The role of ponds in pesticide dissipation at the catchment scale: The case of the Save agricultural catchment (Southwestern France).

Pesticides are a major source of pollution for ecosystems. In agricultural catchments, ponds serve as buffer areas for pesticide transfers and biogeochemical hotspots for pesticide dissipation. Some studies have highlighted the specific impact of ponds on the dynamics of pesticides, but knowledge of their cumulative effect at the watershed scale is scarce. Hence, using a modelling approach, we assessed the cumulative role of ponds in pesticide transfer in an agricultural basin (Southwest of France, 1110 km2). The Soil and Water Assessment Tool (SWAT) model was used to model the Save basin, including 197 ponds selected with a Multi-Criteria Decision Aiding Model based on their pesticide interception capacities. The daily discharge, the suspended sediment loads and two herbicide loads (i.e. S-metolachlor and aclonifen) in dissolved and particulate phases were accurately simulated from January 2002 to July 2014 at a daily time step. The presence of ponds resulted in a yearly mean reduction at the watershed outlet of respectively 61 % and 42 % of aclonifen and S-metolachlor fluxes compared to the simulations in the absence of ponds. Sediment-related processes were the most efficient for pesticide dissipation, leading to a mean dissipation efficiency by ponds of 51.0 % for aclonifen and 34.4 % for S-metolachlor. This study provides a first quantification of the cumulative role of ponds in pesticide transfer at the catchment scale in an intensive agricultural catchment.

Assessing fluvial organic carbon flux and its response to short climate variability and damming on a large-scale tropical Asian river basin.

Fluvial organic carbon (OC) transfer is an essential resource for downstream ecosystems. Multiple factors affect its transfer process, e.g., climate or anthropogenic activities. Quantifying OC fluxes with fine spatiotemporal resolution is challenging in anthropised catchments. This study aims to quantify daily OC dynamics and to assess the impacts of short climate variability and damming on OC spatiotemporal transfer processes in a large tropical Asian river basin (the Red River) for an extended period (2003-2013) by combining empirical equations with modelling outputs. Firstly, empirical equations for calculating dissolved (DOC) and particulate OC (POC) concentrations were calibrated based on in-situ sampling data. Then, simulated daily discharge (Q) and suspended sediment concentrations were used to quantify daily OC fluxes. Results show that the parameters of the DOC and POC equations well represent the subbasins characteristics, underlining the effects of soil OC content, mean annual Q and Chlorophyll a. DOC and POC exports reached 222 and 406 kt yr-1 at the basin outlet, accounting for 0.38 % of the total OC (TOC) exported by Asian rivers to the ocean. However, the specific yields of DOC (1.62 t km-2 yr-1) and POC (2.96 t km-2 yr-1) of the Red River basin were ~ 1.5 times those of other Asian basins. By comparing a reference scenario (without dams) to current conditions, we estimated 12 % and 88 % decreases in DOC and POC fluxes between 2008-2013 and 2003-2007, mainly due to damming. This study shows that climate variability may not impact OC dynamics in rivers as it explained <2 % of the variations. However, dam management, especially recent ones operating since 2008, deeply influences OC variations as the POC/TOC ratio decreased from 86 % to 47 %. Damming significantly decreased POC exports due to sediment retention, altering the equilibrium of OC cycling downstream, which may impact the food chain.

Identification of critical effect factors for prediction of spatial and intra-annual variability of shallow groundwater nitrate in agricultural areas.

Shallow groundwater nitrate nitrogen (NO3--N) concentrations in agricultural areas usually show high spatial and intra-annual variability. It is hard to predict such concentrations due to the complexity of influencing factors (e.g., different forms of N in soil, vadose zone characteristics, and groundwater physiochemical conditions). Here, a large number of groundwater and soil samples were collected monthly over two years at 14 sites to analyze the soil and groundwater physiochemical properties and the stable isotopes of δ15N and δ18O of groundwater NO3--N in agricultural areas. Based on field observations, a random forest (RF) model was used to predict the groundwater NO3--N concentrations and reveal the importance of effect factors. The results show that there are large spatiotemporal variations in NO3--N, δ15N-NO3-, and δ18O-NO3- in groundwater. NO3--N is the major dominant specie of inorganic N in groundwater, and the groundwater NO3--N concentration in 24 % of the samples failed to meet the drinking water standard of the WHO (10 mg L-1). The RF model satisfactorily predicted groundwater NO3--N concentrations with R2 of 0.90-0.94, RMSE of 4.54-5.07, and MAE of 2.17-3.38. Groundwater nitrite and ammonium are the most important factors related to NO3--N consumption and production, respectively, in groundwater. Denitrification and nitrification were further identified by the relationships among δ15N-NO3-, δ18O-NO3-, and NO3--N, and by the ranges of δ15N-NO3-, δ18O-NO3-, temperature, pH, DO, and ORP in groundwater. Soil-soluble organic nitrogen (S-SON) and the depth of groundwater table were identified as vital factors related to N sourcing and leaching. Overall, as a first approach to adopting a RF model for high spatiotemporal-resolution prediction of groundwater NO3--N variations, the findings of this study enable a better understanding of groundwater N pollution in agricultural areas. Optimizing management of irrigation and N inputs is anticipated to reduce S-SON accumulation and mitigate the threat to groundwater quality in agricultural areas.

Global carbon sequestration through continental chemical weathering in a climatic change context.

This study simulates carbon dioxide (CO2) sequestration in 300 major world river basins (about 70% of global surface area) through carbonates dissolution and silicate hydrolysis. For each river basin, the daily timescale impacts under the RCP 2.6 and RCP 8.5 climate scenarios were assessed relative to a historical baseline (1969-1999) using a cascade of models accounting for the hydrological evolution under climate change scenarios. Here we show that the global temporal evolution of the CO2 uptake presents a general increase in the annual amount of CO2 consumed from 0.247 ± 0.045 Pg C year-1 to 0.261 and 0.273 ± 0.054 Pg C year-1, respectively for RCP 2.6 and RCP 8.5. Despite showing a general increase in the global daily carbon sequestration, both climate scenarios show a decrease between June and August. Such projected changes have been mapped and evaluated against changes in hydrology, identifying hot spots and moments for the annual and seasonal periods.

Evolution of N-balance with qualitative expert evaluation approach.

Pollution of rivers by nitrate is a major issue. Many land use units are considered as net nitrate producers when the input dominates the uptake (e.g. agricultural areas), or in the opposite, net consumers (e.g. wetlands), but the role of their spatial organization and temporal dynamics together across the watershed is unclear. Here, we used a Nitrate-related Ecological Functions (NEF) concept, together with an expert-based analysis in a Geographical Information System, to investigate the role of two opposite landscape types in the nitrate regulation across the Garonne river watershed (France). At any point in a watershed, there is nitrate production (NP) and nitrate removal (NR). The nitrate net balance (NNB) between NP and NR functions can be neutral (NB, Neutral Balance) when nitrate fluxes balance over space and time. The first landscape type, called Actual, was obtained using a set of 7 actual environmental variables, as land cover types, soil organic matter content and wetlands presence. The second landscape type, called Natural, described a non-anthropized landscape, using the same layer types as the Actual landscape. Potentials in NP and NR for each class in each map layer were rated by a set of experts according to their scientific knowledge. NP, NR and by difference, NNB maps were obtained, overlaid and compared to provide an evaluation of the potential for each landscape. In both landscapes, NNB were largely balanced (Actual = 48% and Natural = 67%). In the Actual landscape, NNB were secondly dominated by an imbalance toward NP (43%) and in the Natural landscape secondly imbalanced toward NR (32%). We constructed 'maps of disagreement' between both landscapes to provide a spatially explicit assessment of NNB evolution caused by changing land cover. We found that 67% of the agricultural areas and 60% of the artificial areas of the watershed had been subjected to a loss of nitrate ecological functions from Natural to Actual landscapes. Some management practices able to modify these factors may improve ecological functions and diminish the NEF disagreement of the watershed.

Denitrification in wetlands: A review towards a quantification at global scale.

Research to understand the nitrogen cycle has been thriving. The production of reactive nitrogen by humans exceeds the removal capacity through denitrification of any natural ecosystem. The surplus of reactive nitrogen is also a significant pollutant that can shift biological diversity and distribution, promotes eutrophication in aquatic ecosystems, and affects human health. Denitrification is the microbial respiration in anoxic conditions and is the main process that removes definitively nitrates from the ecosystem by returning of reactive nitrogen (Nr) to the atmosphere as N2 and N2O emissions. This process occurs in the oceans, aquatic ecosystems and temporary flooded terrestrial ecosystems. Wetlands ecosystems are rich in organic matter and they have regular anoxic soil conditions ideal for denitrification to occur. In the current paper, we provide a meta-analysis that aims at exploring how research around global nitrogen, denitrification and wetlands had evolved in the last fifty years. Back in the time, wetland ecosystems were seen as non-exploitable elements of the landscape, and now they are being integrated as providers of ecosystem services. A significant improvement of molecular biology techniques and genetic extraction have made the denitrification process fully understood allowing constructed wetlands to be more efficient and popular. Yet, large uncertainties remain concerning the dynamic quantification of the global denitrification capacity of natural wetland ecosystems. The contribution of the current investigation is to provide a way forward for reducing these uncertainties by the integration of satellite-based Earth Observation (EO) technology with parsimonious physical based models.

Applications of a SWAT model to evaluate the contribution of the Tafna catchment (north-west Africa) to the nitrate load entering the Mediterranean Sea.

An elevated nitrogen concentration in water is one of the main problems affecting water quality in Mediterranean rivers. The objectives of this study were (1) to evaluate the contribution of the Tafna catchment to the nitrate load entering the Mediterranean Sea, (2) to quantify the impact of agriculture on the nitrate concentration in water bodies, (3) to evaluate nitrate loads entering groundwater, and (4) to quantify the role of reservoirs in nitrate retention. A SWAT model was applied during the period 2003 to 2011. The discharge calibration was based on a previous study by Zettam et al. (2017). NSE efficiencies ranged from 0.421 to 0.75, R2 ranged from 0.25 to 0.84, and PBIAS ranged from 3.68 to 39.42. The simulations of monthly nitrate loads were satisfactory in the upstream sampling stations, with NSE between 0.48 and 0.65 and R2 between 0.63 and 0.68. The PBIAS was satisfactory in all the sampling stations (- 36.30 to 10.42). In the downstream sampling stations, the calibration of the monthly nitrate loads was unsatisfactory (NSE ranged from - 0.26 to 0.21 and R2 ranged from 0.02 to 0.25). Fertilisation was the main N input in the catchment, while the main N output was plant uptake. The Tafna River carried an annual average of 37 to 85.5 t N year-1 into the Mediterranean Sea. The simulation also showed that reservoirs in the Tafna basin contain a large quantity of nitrates, i.e. 62% on average of the total amount of nitrates transported annually by the Tafna River.

Can Recent Global Changes Explain the Dramatic Range Contraction of an Endangered Semi-Aquatic Mammal Species in the French Pyrenees?

Species distribution models (SDMs) are the main tool to predict global change impacts on species ranges. Climate change alone is frequently considered, but in freshwater ecosystems, hydrology is a key driver of the ecology of aquatic species. At large scale, hydrology is however rarely accounted for, owing to the lack of detailed stream flow data. In this study, we developed an integrated modelling approach to simulate stream flow using the hydrological Soil and Water Assessment Tool (SWAT). Simulated stream flow was subsequently included as an input variable in SDMs along with topographic, hydrographic, climatic and land-cover descriptors. SDMs were applied to two temporally-distinct surveys of the distribution of the endangered Pyrenean desman (Galemys pyrenaicus) in the French Pyrenees: a historical one conducted from 1985 to 1992 and a current one carried out between 2011 and 2013. The model calibrated on historical data was also forecasted onto the current period to assess its ability to describe the distributional change of the Pyrenean desman that has been modelled in the recent years. First, we found that hydrological and climatic variables were the ones influencing the most the distribution of this species for both periods, emphasizing the importance of taking into account hydrology when SDMs are applied to aquatic species. Secondly, our results highlighted a strong range contraction of the Pyrenean desman in the French Pyrenees over the last 25 years. Given that this range contraction was under-estimated when the historical model was forecasted onto current conditions, this finding suggests that other drivers may be interacting with climate, hydrology and land-use changes. Our results imply major concerns for the conservation of this endemic semi-aquatic mammal since changes in climate and hydrology are expected to become more intense in the future.

Water age prediction and its potential impacts on water quality using a hydrodynamic model for Poyang Lake, China.

The water quality in Poyang Lake, the largest freshwater lake in China, has deteriorated steadily in recent years and local governments have made efforts to manage the potential eutrophication. In order to investigate the transport and retention processes of dissolved substances, the hydrodynamic model, Environmental Fluid Dynamics Code (EFDC) was applied by using the concept of water age. The simulated results showed agreement with the measured water level, discharge, and inundation area. The water age in Poyang Lake was significantly influenced by the variations of hydrological conditions. The annual analysis revealed that the largest averaged water age was observed during the wet year (2010) with 28.4 days at Hukou, the junction of the Yangtze River and Poyang Lake. In the normal season (April), the youngest age with 9.1 days was found. The spatial distribution of water quality derived from the remote sensing images suggested that a higher chlorophyll-a concentration, lower turbidity, and smaller water age in the eastern area of Poyang Lake might threaten the regional aquatic health. The particle tracking simulation reproduced the trajectories of the dissolved substances, indicating that the water mass with greater nutrient loading would further lead to potential environmental problems in the east lake. Moreover, the water transfer ability would be weakened due to dam (Poyang Project) construction resulting in the rising water levels in periods of regulation. Generally, this study quantified an indicative transport timescale, which could help to better understand the complex hydrodynamic processes and manage wetland ecosystems similar to Poyang Lake.

New insight into pesticide partition coefficient Kd for modelling pesticide fluvial transport: application to an agricultural catchment in south-western France.

Pesticides applied on crops are leached with rainfall to groundwater and surface water. They threat the aquatic environment and may render water unfit for human consumption. Pesticide partitioning is one of the pesticide fate processes in the environment that should be properly formalised in pesticide fate models. Based on the analysis of 7 pesticide molecules (alachlor, atrazine, atrazine's transformation product deethylatrazine or DEA, isoproturon, tebuconazole and trifluralin) sampled from July 2009 to October 2010 at the outlet of the river Save (south-western France), the objectives of this study were (1) to check which of the environmental factors (discharge, pH, concentrations of total suspended matter (TSM), dissolved organic carbon (DOC) and particulate organic carbon (POC) could control the pesticide sorption dynamic, and (2) to establish a relationship between environmental factors, the partition coefficient Kd and the octanol/water distribution coefficient Kow. The comparison of physico-chemical parameters values during low flow and high flow shows that discharge, TSM and POC are the factors most likely controlling the pesticide sorption processes in the Save river network, especially for lower values of TSM (below 13mgL(-1)). We therefore express Kd depending on the widely literature-related variable Kow and on the commonly simulated variable TSM concentration. The equation can be implemented in any model describing the fluvial transport and fate of pesticides in both dissolved and sorbed phases, thus, Kd becomes a variable in time and space. The Kd calculation method can be applied to a wide range of catchments and organic contaminants.

Simulating land management options to reduce nitrate pollution in an agricultural watershed dominated by an alluvial aquifer.

The study area (Alegria watershed, Basque Country, Northern Spain) considered here is influenced by an important alluvial aquifer that plays a significant role in nitrate pollution from agricultural land use and management practices. Nitrates are transported primarily from the soil to the river through the alluvial aquifer. The agricultural activity covers 75% of the watershed and is located in a nitrate-vulnerable zone. The main objective of the study was to find land management options for water pollution abatement by using model systems. In a first step, the SWAT model was applied to simulate discharge and nitrate load in stream flow at the outlet of the catchment for the period between October 2009 and June 2011. The LOADEST program was used to estimate the daily nitrate load from measured nitrate concentration. We achieved satisfactory simulation results for discharge and nitrate loads at monthly and daily time steps. The results revealed clear variations in the seasons: higher nitrate loads were achieved for winter (20,000 kg mo NO-N), and lower nitrate loads were simulated for the summer (<1000 kg mo NO-N) period. In a second step, the calibrated model was used to evaluate the long-term effects of best management practices (BMPs) for a 50-yr period by maintaining actual agricultural practices, reducing fertilizer application by 20%, splitting applications (same total N but applied over the growing period), and reducing 20% of the applied fertilizer amount and splitting the fertilizer doses. The BMPs were evaluated on the basis of local experience and farmer interaction. Results showed that reducing fertilizer amounts by 20% could lead to a reduction of 50% of the number of days exceeding the nitrate concentration limit value (50 mg L) set by the European Water Framework Directive.

Daily nitrate losses: implication on long-term river quality in an intensive agricultural catchment of southwestern france.

High nitrate concentrations in streams have become a widespread problem throughout Europe in recent decades, damaging surface water and groundwater quality. The European Nitrate Directive fixed a potability threshold of 50 mg L for European rivers. The performance of the Soil and Water Assessment Tool model was assessed in the 1110-km Save catchment in southwestern France for predicting water discharge and nitrate loads and concentrations at the catchment outlet, considering observed data set uncertainty. Simulated values were compared with intensive and extensive measurement data sets. Daily discharge fitted observations (Nash-Sutcliffe efficiency coefficient = 0.61, = 0.7, and PBIAS = -22%). Nitrate simulation (1998-2010) was within the observed range (PBIAS = 10-21%, considering observed data set uncertainty). Annual nitrate load at the catchment outlet was correlated to the annual water yield at the outlet ( = 0.63). Simulated annual catchment nitrate exportation ranged from 21 to 49 kg ha depending on annual hydrological conditions (average, 36 kg ha). Exportation rates ranged from 3 to 8% of nitrogen inputs. During floods, 34% of the nitrate load was exported, which represented 18% of the 1998-2010 period. Average daily nitrate concentration at the outlet was 29 mg L (1998-2010), ranging from 0 to 270 mg L. Nitrate concentration exceeded the European 50 mg L potability threshold during 244 d between 1998 and 2010. A 20% reduction of nitrogen input reduced crop yield by between 5 and 9% and reduced by 62% the days when the 50 mg L threshold was exceeded.

Interaction between local hydrodynamics and algal community in epilithic biofilm.

Interactions between epilithic biofilm and local hydrodynamics were investigated in an experimental flume. Epilithic biofilm from a natural river was grown over a 41-day period in three sections with different flow velocities (0.10, 0.25 and 0.40 m s(-1) noted LV, IV and HV respectively). Friction velocities u* and boundary layer parameters were inferred from PIV measurement in the three sections and related to the biofilm structure. The results show that there were no significant differences in Dry Mass and Ash-Free Dry Mass (g m(-2)) at the end of experiment, but velocity is a selective factor in algal composition and the biofilms' morphology differed according to differences in water velocity. A hierarchical agglomerative cluster analysis (Bray-Curtis distances) and an Indicator Species Analysis (IndVal) showed that the indicator taxa were Fragilaria capucina var. mesolepta in the low-velocity (u*. = 0.010-0.012 m s(-1)), Navicula atomus, Navicula capitatoradiata and Nitzschia frustulum in the intermediate-velocity (u*. = 0.023-0.030 m s(-1)) and Amphora pediculus, Cymbella proxima, Fragilaria capucina var. vaucheriae and Surirella angusta in the high-velocity (u*. = 0.033-0.050 m s(-1)) sections. A sloughing test was performed on 40-day-old biofilms in order to study the resistance of epilithic biofilms to higher hydrodynamic regimes. The results showed an inverse relationship between the proportion of detached biomass and the average value of friction velocity during growth. Therefore, water velocity during epilithic biofilm growth conditioned the structure and algal composition of biofilm, as well as its response (ability to resist) to higher shear stresses. This result should be considered in modelling epilithic biofilm dynamics in streams subject to a variable hydrodynamics regime.

Occurrence of metolachlor and trifluralin losses in the Save river agricultural catchment during floods.

Rising pesticide levels in streams draining intensively managed agricultural land have a detrimental effect on aquatic ecosystems and render water unfit for human consumption. The Soil and Water Assessment Tool (SWAT) was applied to simulate daily pesticide transfer at the outlet from an agriculturally intensive catchment of 1110 km(2) (Save river, south-western France). SWAT reliably simulated both dissolved and sorbed metolachlor and trifluralin loads and concentrations at the catchment outlet from 1998 to 2009. On average, 17 kg of metolachlor and 1 kg of trifluralin were exported at outlet each year, with annual rainfall variations considered. Surface runoff was identified as the preferred pathway for pesticide transfer, related to the good correlation between suspended sediment exportation and pesticide, in both soluble and sorbed phases. Pesticide exportation rates at catchment outlet were less than 0.1% of the applied amount. At outlet, SWAT hindcasted that (i) 61% of metolachlor and 52% of trifluralin were exported during high flows and (ii) metolachlor and trifluralin concentrations exceeded European drinking water standards of 0.1 µg L(-1) for individual pesticides during 149 (3.6%) and 17 (0.4%) days of the 1998-2009 period respectively. SWAT was shown to be a promising tool for assessing large catchment river network pesticide contamination in the event of floods but further useful developments of pesticide transfers and partition coefficient processes would need to be investigated.

Temporal variability of nitrate transport through hydrological response during flood events within a large agricultural catchment in south-west France.

The temporal variability of nitrate transport was monitored continuously in a large agricultural catchment, the 1110km(2) Save catchment in south-west France, from January 2007 to June 2009. The overall aim was to analyse the temporal transport of nitrate through hydrological response during flood events in the catchment. Nitrate loads and hysteresis were also analysed and the relationships between nitrate and hydro-climatological variables within flood events were determined. During the study period, 19 flood events were analysed using extensive datasets obtained by manual and automatic sampling. The maximum NO(3)(-) concentration during flood varied from 8.2mgl(-1) to 41.1mgl(-1) with flood discharge from 6.75m(3)s(-1) to 112.60m(3)s(-1). The annual NO(3)(-) loads in 2007 and 2008 amounted to 2514t and 3047t, respectively, with average specific yield of 2.5tkm(-12)yr(-1). The temporal transport of nitrate loads during different seasonal flood events varied from 12t to 909t. Nitrate transport during flood events amounted to 1600t (64% of annual load; 16% of annual duration) in 2007 and 1872t (62% of annual load; 20% of annual duration) in 2008. The level of peak discharge during flood events did not control peak nitrate concentrations, since similar nitrate peaks were produced by different peak discharges. Statistically strong correlations were found between nitrate transport and total precipitation, flood duration, peak discharge and total water yield. These four variables may be the main factors controlling nitrate export from the Save catchment. The relationship between nitrate and discharge (hysteresis patterns) investigated through flood events in this study was mainly dominated by anticlockwise behaviour.