationary systems. As a consequence of the Common Agriculture Policy, that was put in practice at the beginning of this century, namely almost disappearing with tobacco, the irrigated area decreased to less than half after 2005 irrigation season, with a corresponding increase in fallow area. This reduction has clear implications on the hydrologic behavior of the basin during the irrigation season and subsequent rainfall season [15].

Soils in the catchment were mapped from preexisting unpublished studies, field inspections, photo-interpretation techniques, and detailed characterization of one soil profile per soil group. According to the FAO classification system [16, 17], the predominant soil groups in the watershed are Luvisols and Cambisol, these ones originated from fluvial deposits associated with the tributaries of the Tagus River (Figure 7). Fluvisols are also present in the catchment area, these originated from alluvial deposits associated with the main creek. An impermeable, fractured soil layer underlies the Luvisols classes at a depth of approximately 0.4 m. Luvisols1 are soils with a subsurface horizon of high activity clay accumulation and high base saturation and show marked textural differences within the profile. The surface horizon is depleted in clay while the subsurface argic horizon has accumulated clay (Bt horizon). A wide range of parent materials and environmental conditions lead to a great diversity of soils in this Reference Soil Group. Other names used for this soil type include sols lessivés (France), Alfisols (Soil Taxonomy), and Mediterranean soils (Portugal). The Luvisols of the Mediterranean region are widely distributed throughout Portugal, Spain, Italy, Greece, Albania, Croatia, Turkey, and Cyprus, which represent a significant percentage of the total area of soils in these countries. Luvisol is a fertile soil suitable for a wide range of agricultural uses. In the Mediterranean, it is commonly used for cereals. On sloping land, it requires measures such as man-made terraces to control erosion, and it is best suited for fruit trees, vineyards, olives, and grazing [13].

Figure 7.

Soils in the study catchment (FAO nomenclature), and a soil profile from a Luvisol.

A hydrological station was constructed and installed in 2004 at the outlet of a catchment (39°50’48″ N, 7°10’00″ W). The station consisted of (i) a long-throated flume, with a triangular control section for small water depths and triangular/trapezoidal control section for large water depths [18] and (ii) an ultrasonic sensor (“The Probe,” manufactured by Milltronics, Siemens Milltronics Process Instruments Inc., Ontario, Canada) connected to a data logger continuously measuring and recording the water level at the flume (Figure 8). Pollutants, nitrates, sediments, and salts, were evaluated with a multiparameter probe for monitoring and logging water quality (TROLL 9500 Water Quality Instruments, manufactured by In-Situ Inc., Fort Collins, Colorado, USA). Farming practices, related to Mediterranean crops rotations, were recorded by farmers and verified by direct observations during visits to the catchment to apply to a computer model. The irrigation methods used in the study watershed are sprinkler center pivot and stationary sprinkler (in the areas not covered by the pivot machines).

Figure 8.

Hydrometric and water quality station located at the outlet of the study catchment.

Several available hydrologic models were evaluated and the Annualized Agricultural Non-Point Source (AnnAGNPS) pollution model was selected as the simulation tool to be used in this study. AGNPS is a suite of computer models resulting from the joint effort between the Agricultural Research Service (ARS) and Natural Resources Conservation Service (NRCS) agencies and it has been developed to evaluate farming and conservation practices through prediction of nonpoint source pollutant loadings within agricultural watersheds. Within AGNPS, AnnAGNPS is a continuous simulation, mixed-land use, watershed-scale computer model designed to predict the origin and movement of water, sediment, and chemicals at any location in agricultural watersheds [19]. The model estimates erosion caused by different processes such as sheet and rill, tillage-induced gullies, classical gullies, and streambed and bank sources [20]. AnnAGNPS has been calibrated, validated, and applied for runoff and sediment yield losses from watersheds in different geographic locations, conditions, and management practices [21, 22]. The AnnAGNPS model was applied in this study and afterward was calibrated and validated to the conditions in the study catchment [23]. Table 1 lists the main input parameters commonly used in AnnAGNPS simulations.

Group of parameters Input variables Methods
Climate Daily rainfall
Daily maximum and minimum temperatures, wind direction and speed, daily percentage cloud cover and dew point temperature
Annual distribution EI30
Type of rainfall distribution (TR-55)
Two year 24 h precipitation
Measurements at the Ladoeiro [24] and Ribeiro de Freixo stations (data not published)
Measurements at the Ribeiro de Freixo station (data not published)
Calculated with measured data, from Ladoeiro data [24] and the methodology described by Wischmeier and Smith [25]
Comparison with the 24 h rainfall distribution curve calculated with the data from Ladoeiro station [24]
Gumbel method [26] applied to the data from Ladoeiro station [24]
Topography Drainage area and limit of the catchment, cells area, reaches length, mean slope of cells and reaches, RUSLE LS factors
Topographic and hydrologic configuration
Application of TopAGNPS and AgFlow [27]—ArcView interface 3.2, using a DEM with 1 m vertical resolution
Manipulation of the CSA and MSCL parameters, and comparison to observed and simulated natural stable reaches
Soils Depth (horizons)
Saturated hydraulic conductivity
Bulk density
Field capacity and wilting point
Percentage of organic matter
Coarse elements (%)
Field observations, cleaning of profiles
Method of Robison pipette [28]
Rawls and Brakensiek [29]
Mass/volume of clods with wax to measure their submerged weight
Richard’s methodology
Methodology of Walkley-Black [30]
ISO [31]
Field observations
Operations and Management Crop data
Crop operations, included irrigation
Irrigation application rate
RUSLE-factors C and P
Farmer’s information and some bibliography related with crops in the catchment
Farmer’s information and some observation in the fields
Measured in field [32, 33]
Wischmeier and Smith [25]
Others CN
Manning’s n
TR-55 [34]
Wischmeier and Smith [25]

Table 1.

AnnAGNPS Input parameters and methods used in their evaluation.

Note: DEM, digital elevation model; CSA, critical source area; MSCL, minimum source critical length; CN, curve number.

Irrigation performance and catchment water usage were quantified for various irrigation and rainfed seasons using the following indicators [35, 36]:

Runoff coefficient: R O C = R O R + I E1
Relative irrigation supply: R I S = I I required E2

where ROC is the fraction of rainfall and irrigation that contribute to runoff verified at the outlet of the basin, RO (mm) is the runoff verified at the outlet of the study catchment, R (mm) is rainfall, I is irrigation applied on the irrigated fields in the study basin, and Irequired is the required irrigation calculated by FAO methodology, as the water requirements to satisfy the full water needs of the crop [37].


3. Results

In Mediterranean climatic conditions, the accumulated erosivity curve shows a stationary phase in the summer months (June, July, August, and September), because it has not practically rained, and two phases of growth, one in the first five months of the year and another, more intense, in the last months of the year. This phase is particularly important, specifically for winter crops such as cereals, since the surface of the soil is unprotected [38, 39]. In contrast, the annual erosive rain is less than 50% of the total annual rain, according to the results of 14 years (2001–2014). The adjustment of linear tendency shows that the larger the total rainfall, the less is the erosive rain (Figure 9). Therefore, the analysis of rainfall and erosivity (EI30 methodology, as a compound index of kinetic energy of the rain, E, and the maximum 30-min intensity) [25] distribution over many years (2001–2014) shows typically Mediterranean climate conditions.

Figure 9.

Distribution of average EI30 factor (MJ mm ha−1 h−1 yr−1), total rainfall, and erosive rainfall.

Case of a single event (Figure 10) shows that the superficial runoff dominates the hydrological response of this basin during the most significant events, known as Hortonian model [40]. This behavior is in part responsible for an almost impermeable layer (horizon Bt), characteristic of Luvisols. The significant variability in runoff is justified not only by differences in characteristics of storms and surface soil moisture content, but also by preceding long dry periods which were likely to ease the effects of rainfall [41, 42]. So, the event occurred on October 20, 2004, in dry soil conditions, reached a discharge equal to 8289 m3 in sequence of 39.9 mm of precipitation (Figure 9). In a subsequent event, on October 27, 2004, in wet soil conditions, the total discharge was 80,457 m3 derived from 42.3 mm of precipitation (Figure 10). This significant difference between two runoff events caused by almost the same precipitation illustrates the decisive influence of the antecedent soil moisture conditions in the magnitude of the flash floods at the small basin scale. It is considered by some authors (e.g., [43]) as the most important soil factor for rapid runoff and flash flooding, with a tendency to produce in Mediterranean region more extreme floods than other European regions [44].

Figure 10.

Precipitation and hydrograph (dotted line), recorded between November 30 and December 2, 2004, showing a hortonian behavior [40], and influence of the antecedent soil moisture conditions on the magnitude of hydrologic events.

The relation between the water entry in the catchment and the runoff coefficient is completely different under irrigation and rainy seasons. In fact, it is possible to establish a good positive correlation in irrigation, while in the rainy season no relation is possible (Figure 11). As is comprehensible, in irrigation conditions, the runoff coefficient is proportional to the volume of irrigation water provided to the fields, given that the water is supplied by a regular behavior. Nevertheless, only from a minimum value of water supplied to the irrigation fields, start to have runoff at the outlet of the catchment. For example, in the 2007 and 2009 irrigations seasons, the water supplied to the irrigations fields was 56 mm and 54 mm respectively (Table 2), and the runoff observed in the outlet of the study basin was 0 mm in both irrigation seasons (Figure 11, left chart). On the other hand, as was concluded earlier [14], the reduction of the runoff coefficient is more than proportional relatively to the reduction of water supplied to the catchment. Therefore, it is necessary to make difference on the concept irrigation efficiency at different territorial scales; performance indicators of irrigation at field scale or at catchment scale are not the same [45]. During usual rainy seasons, as in the irrigation seasons, it is possible to obtain a good positive relation between rainfall and runoff coefficient, but in some unusual years the relation is completely shuffled. That is characteristic for Mediterranean climatic conditions, namely precipitation patterns (Figure 11, right chart). The highest value of runoff coefficient verified in the study catchment (0.36) is obtained for the rainfall of only 302 mm in a very dry season (2004/2005).

Figure 11.

Runoff coefficient (ROC) in irrigation and rainfed seasons evaluated in the study catchment.

In many regions in the Mediterranean basin, like in the study catchment, the indicator cropping intensity (CI), defined as the ratio of arable to cultivated area, is low due to some climatic, institutional, or socio-economic factors. This is one of the threats to the region, as a result of the higher susceptibility to environmental damage, like wild fires, and the consequences of soil erosion [46]. In the last century, many rural regions in the Mediterranean basin had significant demographic and socio-economic changes, which determined a substantial decrease in population, abandonment of agricultural land and increase of the human-induced desertification process [47]. Water scarcity and more frequent droughts are other serious problems in the Mediterranean region, which require an enhancement in water management and more efficient water allocation, distribution, and use [48]. Relative water supply (RWS) is always below the unity value, except in the irrigation seasons of 2004 and 2008 in the study catchment, and sometimes even below the value 0.5, clearly denoting water scarcity in most of the irrigation seasons (Table 2). Both, deficient water distribution in many irrigation districts and/or low water storage in the reservoirs, are the most frequent reasons that determine water scarcity conditions on the irrigation fields.

Cropping season Year Rain (mm) Irrigation (mm) Runoff (mm) CI ROC RIS
Irrigation 2004 346 32 0.81 0.09 1.11
2005 118 6 0.39 0.05 0.66
2006 90 1 0.38 0.01 0.62
2007 56 0 0.24 0.00 0.60
2008 286 25 0.64 0.09 1.03
2009 54 0 0.16 0.00 0.42
2012 43 0 0.21 0 0.49
Rainfed 2004–2005 302 109 0.23 0.36
2005–2006 591 112 0.14 0.19
2006–2007 1021 255 0.14 0.25
2007–2008 422 62 0.06 0.15
2008–2009 333 36 0.16 0.11
2009–2010 601 180 0.32 0.30
2010–2011 579 134 0.24 0.23
2011–2012 288 28 0.24 0.10

Table 2.

Level of agricultural intensification (CI), runoff coefficient (ROC), and relative irrigation supply (RIS), in the study catchment, during seven irrigation seasons and eight rainfed seasons.

Mediterranean climate, as referred earlier, presents a strong intra- and inter-annual irregularity, being especially evident in the rainfall patterns [49]. Additionally, studies of hydrological behavior in Mediterranean climate using precipitation and discharge averages are not relevant, since this kind of climate is dominated by extreme events [50]. The two graphics in Figure 12 show exactly what is a very abnormal (2004–2005) and normal year (2005–2006), as to the quantity and the distribution of rainfall and runoff. In fact, in 2004–2005, the amount of precipitation was 302 mm, distributed mainly into five events and giving a few more intense runoff events, highlighting the first event with a great magnitude. In the same period of time, the amount of runoff was 109 mm, determining a runoff coefficient equal to 0.36, a value which is uncommon in these conditions. In the hydrological year 2005–2006, the amount of rainfall (591 mm) was close to the average value in this region (638 mm) and was distributed by many more rainfall events. It should be noted that the amount of runoff (112 mm) was almost the same as that of the previous year, but giving a lower runoff coefficient (0.19). In the first year of analysis (2004–2005), the runoff started on October 19 and finished on April 20, without any gap in this period (Figure 12). Considering that, between December 13 and March 20 the amount of precipitation was 15.3 mm, and the continuous runoff was possible only because the natural channels’ network was not disconnected from the aquifers or groundwater table [51].

Figure 12.

Rainfall and runoff verified in the rainfed seasons 2004–2005, as a dry year, and 2005–2006, as a meteorologically normal year.

Mediterranean rainfall can reach, sometimes with very high intensity, very important runoff events, like this one illustrated in Figure 13, which occurred in the hydrological year 2004–2005. This event occurred in the conditions of saturated soil, due to the amount of rainfall in the previous days (October 21–23; 77.9 mm), but, as noted before [52], big events can be reached when the intensity of precipitation exceeds the initial capacity of soil infiltration, or the maximum intensity of precipitation occurs in a favorable moment of the storm. In the present event, verified in a very dry year (302 mm), the runoff coefficient was equal to 0.85, characteristic to streams with torrential patterns, and the erosion annual rate in this event was 79.1%; with the potential to carry coarse elements outside the basin. Among the two different types of events identified by Camarasa-Belmonte [53], this one is related to the type with high intensity, occurring at the beginning of the storm and generating hydrographs that are very similar to the hyetographs. The other type of events, according to Camarasa-Belmonte [53], occur in the conditions of low intensity of precipitation and rain peaks at the end of the event, generating hydrographs quite different from hyetographs. Both rising time and recession time are reached relatively quickly, due to the characteristics of this storm and the physical conditions of the catchment. Since this catchment does not have important water storage in the channels, soil and surface, rapid recession time is verified, when the storm stops [54].

Figure 13.

Runoff and erosion rates verified in a very strong event in October 2004.

The effects of surface runoff and soil erosion, as well as nonpoint source pollution at watershed scale, can be predicted with hydrological models. They are effective to simulate various combinations and scenarios, and indicate the best management practices to minimize these processes [55]. Several hydrological models were evaluated to simulate runoff and soil erosion, as well as nonpoint source pollution by nitrates and phosphorus. The selection indicates the AnnAGNPS model (Annualized Agricultural NonPoint Source), as the most suitable for this task. As for spatial distribution of runoff volume and the ratio of erosion in the basin, the results based on observations (Figure 14) seem to indicate that there is a distinction between the two runoff-producing areas in the basin: one that generates low runoff (0–40 m3 ha−1 yr−1), which corresponds to the uncultivated area in the basin (58.6 ha), and another, which corresponds to the remaining area of the watershed (130.4 ha) occupied by various crops, producing higher runoff (41–140 m3 ha−1 yr−1). The variation of runoff in the cultivated part of the basin depends on the topography, the type of soil, and their intrinsic characteristics. The relations showed in Figure 6 enable to conclude that there is no correlation between the average sediment production and the average runoff volume in most of the cells in the basin [38]. Although having an influence on the erosion process, the average volume of runoff is not the determining factor. Other factors such as land cover throughout the year, especially during the concentration of rainfall, and the occurrence of more or less erosive storms appear to have more significant influence [56, 39].

Figure 14.

Average simulated runoff and erosion rates, generated in each cell and subcatchment, by AnnAGNPS model (period 2003–2005).

The massive use of fertilizers is common to all the agricultural systems around the world. Nitrogen losses, from the agricultural nonpoint sources, has become one the most significant threats to the quality of water bodies [57]. Due to the solubility of this contaminant (inorganic-N), there is a close relationship between hydrological processes and the loss of nitrogen, with the superficial and subsuperficial runoff. Organic-N is adsorbed in the fine soil particles, while their loss occurs preferably through soil erosion [58]. The loss of nitrates is not affected by rainfall variability as the sediment loss. According to some earlier studies, the base flow is the main source of nitrates at the surface of water bodies [59]. Regarding the nonpoint source pollution of nitrates (Figure 15), the research has shown that the process occurs when there is significant runoff, given the solubility of nitrates, and when there is availability of this nutrient in the soil, in sequence of more or less intense fertilization [60]. The values simulated by AnnAGNPS are similar to the reference values obtained in other studies: pristine watersheds and organic agriculture (0.76–10.85 kg NO3-N ha−1 yr−1) [61]; nonirrigated agriculture (26.10 kg NO3-N ha−1 yr−1) [59]; irrigated agriculture (59.00 kg NO3-N ha−1 yr−1) [60]. Moreover, only about 0.3 mg l−1 of nitrates is needed for growth of algal and to promote the eutrophication [62]. The critical value was largely surpassed in the period of simulation, as shown in Figure 15. According to the efficiency of nitrogen fertilization practices, the loss of nitrates can reach high values (30–80%) in intensive cropping systems [63], like maize or tobacco in the study catchment. So, the riparian buffers and vegetation strips on stream water can be a very good measure to improve water quality downstream of the agricultural fields, as observed and analyzed in some earlier studies [64, 65].

Figure 15.

Average simulated runoff and nitrates load, generated in each cell and subcatchment, by AnnAGNPS model (period 2003–2005).

The continuous use of phosphorus in agriculture, many times with low P-use efficiency by crops, increases in phosphorus concentration in the soil, and increasing the risk of phosphorus in runoff along with leaching to water bodies, contribute to the eutrophication [66, 67]. Phosphorus is an element with little mobility in the soil, and can be transferred from agricultural systems to water bodies dissolved in the superficial runoff, leaching to deeper soil layers, or in conjunction with mineral and organic sediment in the water erosion process. Similar to this study, Olness et al. [68] observed losses of P to agricultural watersheds lower than 5% of the most recent P fertilization, with a total P discharge in runoff waters ranging from 1.0 to 11.5 kg P ha−1. Considering the volume of runoff verified in our basin, the value highlighted by some authors as the level to prevent the risk of eutrophication (0.05 mg L−1) [69] is probably often exceeded. Most of the phosphorus used in the agricultural fields, as in our basin, is for winter cereals crops (oat, wheat, rye), most common in Mediterranean agricultural systems. Phosphorus can load in solution outside the basin in surface or subsurface runoff, depending on the nature of P-forms (mineral or organic) and in the soil texture, or attached to the finest soil particles [70, 56]. The analysis of the relation showed in Figure 16, referent to the simulated phosphorus load by AnnAGNPS model, suggests that most of the phosphorus is lost in solution and not with the soil colloidal particles loaded outside the watershed, given the subcatchments where high values of P-load occurred (10–12 kg ha−1.yr), and relatively low rates of soil erosion were verified. The phosphorus loss clearly occurred mostly in the agricultural fields and almost null in the forested area, while the high loss values were verified in the fields occupied by winter cereals.

Figure 16.

Simulated average phosphorus load by AnnAGNPS model, generated in each subcatchment (period 2003–2005).


4. Conclusions

Both soil and water resources in the Mediterranean environmental conditions are exposed to physical, chemical, and biological degradation. The phase of pronounced erosivity in the Mediterranean climatic condition corresponds to the last 3 months of the year and it is particularly important, mainly for winter crops such as cereals, since the surface of the soil is unprotected. The superficial runoff dominates the hydrological response of the study basin during the most significant events. The antecedent soil moisture condition is considered a factor of greater importance. The irregularity of intra- and inter- annual precipitation is a pattern specific to the Mediterranean climate, that usually determines the grade of soil cover and the soil erosion process. The climate change can additionally aggravate this situation. Water scarcity is another threat in the irrigated agricultural systems of the Mediterranean region. It is verified in many irrigation seasons in this study, where the indicator RWS was much higher. Mediterranean rainfall, sometimes with very high intensity and verified in dry years, can reach very important runoff events like the one that occurred at the end of October in the hydrological year 2004–2005. Runoff coefficient was equal to 0.85 and an annual erosion rate equal to 79.1%. Spatial distribution of runoff was primarily influenced by topography and soil management, which is common to Mediterranean agricultural systems, namely in grain crops systems such as oats and wheat. The simulation of spatial distribution of nitrate loads shows a dependence of the spatial distribution of runoff, due to its high solubility. Despite the close relation between soil erosion, water and runoff, the simulation using the AnnAGNPS model of spatial distribution of soil erosion by water indicates that the process does not directly depend on it. Therefore, soil erosion is mostly influenced by deficient land cover whenever erosivity of rainfall is strong. Phosphorus losses were less than nitrate losses, due to their lower water solubility and mobility in soil.



The author gratefully acknowledges funding from Portuguese Foundation for Science and Technology (FCT) for financially supporting this research through the project GeoBioSciences, GeoTechnologies, and GeoEngineering (UID/GEO/04035/2013).


  1. 1. Thornton JA, Rast W, Holland MM, Jolankai G, Ryding SO. Man and the Biosphere Series. Paris: UNESCO; 1999. p. 478
  2. 2. EEA—European Environment Agency. Nutrients in Europe Ecosystems. Environmental Assessment Report No. 4 EEA. 1999
  3. 3. Berry JK, Delgado JA, Khosla R, Pierce FJ. Precise conservation for environmental sustainability. Journal of Soil and Water Conservation. 2003;58(6):332-339
  4. 4. European Community. Directive 2000/60/EC of October 23 2000 of the European Parliament and of the Council establishing a framework for community action in the field of water policy. Official Journal of the European Communities. 2000;L327:1-72
  5. 5. García-Barron L, Aguilar M, Sousa A. Evolution of annual rainfall irregularity in the southwest of the Iberian Peninsula. Theoretical and Applied Climatology. 2011;103(1-2):13-26
  6. 6. Lionello P, Malanotte P, Boscolo R. Mediterranean Climate Variability. Amsterdam: Elsevier; 2006
  7. 7. Poesen J, Nachtergaele J, Verstraeten G, Valentin C. Gully erosion and environmental change: Importance and research needs. Catena. 2003;50(2-4):91-133
  8. 8. DeSantisteban L, Casalí J, López JJ. Assessing soil erosion rates in cultivated areas of Navarre (Spain). Earth Surface Processes and Landforms. 2006;31(4):487-506
  9. 9. Pereira LS, Cordery I, Lacovides I. Coping with Water Scarcity-Addressing the Challenges. Springer; Dordrecht (Netherlands), 2009. p. 382. DOI: 10.1007/978-1-4020-9579-5
  10. 10. Stoof CR, Vervoort RW, Iwema J, Elsen E, Ferreira AJ, Ritsema CJ. Hydrological response of a small catchment burned by experimental fire. Hydrology and Earth System Sciences. 2012;16(267-285):1027-5606
  11. 11. Mohammed H, Yohannes F, Zeleke G. Validation of agricultural non-point source (AGNPS) pollution model in Kori watershed, South Wollo, Ethiopia. International Journal of Applied Erath Observation and Geoinformation. 2004;6:97-109
  12. 12. He C. Integration of geographic information systems and simulation model for watershed management. Environmental Modelling & Software. 2003;18:809-813
  13. 13. EC, European Commission, Joint Research Centre & European Soil Bureau, editors. Soil Atlas of Europe. Luxembourg: European Commission; 2005
  14. 14. Duarte AC. Non-point source pollution caused by agricultural activity, at watershed scale [thesis]. Córdoba (Spain): University of Córdoba; 2006. p. 192
  15. 15. Duarte AC. Non-point source pollution in an agricultural catchment and the quality of return flows, under Mediterranean conditions. Fresenius Environmental Bulletin. 2014;23/No12a/2014:3237-3247
  16. 16. FAO, Food and Agriculture Organization of the United Nations, editors. World Reference Base for Soil Resources-FAO World Soil Resources Report 84. Rome; 1998
  17. 17. FAO, Food and Agriculture Organization of the United Nations, editors. World reference base for soil resources 2014. International soil classification system for naming soils and creating legends for soil maps. Update 2015. World Soil Resources Reports No. 106. Rome; 2016
  18. 18. Bos MG, Replogle JA, Clemmens AJ. Flow measuring flumes for open channel systems. New York: John Wiley & Sons; 1991. p. 321
  19. 19. Cronshey RG, Theurer FG. AnnAGNPS—Non-point pollutant loading model. In: Proceedings of First Federal Interagency Hydrologic Modelling Conference; 19-23 April; Las Vegas, NV, ASAE (American Society of Agricultural Engineers), St. Joseph-MI (USA); 1998
  20. 20. Bingner RL, Wells RR, Momm HG, Theurer FD, Frees LD. Development and application of gully erosion components within the USDA AnnAGNPS watershed model for precision agriculture. In: Proceedings of the International Conference on Precision Agriculture; July 18-21; Denver, Colorado, International Society of Precisition Agriculture (ISPA), Monticello-IL (USA); 2010
  21. 21. Sadeghi A, Kwang Y, Graff C, McCarty G, McConnell L, Shirmohammadi A, Hively D, Sefton KA. Assessing the performance of SWAT and AnnAGNPS Models in a Coastal Plain Watershed, Choptank River, Maryland, USA. In: ASABE Meeting. St. Joseph, MI: American Society of Agricultural and Biological Engineers; 2007
  22. 22. Licciardello F, Zema DA, Zimbone SM, Bingner RL. Runoff and soil erosion evaluation by AnnAGNPS model in a small Mediterranean watershed. Transactions of the ASABE. 2007;59(5):1585-1593
  23. 23. Duarte AC, Mateos L, Fereres E. Modelling runoff with AnnAGNPS model in a small agricultural catchment, in Mediterranean environment. In: Proceedings of the 21st Century Watershed Technology Conference and Workshop; May 27-June 1, 2012; Bari (Italy). St. Joseph MI: ASABE; 2012
  24. 24. INAG-Serviço Nacional de Informação de Recursos Hídricos. Boletim de Precipitação Anual—Estação do Ladoeiro-14N/02U [Internet]. 2008. Available from: http// [Accessed: January 25, 2008]
  25. 25. Wischmeier WH, Smith DD. Predicting Rainfall Erosion Losses—Agricultural Handbook No.537. Washington, DC: United States Department of Agriculture; 1978
  26. 26. Gumbel EJ. Statistics of Extremes. New York: Columbia University Press; 1958
  27. 27. Garbrecht J, Martz LW. TOPAGNPS, an automated digital landscape analysis tool for topographic evaluation, drainage identification watershed segmentation and subcatchment parametrization for AGNPS 2001. In: Agricultural Research Service, editor. Washington, DC: Watershed Modelling Technology; 1999
  28. 28. USDA-SCS, United States Department of Agriculture-Soil Conservation Service. Soil Survey Laboratory Methods and Procedures for Collecting Soil Samples-Soil Survey Report 1. Washington, DC; 1982
  29. 29. Rawls WJ, Brakensiek DL. Estimation of soil and water retention and hydraulic properties. In: Morel-Seytoux H.J., editor. Unsatured Flow in Hydrologic Modelling Theory and Practice. Beltsville, MD: Kluwer Academic Publishers; 1989. pp. 275-300
  30. 30. Nelson DW, Sommers LE. Total carbon, organic carbon, and organic matter. In: Page AL, Miller H, Keeney DR, editors. Methods of Soil Analysis: Chemical and Microbiological Properties. 2nd ed. Madison, WI: Soil Science Society of America; 1982. pp. 539-577
  31. 31. International Organization for Standardization. Soil Quality—Determination of pH—Standard ISO 10390:2005. Geneva, Switzerland: Standards ISO; 2005
  32. 32. ASAE, American Society of Agricultural Engineers, editor. ASAE Standards 2005. St. Joseph, MI, American Society of Agricultural Engineers (ASAE); 2005
  33. 33. Merriam JL, Keller J. Farm Irrigation Systems Evaluation: A Guide for Management. Logan, UT: Utah State University; 1978
  34. 34. USDA-NRCS, United States Department of Agriculture-Natural Resources Conservation Service, editor. Urban Hydrology for Small Watersheds-Technical Release 55. Washington, DC; 1986
  35. 35. Molden DJ, Gates TK. Performance measures for evaluation of irrigation-water-delivery systems. Journal of Irrigation and Drainage Engineering. 1990;116:804-823
  36. 36. Burt CM, Clemmens AJ, Strelkoff TS, Solomon KH, Bliesner RD, Hardy LA, Howell TA, Eisenhauer DE. Irrigation performance measures: Efficiency and uniformity. Journal of Irrigation and Drainage. 1997;123:423-442
  37. 37. Allen RG, Pereira LS, Raes D, Smith M. Crop Evapotranspiration. Guidelines for Computing Crop Water Requirements-FAO Irrigation and Drainage Paper No 56. Rome: FAO, Food and Agriculture Organization of the United Nations; 1998
  38. 38. Pathak P, Wani SP, Singh P, Sudi R. Sediment flow behaviour from small agricultural watershed. Agricultural Water Management. 2004;67:105-117. DOI: 10.1016/j.agwat.2004.02.005
  39. 39. Duarte AC, Mateos L, Fereres E. Application of the AnnAGNPS model to a small agricultural watershed, to analyse the spatial and temporal distribution of the pollutants and runoff. In: Proceedings of the 5th International Congress of the European Society for Soil Conservation; 25-28 June; Palermo (Italy), European Society for Soil Conservation (ESSC); 2007
  40. 40. Horton RE. Erosional development of streams and their drainage basins: Hydrophysical approach to quantitative morphology. Bulletin of the Geological Society of America. 1945;56:275-370
  41. 41. Taguas EV, Ayuso JL, Peña A, Yuan Y, Pérez R. Evaluating and modelling the hidrological and erosive behaviour of an olive orchard microcatchment under no-tillage with bare soil in Spain. Earth Surface Processes and Landforms. 2009;34:738-751
  42. 42. Beven KJ. Rainfall-Runoff Modelling—The Primer. 2nd ed. Chichester: Wiley-Blackwell; 2012. p. 488
  43. 43. Grillakis MG, Koutroulis AG, Kommab J, Tsanis IK, Wagner W, Bloschl G. Initial soil moisture effects on flash flood generation—A comparison between basins of contrasting hydro-climatic conditions. Journal of Hydrology. 2016;541:206-217
  44. 44. Gaume E, Bain V, Bernardara P, Newinger O, Barbuc M, Bateman A, Blaskovicová L, Bloschl G, Borga M, Dumitrescu A, Daliakopoulos I, Garcia J, Irimescu A, Kohnova S, Koutroulis A, Marchi L, Matreata S, Medina V, Preciso E, Sempere-Torres D, Stancalie G, Szolgay J, Tsanis I, Velasco D, Viglione A. A compilation of data on European flash floods. Journal of Hydrology. 2009;367:70-78
  45. 45. Lorite IJ, Mateos L, Fereres E. Evaluating irrigation performance in a Mediterranean environment. II. Variability among crops and farmers. Irrigation Science. 2004;23:85-92
  46. 46. Lourenço L, Nunes AN, Bento-Gonçalves A, Vieira A. Soil erosion after wildfires in Portugal: What happens when heavy rainfall events occurs. In: IN-TECH, editor. Research on Soil Erosion. IN-TECH, Rijeka (Croatia); 2012. pp. 65-88. DOI:
  47. 47. Moreira F, Viedma O, Arianoutsou M, Curt T, Koutsias N, Rigolot E. Landscape and wild fire interactions in southern Europe: Implications for landscape management. Journal of Environmental Management. 2011;92:2389-2402
  48. 48. Esteve P, Varela-Ortega C, Blanco-Gutierrez I, Downing TE. A hydro-economic model for the assessment of climate change impacts and adaptation in irrigated agriculture. Ecological Economics. 2015;120:49-58
  49. 49. García-Barrón L, Camarillo JM, Morales J, Sousa A. Temporal analysis (1940-2010) of rainfall aggressiveness in the Iberian Peninsula basins. Journal of Hydrology. 2015;525:747-759
  50. 50. Brakenridge GR. River flood regimen and floodplain stratigrafy. In: Baker VR, Kochel RC, Patton PC, editors. Flood Geomorphology. Wiley and Sons, New York (USA); 1988. pp. 139-155
  51. 51. Longobardi A, Villani P, Grayson RB, Western AW. On the relationship between runoff coefficient and catchment initial conditions. In: Proceeding of MODSIM 2003; 2003. pp. 867-872
  52. 52. Cerdà A, Doerr SH. Soil wettability, runoff and erodibility of major dry-Mediterranean land uses types on calcareous soils. Hydrological Processes. 2007;21:2325-2336
  53. 53. Camarasa-Belmonte AM. Flash floods in Mediterranean ephemeral streams in Valencia Region (Spain). Journal of Hydrology. 2016;541:99-115
  54. 54. Bras R. Hydrology. An Introduction to Hydrologic Science. New York: Addison-Wesley Pub; 1990. p. 633
  55. 55. Zema DA, Bombino G, Denisi P, Licciardello F, Zimbone SM. Prediction of surface runoff and soil erosion at watershed scale: Analisys of the AnnAGNPS model in different environmental conditions. In: Godone D, Stanchi S, editor. Research on Soil Erosion. IN-TECH, Rijeka (Croatia); 2012. pp. 4-31. DOI:
  56. 56. Lemunyon JL, Daniel TC. Quantifying phosphorus losses from the agricultural system. Journal of Soil and Water Conservation. 2002;57(6):399-401
  57. 57. Carpenter SR, Caraco NF, Correl DL, Howarth RW, Sharpley AN, Smith VH. Non-point pollution of surface waters with phosphorus and nitrogen. Ecological Applications. 1998;8(3):559-568
  58. 58. Dekun S. Characteristics of nitrogen loss on slope under simulated rainfall. Bulletin of Soil and Water Conservation. 2009;29(5):98-101
  59. 59. Schilling K, Zhang YK. Baseflow conntribution to nitrate-nitrogen export from a large agricultural watershed, USA. Journal of Hydrology. 2004;295:305-316
  60. 60. Causapé J, Quilez D, Aragues R. Salt and nitrate concentrations in the surface waters of the CR-V irrigation district (Bardenas I, Spain): diagnosis and prescriptions for reducing off-site contamination. Journal of Hydrology. 2004;295(1-4):87-100
  61. 61. Honisch M, Hellmeier C, Weiss K. Response of surface and subsurface water. Geoderma. 2002;105(3-4):277-298
  62. 62. Brooks KN, Folliott PF, Gregersen HM, DeBano LF. Hydrology and the Management of Watersheds. 2nd ed. Ames: Iowa State University Press; 1997. p. 502
  63. 63. Angás P, Lampurlané J, Cantero-Martinez C. Tillage and N fertilization effects on N dynamics and barley yield under semiarid Mediterranean conditions. Soil &Tillage Research. 2006;87:59-71
  64. 64. Broadmeadow S, Niesbet TR. The effects of riparian forest management on the freshwater environment: A literature review of best management practice. Hydrology and Earth System Science. 2004;8(3):286-305
  65. 65. Fiener P, Auerswald K. Seasonal variation of grassed waterway effectiveness in reducing runoff and sediment delivery from agricultural watersheds in temperate Europe. Soil & Tillage Research. 2006;87:48-58
  66. 66. Kleinman PJA, Sharpley AN, McDowell RW, Flaten DN, Buda AR, Tao L, Bergstrom L, Zhu Q. Managing agricultural phosphurus for water quality protection: Principles for progress. Plant and Soil. 2011;349:169-182. DOI: http//
  67. 67. Sharpley AN, Daniel TC, Gibson G, Bundy L, Cabrera M, Sims T, Stevens R, Lemunyon J, Kleinman PJA, Parry R. Brest Management Practices to Minimize Agricultural Phosphurus Impacts on Water Quality. USDA-ARS Publication 163. Washington DC: U.S. Gov. Printing Office; 2006
  68. 68. Olness A, Smith SJ, Rhoades ED, Menzel RG. Nutrient and sediment discharge from agricultural watersheds in Oklahoma. Journal of Environmental Quality. 1975;4:331-336
  69. 69. Golterman HL, Oude NT. Eutrophication of lakes, rivers and coastal seas. In: Hutzinger O, editor. The Handbook of Environmental Chemistry, Vol. 5 Part A. Berlin: Springer-Verlag; 1991. pp. 80-124
  70. 70. Kronvang B, Bechmann M, Lundekvam H, Behrendt H, Rubaek GH, Schoumans OF, Syversen N, Andersen HE, Hoffman CC. Phosphorus losses from agricultural areas in river basins: Effects and uncertainties of target mitigation measures. Journal of Environmental Quality. 2005;34:2129-2145. DOI: 10.2134/jeq2004.0439


  • From the Latin, luere, meaning to wash.

Written By

António Canatário Duarte

Submitted: November 29th, 2016 Reviewed: June 9th, 2017 Published: November 8th, 2017