Water quality predictions through linear regression - A brute force algorithm approach.

Linear regression is one of the oldest statistical modeling approaches. Still, it is a valuable tool, particularly when it is necessary to create forecast models with low sample sizes. When researchers use this method and have numerous potential regressors, choosing the group of regressors for a model that fulfills all the required assumptions can be challenging. In this sense, the authors developed an open-source Python script that automatically tests all the combinations of regressors under a brute-force approach. The output displays the best linear regression models, regarding the thresholds set by users for the required assumptions: statistical significance of the estimations, multicollinearity, error normality, and homoscedasticity. Further, the script allows the selection of linear regressions with regression coefficients according to the user's expectations. This script was tested with an environmental dataset to predict surface water quality parameters based on landscape metrics and contaminant loads. Among millions of possible combinations, less than 0.1 % of the regressor combinations fulfilled the requirements. The resulting combinations were also tested in geographically weighted regression, with similar results to linear regression. The model's performance was higher for pH and total nitrate and lower for total alkalinity and electrical conductivity.•A Python script was developed to find the best linear regressions within a dataset.•Output regressions are automatically selected based on regression coefficient expectations set by the user and the linear regression assumptions.•The algorithm was successfully validated through an environmental dataset.

Is it safe to remove a dam at the risk of a sprawl by exotic fish species?

The longitudinal dimension of river connectivity has been significantly disrupted by barriers to compensate for water demand in the long periods of water scarcity in the Iberian Peninsula. The scale of this modification is widespread in the Portuguese part of Douro River network where, side to side with agriculture water demand, there is a constant increase in hydropower production. Thus, native species in Iberian freshwater systems, performing reproductive migrations along the rivers, are strongly affected by the amplification of fragmentation caused by the tremendous density of transversal obstacles in this river basin. We aimed to prioritize dam removal in the Portuguese part of Douro River, mainly considering obsolete barriers (small dams, weirs) based on a spatial multicriteria decision analysis (GIS-MCDA) based on a prioritization procedure. A diversity of parameters were used to prioritize (rank) the dam's suitability for removal, considering the losses of connectivity and fish biodiversity, habitat degradation, negative effects on water quality and ecological conditions, and socio-economic factors. Different weights were assigned to the different attributes in each criterion according to their importance. The analysis also included a significant constraint: the potential spreading of exotic invasive fish species if connection was reestablished through dam removal. This procedure started with the georeferencing of 1201 transversal obstacles that were further characterized for their relative permeability to fish migration. In conclusion the model used allowed to identify 158 priority barriers, as well as the 5 most fragmented tributaries, which means the most impacted by river regulation. In 8 cases the barriers were big dams (> 15 m), whereas in the remaining 150 were weirs. From a final rank of 20 most impacting structures, the MCDA results also identified two cases where potential removal could trigger the additional impact to native fish species related to the sprawl of alien populations.

A structural equation model to predict macroinvertebrate-based ecological status in catchments influenced by anthropogenic pressures.

A Partial Least Squares-Path Model (PLS-PM) was developed for the Ave River Basin (North of Portugal), and the results used in a scenario analysis. The data for PLS-PM comprised a set of anthropogenic pressures, water quality parameters, and a macroinvertebrate-based biodiversity index (IPtIN) used to assess the ecological status of streams. These groups of measured parameters (called latent variables) were given the names "Pressures", "Contamination" and "Ecological Integrity". Besides, latent variables were connected through path coefficients representing potential causal effects among them. In a large portion of Ave the ecological status of streams is currently bad or poor. Nitrate and coliforms were the most weighted measured variables of latent variable "Contamination", with w ≈ 0.7 and w ≈ 0.2, respectively. The highest "Pressures" weights were ascribed to livestock farming (0.7) and population density (0.4). The connections "Pressures"-"Contamination" and "Contamination" - "Ecological Integrity" exposed a sequence of direct negative effects between the three variables, expressed in the corresponding path coefficients (pc = 0.87 and pc = -1.11). Paradoxically, a direct negative effect of "Pressures" over "Ecological Integrity" was absent (pc = 0.29). Therefore, the poor ecological status of local stream waters might not be directly related to the presence of potentially threatening contaminant sources (the "Pressures"), but to ineffective monitoring of livestock farming and wastewater treatment activities that potentiate (accidental) releases of contaminants into the streams. The lack of a direct link "Pressures" - "Ecological Integrity" supported the results of pressure change versus IPtIN change scenarios. Regardless of some significant reductions of anthropogenic activity and population density until 2027, announced by the Portuguese Environmental Agency, the scenarios could not predict improvement of ecological status beyond the "moderate" category. The study recommendations were therefore to prevent contamination through proper implementation and monitoring of existing watershed management plans. The adequate treatment of domestic effluents and the control of livestock farming residues are urgent.

An approach to validate groundwater contamination risk in rural mountainous catchments: the role of lateral groundwater flows.

Computer models dedicated to the validation of groundwater contamination risk in the rural environment, namely the risk of contamination by nitrate leachates from agriculture fertilizers, are frequently based on direct comparison of risky areas (e.g., cropland, pastures used for livestock production) and spatial distributions of contaminant (nitrate) plumes. These methods are fated to fail where lateral flows dominate in the landscape (mountainous catchments) displacing the nitrate plumes downhill and from the risky spots. In these cases, there is no connection between the spatial location of risky areas and nitrate plumes, unless the two locations can be linked by a contaminant transport model. The main purpose of this paper is therefore to introduce a method whereby spatio-temporal links can be demonstrated between risky areas (contaminant sources), actual nitrate plumes (contaminant sinks) and modeled nitrate distributions at specific groundwater travel times, thereby validating the risk assessment. The method assembles a couple of well known algorithms, namely the DRASTIC model [1,2] and the Processing Modflow software (https://www.simcore.com), but their combination as risk validation method is original and proved efficient in its initial application, the companion paper of Pacheco et al. [3].

A partial least squares - Path modeling analysis for the understanding of biodiversity loss in rural and urban watersheds in Portugal.

The main purpose of this study was to use Partial Least Squares - Path Modeling (PLS-PM) to quantify the contributions of natural and human-induced threats to biodiversity loss in rural and urban watersheds. The study area comprised the Sabor and Ave river basins, located in northern Portugal. The Sabor is rural and sparsely populated while the Ave is urbanized, industrialized and densely populated. Within PLS-PM, threats are called exogenous latent variables while the ultimate environmental consequence (biodiversity loss) is termed endogenous latent variable. Latent variables are concepts represented by numerical parameters called formative variables. The selected latent variables were given the names "pressures", "contamination" and "ecological integrity". The most important "pressures" were the wildfire risk, the percentage of urban area in sub-catchments, the diffuse emissions of livestock nitrogen (N) and agriculture/forest phosphorus (P), and the point source emissions of urban N, P and biochemical oxygen demand, as well as of industrial N. The latent variable called "contamination" was primarily represented by stream water concentrations of phosphate, suspended solids and dissolved oxygen. And finally, the "ecological integrity" was represented by the he North Invertebrate Portuguese Index. The results unequivocally showed that point source emissions in the Sabor (except industrial N) and stream water contamination in the Ave determine biodiversity loss. These contrasting influences suggest that Ave basin has evolved from a catchment where man once produced localized negative effects on stream ecological integrity (a condition still observed in the Sabor basin) to a catchment where the dense human occupation has covered the entire area with urban contaminant sources, somewhat generalizing the local effects. The attribution of local effects to biodiversity loss in the rural catchment and of regional effects in the urban catchment is confirmed by the results of a study covering the entire planet.

Rainwater harvesting in catchments for agro-forestry uses: A study focused on the balance between sustainability values and storage capacity.

Rainwater harvesting (RWH) is used to support small-scale agriculture and handle seasonal water availability, especially in regions where populations are scattered or the costs to develop surface or groundwater resources are high. However, questions may arise as whether this technique can support larger-scale irrigation projects and in complement help the struggle against wildfires in agro-forested watersheds. The issue is relevant because harvested rainwater in catchments is usually accumulated in small-capacity reservoirs created by small-height dams. In this study, a RWH site allocation method was improved from a previous model, by introducing the dam wall height as evaluation parameter. The studied watershed (Sabor River basin) is mostly located in the Northeast of Portugal. This is a rural watershed where agriculture and forestry uses are dominant and where ecologically relevant regions (e.g., Montezinho natural park) need to be protected from wildfires. The study aimed at ranking 384 rainfall collection sub-catchments as regards installation of RWH sites for crop irrigation and forest fire combat. The height parameter was set to 3m because this value is a reference to detention basins that hold sustainability values (e.g., landscape integration, environmental protection), but the irrigation capacity under these settings was smaller than 10ha in 50% of cases, while continuous arable lands in the Sabor basin cover on average 222ha. Besides, the number of sub-catchments capable to irrigate the average arable land was solely 7. When the dam wall height increased to 6 and 12m, the irrigation capacity increased to 46 and 124 sub-catchments, respectively, meaning that more engineered dams may not always ensure all sustainability values but warrant much better storage. The limiting parameter was the dam wall height because 217 sub-catchments were found to drain enough water for irrigation and capable to store it if proper dam wall heights were used.

Integrative assessment of river damming impacts on aquatic fauna in a Portuguese reservoir.

The impacts of river damming on aquatic fauna are assessed and then integrated across hierarchical scales, portrayed as nested circles. The outer circles characterize the studied site for habitat disturbance and hydrologic regime during the construction and filling phases. The inner circles characterize the construction and filling phase zones for water quality parameters and aquatic fauna. The procedure for integrating the impacts comprises two consecutive stages: in the task stage, inner circles are characterized through field and laboratory work involving sampling and analyses of water aliquots and aquatic fauna; in the impact assessment stage, circle data are processed in ecological and statistical algorithms, which allow identification of changes in abundance and composition of aquatic fauna communities and their integration with changes in water quality parameters across the construction and filling phase zones. The integrative assessment of river damming impacts on aquatic fauna was carried out in the Sabor River dam (Portugal). This dam created two reservoirs: primary and secondary. Changes in water quality caused by dam construction and stream water impoundment were significant, marked by increases in temperature and electric conductivity downstream, accumulation of phosphorus and nitrogen in the reservoirs triggering the growth of algae and the increase of chlorophyll a, and drop of transparency. These changes were aggravated in the secondary reservoir. The consequences of water deterioration for aquatic fauna were severe, marked by abrupt declines of native fish species and invasion of exotic species even upwards the reservoirs. The ecological status determined from ecological quality ratios of benthic macroinvertebrate assemblages were also affected, changing from good-fair in the unaffected watercourses to fair-poor in the lakes.

The impact of freshwater metal concentrations on the severity of histopathological changes in fish gills: A statistical perspective.

The purpose of this study was to relate the severity of histopathological changes in fish gills with changes in metal concentrations of freshwater samples, and to use the relationships as premature warnings of impairment in aquatic fauna populations. The investigated species were the native barbel (Luciobarbus bocagei) and boga (Pseudochondrostoma sp.), and the introduced trout (Oncorhynchus mykiss), collected from 6 northern Portuguese rivers in a total of 249 individuals. The sampling sites have been linked to different ecological status by the official authorities. The sampling has been repeated 4 times to cover different hydrologic and environmental conditions. The analyzed metals were aluminum, arsenic, cadmium, cobalt, chromium, copper, manganese, nickel, lead and zinc. For each fish, 30 filaments of a gill arch were observed in a light microscope, and the histopathological changes evaluated according to a 6-degree gradation scale that combines the extent and severity of each lesion. The relationships between the histopathological and the chemical results were investigated by the non-parametric Goodman Kruskal gamma correlation and Partial Least Squares regression (PLS). The statistical results highlighted the importance of filament epithelium proliferation (FEP) as key biomarker to the toxicity of sub lethal concentrations of metals, because FEP was significantly correlated with all analyzed metals and explained through PLS regression by concentration changes of Cu, Zn, Mn, Cr and As. A refined regression analysis, where histopathological data on the 3 species were processed in separate, revealed that FEP severity is especially sensitive to changes in metal concentrations in boga. Thus, monitoring studies on the ecological status of northern Portuguese rivers would benefit in time and cost if FEP is used as biomarker and boga as species. Naturally, the option for this species depends on the availability of boga individuals along the stream reaches selected for the monitoring programs.

Assessing anthropogenic impacts on riverine ecosystems using nested partial least squares regression.

The results of three Partial Least Squares (PLS) regression models were used to gain a holistic view on the consequences of natural processes and anthropogenic pressures for water quality degradation and biodiversity decline in a multi-use watershed. The processes were soil erosion and wildfire risk; the pressures comprised land use conflicts, leachates from domestic and industrial waste, arable farming intensity and livestock density. Water quality was characterized for concentrations of nutrients (nitrate, phosphate), oxygen demands (Biochemical Oxygen Demand - BOD5, Chemical Oxygen Demand - COD) and various metals (e.g., As, Cr). Ecological integrity was assessed by the recently developed MELI (Multiple Ecological Level Index). In total, 18 variables were processed in the regression models. Two models were called "nested models" because they dealt with initial (pressures), intermediate (water quality) and final (MELI) environmental descriptors, used as dependent (MELI, quality) or independent (quality, pressures) variables. The third was called "bypass model" because it dealt solely with initial and final descriptors. Overall, the results of PLS regression linked the ineffective treatment of domestic sewage to water quality and ecological integrity declines in the studied watershed. Put another way, all models recurrently affirmed the major role of local factors, meaning of point source pollution, in determining the quality of stream water and the integrity of freshwater ecosystems. Sources of diffuse pollution were accounted for as contributing factors in the PLS regressions, but their influence was scarcely perceptible in the results. The poor treatment of domestic effluents is a public concern. In their strategic plans for mitigating this problem in the forthcoming years, administrative authorities are concentrated on management initiatives to improve the quality of provided services, instead of considering the construction of new wastewater treatment plants.

The role of environmental land use conflicts in soil fertility: A study on the Uberaba River basin, Brazil.

In the Uberaba River basin (state of Minas Gerais, Brazil), pastures for livestock production have invaded areas of native vegetation (Cerrado biome), while already existing pastures were invaded by crop agriculture, with an expansion of sugar cane plantations in the most recent years. In some areas of the basin, these land use changes were classified as environmental land use conflicts because the new uses were not conforming to land capability, i.e. the soil's natural use. Where the areas in conflict became dense, some soil properties have changed significantly, namely the organic matter content and the exchangeable potassium concentration, which have decreased drastically (5kg/m(3) per 10% increase in the conflict area) threatening the fertility of soil. Besides, these changes may have triggered a cascade of other environmental damages, specifically the increase of soil erosion and the degradation of water quality with negative impacts on aquatic biodiversity, related to a disruption of soil organic matter structural functions. Because half the Uberaba catchment has been considered is a state of accentuated environmental degradation, not only caused by environmental land use conflicts, conservation measures have been proposed and requested for immediate implementation across the watershed.

From catchment to fish: Impact of anthropogenic pressures on gill histopathology.

Gill histopathology was investigated in barbel (Luciobarbus bocagei) and nase (Pseudochondrostoma sp.) in sub-catchments of Paiva River (Portugal) located upstream and downstream of a Waste Water Treatment Plant (WWTP). Multivariate statistical analyses were performed to set up correlations between the species sample (n=24) and injury types (8). The results discriminate well edema and vasodilatation between reference (upstream) and disturbed (downstream) samples. Using a watershed model, time series of physico-chemical parameters and heavy metal concentrations were calibrated and validated for the entire Paiva River basin as to investigate the relationship between water quality and the gill histopathology results. Increased concentrations of heavy metal downstream, specifically of zinc and lead, coincided with a higher severity of histopathological alterations in the fish gills. Significant but less evident relationship between water quality parameters and severity of gill injuries in the analyzed fish species were also observed for fecal coliforms, water temperature and manganese. Notwithstanding the location of the samples upstream and downstream of the WWTP, contamination of Paiva River and its effect on gill injuries cannot be disconnected from other punctual and diffuse pollution sources acting in different sectors within the watershed, namely agriculture and forest management. The severity of histopathological alterations in the fish gills reflected differences in the type and concentration of contaminants in Paiva River, and consequently can be viewed as valuable indicator of water quality.

Environmental land use conflicts in catchments: A major cause of amplified nitrate in river water.

Environmental land use conflicts are uses of the land that ignore soil capability. In this study, environmental land use conflicts were investigated in mainland Portugal, using Partial Least Squares (PLS) regression combined with GIS modeling and a group of 85 agricultural watersheds (with >50% occupation by agriculture) as work sample. The results indicate a dominance of conflicts in a region where vineyards systematically invaded steep hillsides (the River Douro basin), where forests would be the most appropriate use. As a consequence of the conflicts, nitrate concentrations in rivers and lakes from these areas have increased, sometimes beyond the legal limit of 50mg/L imposed by the European and Portuguese laws. Excessive nitrate concentrations were also observed along the Atlantic coast of continental Portugal, but associated to a combination of other factors: large population densities, and incomplete coverage by sewage systems and inadequate functioning of wastewater treatment plants. Before this study, environmental land use conflicts were never recognized as possible boost of nitrate concentrations in surface water. Bearing in mind the consequences of drinking water nitrate for human health, a number of land use change scenarios were investigated to forecast their impact on freshwater nitrate concentrations. It was seen that an aggravation of the conflicts would duplicate the number of watersheds with maximum nitrate concentrations above 50mg/L (from 11 to 20 watersheds), while the elimination of the conflicts would greatly reduce that number (to 3 watersheds).

Controls and forecasts of nitrate yields in forested watersheds: A view over mainland Portugal.

A study on nitrate yields was conducted in forested watersheds of mainland Portugal. The prime goal was to rank parameters in descending order of their contribution to the export of nitrate towards streams and lakes. To attain the goal, variables like soil loss, rainfall intensity, topography, soil type, forest composition and environmental disturbances such as hardwood harvesting or wildfires were organized in a conceptual yield model. Because some parameters were potentially collinear, a robust multivariate statistical technique was selected to execute the conceptual model and perform the aforementioned ranking, namely Partial Least Squares (PLS) regression. This technique was tested with a sample of 60 forested watersheds (>70% of forest occupation), being subject to a double-validation process to ensure prediction capability. According to final regression coefficients, soil erosion seems to regulate nitrate distribution across the basins, because soil loss and type, rainfall intensity and topography explained around 60% of nitrate yield variance. The major importance of erosion is followed by a moderate role of biochemical processes such as nitrification or nutrient uptake, which accounted for approximately 15% of nitrate yield variance. In this case, deciduous forests and scrubland seem to behave as net sinks of nitrate while coniferous and mixed forests seem to act dually, as net sources or sinks. The least important parameters are the environmental disturbances, explaining no more than 5% of nitrate yield variance. The results of PLS regression were coupled in a scenario analysis with measures designed to protect soil from erosion and surface water from eutrophication. These interventions are to be implemented until 2045, according to regional plans of forest management. Considering the key role of erosion in explaining nitrate dynamics across the catchments, it was not surprising to verify that soil protection measures may reduce nitrate yields by some 35% of their current values.

A framework model for investigating the export of phosphorus to surface waters in forested watersheds: Implications to management.

The present study was developed in four sub-basins of rivers Cávado and Douro, located in the North of mainland Portugal. The goal was to identify main stressors as well as driving and attenuating processes responsible for the presence of phosphorus in masses of surface water in those catchments. To accomplish the goal, the basins were selected where a quality station was present at the outlet, the forest occupation was greater than 75% and the phosphorus concentrations have repeatedly exceeded the threshold for the good ecological status in the period 2000-2006. Further, in two basins the quality station was installed in a lotic (free-flow water) environment whereas in the other two was placed in a lentic (dammed water) environment. The ArcMap GIS-based software package was used for the spatial analysis of stressors and processes. The yields of phosphorus vary widely across the studied basins, from 0.2-30 kg·ha(-1)·yr(-1). The results point to post-fire soil erosion and hardwood clear cuttings as leading factors of phosphorus exports across the watersheds, with precipitation intensity being the key variable of erosion. However, yields can be attenuated by sediment deposition along the pathway from burned or managed areas to water masses. The observed high yields and concentrations of phosphorus in surface water encompass serious implications for water resources management in the basins, amplified in the lentic cases by potential release of phosphorus from lake sediments especially during the summer season. Therefore, a number of measures were proposed as regards wildfire combat, reduction of phosphorus exports after tree cuts, attenuation of soil erosion and improvement of riparian buffers, all with the purpose of preventing phosphorus concentrations to go beyond the regulatory good ecological status.

Water resources planning for a river basin with recurrent wildfires.

Situated in the north of Portugal, the Beça River basin is subject to recurrent wildfires, which produce serious consequences on soil erosion and nutrient exports, namely by deteriorating the water quality in the basin. In the present study, the ECO Lab tool embedded in the Mike Hydro Basin software was used for the evaluation of river water quality, in particular the dissolved concentration of phosphorus in the period 1990-2013. The phosphorus concentrations are influenced by the burned area and the river flow discharge, but the hydrologic conditions prevail: in a wet year (2000, 16.3 km(2) of burned area) with an average flow of 16.4 m(3)·s(-1) the maximum phosphorus concentration was as low as 0.02 mg·L(-1), while in a dry year (2005, 24.4 km(2) of burned area) with an average flow of 2 m(3)·s(-1) the maximum concentration was as high as 0.57 mg·L(-1). Phosphorus concentrations in the water bodies exceeded the bounds of good ecological status in 2005 and between 2009 and 2012, water for human consumption in 2009 and water for multiple uses in 2010. The River Covas, a right margin tributary of Beça River, is the most appropriate stream as regards the use of water for human consumption, because it presents the biggest water potential with the best water quality. Since wildfires in the basin result essentially from natural causes and climate change forecasts indicate an increase in their frequency and intensity in the near future, forestry measures are proposed to include as a priority the conversion of stands of maritime pine in mixed stands of conifer and hardwood species.

Impacts of climate change and land-use scenarios on Margaritifera margaritifera, an environmental indicator and endangered species.

In this study, we assess the impacts of future climate and land-use in the Beça River (northern Portugal) under different scenarios and how this will translate into the conservation status of the endangered pearl mussel Margaritifera margaritifera (Linnaeus, 1758). This species is currently present in several stretches of the Beça River that still hold adequate ecological conditions. However, the species is threatened by projected declines in precipitation for the 21st century, with implication on the river flows and water depths that might decrease below the species requisites. This situation could be especially critical during summer conditions since the ecological flows may not be assured and several river stretches may be converted into stagnant isolated pools. The habitat connectivity will also be affected with reverberating effects on the mobility of Salmo trutta, the host of M. margaritifera, with consequences in the reproduction and recruitment of pearl mussels. In addition, human-related threats mostly associated with the presence of dams and an predicted increases in wildfires in the future. While the presence of dams may decrease even further the connectivity and river flow, with wildfires the major threat will be related to the wash out of burned areas during storms, eventually causing the disappearance of the mussels, especially the juveniles. In view of future climate and land-use change scenarios, conservation strategies are proposed, including the negotiation of ecological flows with the dam promoters, the replanting of riparian vegetation along the water course and the reintroduction of native tree species throughout the catchment.

Factor weighting in DRASTIC modeling.

Evaluation of aquifer vulnerability comprehends the integration of very diverse data, including soil characteristics (texture), hydrologic settings (recharge), aquifer properties (hydraulic conductivity), environmental parameters (relief), and ground water quality (nitrate contamination). It is therefore a multi-geosphere problem to be handled by a multidisciplinary team. The DRASTIC model remains the most popular technique in use for aquifer vulnerability assessments. The algorithm calculates an intrinsic vulnerability index based on a weighted addition of seven factors. In many studies, the method is subject to adjustments, especially in the factor weights, to meet the particularities of the studied regions. However, adjustments made by different techniques may lead to markedly different vulnerabilities and hence to insecurity in the selection of an appropriate technique. This paper reports the comparison of 5 weighting techniques, an enterprise not attempted before. The studied area comprises 26 aquifer systems located in Portugal. The tested approaches include: the Delphi consensus (original DRASTIC, used as reference), Sensitivity Analysis, Spearman correlations, Logistic Regression and Correspondence Analysis (used as adjustment techniques). In all cases but Sensitivity Analysis, adjustment techniques have privileged the factors representing soil characteristics, hydrologic settings, aquifer properties and environmental parameters, by leveling their weights to ≈4.4, and have subordinated the factors describing the aquifer media by downgrading their weights to ≈1.5. Logistic Regression predicts the highest and Sensitivity Analysis the lowest vulnerabilities. Overall, the vulnerability indices may be separated by a maximum value of 51 points. This represents an uncertainty of 2.5 vulnerability classes, because they are 20 points wide. Given this ambiguity, the selection of a weighting technique to integrate a vulnerability index may require additional expertise to be set up satisfactorily. Following a general criterion that weights must be proportional to the range of the ratings, Correspondence Analysis may be recommended as the best adjustment technique.

Groundwater quality in rural watersheds with environmental land use conflicts.

The quality of groundwater was evaluated in a rural watershed of northern Portugal (River Sordo basin) where environmental land use conflicts have developed in the course of a progressive invasion of forest and pasture lands by agriculture, especially by vineyards. The selected groundwater quality parameters were the concentrations of sodium, calcium, bicarbonate, chloride and nitrates, derived from natural and anthropogenic sources. The environmental land use conflicts were revealed by the coupling of land use and land capability raster maps. The land capability evaluation allocated 70.3% of the basin to the practicing of agriculture, 20% to livestock pasturing and 9.7% to a mosaic of land uses including agriculture, livestock pasturing and forestry. The assessment of land use conflicts allocated 93.9% of the basin to no conflict areas. Minor conflict areas (4.1%) were found concentrated in the western region of the watershed. They correspond to an invasion of farmlands towards sectors of the catchment capable for the practicing of livestock pasturing. Moderate (1.6%) and major (0.4%) conflict areas were found limited to the eastern region, matching steep hillsides capable for the practicing of livestock pasturing or forestry but presently occupied with vineyards. The spatial distributions of ion concentrations were generally justified by common geochemical processes. The dominance of high concentration levels in moderate and major conflict areas was justified within the framework of nutrient dynamics in vineyard environment. Nitrate in groundwater was likewise produced via the nitrification of N-fertilizers. Apparently, this process promoted the weathering of plagioclase by the nitric acid reaction, in concurrence with the weathering by the carbonic acid reaction. The impact of nitrification was found more important in moderate and major conflict areas, relative to no conflict areas.

Soil losses in rural watersheds with environmental land use conflicts.

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 km(2) (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.