Diffuse nitrate (NO3) contamination from intense agriculture adversely impacts freshwater ecosystems, and can also result in nitrate concentrations exceeding limits set in drinking water regulation, when receiving surface waters are used for drinking water production. Implementation of near-natural mitigation zones such as reactive swales or wetlands have been proven to be promising measures to reduce nitrate loads in agricultural drainage waters. However, the behavior of these systems at low temperatures and its dependence on systemdesign has not beenwell known until now. In this study, the behavior of a full-scale (length: 45 m) reactive swale treating drainage water from an agricultural watershed in Brittany (France), with high nitrate concentrations in the receiving river, was monitored for one season (6 months). As flow in this full-size field system is usually restricted to winter and spring months (December–May), it usually operates at lowwater temperatures of 5–10 WC. Tracer tests revealed shorter than designed retention times due to high inflows and preferential flow in the swale. Results show a correlation between residence time and nitrate reduction with low removal (<10%) for short residence times (<0.1 day), increasing to >25% at residence times >10 h (0.4 day). Performance was compared to results of two technical-scale reactive swales (length: 8 m) operated for 1.5 yearswith two different residence times (0.4 and 2.5 days), situated at a test site of the German Federal Environmental Agency in Berlin (Germany). Similar nitrate reduction was observed for comparable temperature and residence time, showing that up-scaling is a suitable approach to transferring knowledge gathered from technical-scale experiments to field conditions. For the design of new mitigation systems, one recommendation is to investigate carefully the expected inflow volumes in advance to ensure a sufficient residence time for effective nitrate reduction at low temperatures.

The wash out of agricultural auxiliary chemicals like fertilizer and pesticides via surface run-off or subsurface leaching into drainage systems or ground waters, which discharge into surface waters, presents an increasing risk for drinking water production and biodiversity in rivers and lakes. Mitigation zones are important measures to attenuate contamination at the source and relieve surface waters downstream. Under high flow conditions, as they occur during rainy seasons and snow melt, the effectiveness of such facilities is restricted due to bypass of untreated waters or very short contact times. This study of the Aquisafe 2 project focus on drainage water decontamination and examines mitigation zone designs with organic substrates for their potential to reduce a set of herbicides and nitrate (NO3-), concurrently and efficiently, at short hydraulic residence times (0.2 to 2.5 days) to prepare their implementation in contamination hot spots. The herbicides bentazone, atrazine and isoproturon were classified as most relevant for drinking water production. On the basis of comprehensive literature studies the organic substrates bark mulch and straw and the design of bioretention swales emerged to be of high potential for decontamination of drainage waters in mitigation zones. In laboratory scale studies the substrates were tested in degradation-, sorption- and leaching-experiments at temperatures around 21 °C for their potential to ensure long- lasting hydraulic permeability, denitrification and attenuation of the selected herbicides. The selected organic substrates provide a high and long term stable permeable conductivity to realize and maintain high flow. The effective porosity yielded around 0.45 and reduced within 1.5 years by only 25 %. Straw is a readily available organic carbon source, which can support effective and efficient denitrification at short hydraulic residence times. Bark mulch contains more resistant carbon species, but contributes also to NO3- removal. In mixture with straw the performance of bark mulch as organic carbon source for denitrification increases (co-metabolic decomposition). Organic substrates are characterized by strong wash out of dissolved organic carbon (DOC) and high denitrification rates (15 to 45 g-N m-3 d-1) in the start phase and successive decrease of denitrification performance due to loss of readily available organic carbon. Despite decline of performance, denitrification rates stabilized after one year of operation at constant conditions at a level of 4 to 10 g-N m-3 d-1 (10 to 25 % of input). The potential of the organic substrate to retain the selected herbicides is very different for each compound and bases on different dissipation paths. Denitrifying conditions are in general disadvantageous for retention of the selected herbicides. Bentazone is too persistent and mobile to be considerably retained under high flow conditions. Atrazine can be substantially removed from drainage waters. It is suspected to be attenuated predominantly by formation of bound residues at the organic substrate, especially bark mulch, and partially by degradation to hydroxy-atrazine. Isoproturon seems to be effectively retained under suboxic conditions by degradation to metabolites. At technical scale parallel retention of NO3- and atrazine and NO3- and isoproturon was investigated. The potential of the organic carbon source (mixture of bark mulch and

Diffuse nitrate (NO3) contamination from intense agriculture adversely impacts freshwater ecosystems, and can also result in nitrate concentrations exceeding limits set in drinking water regulation, when receiving surface waters are used for drinking water production. Implementation of near-natural mitigation zones such as reactive swales or wetlands have been proven to be promising measures to reduce nitrate loads in agricultural drainage waters. However, the behavior of these systems at low temperatures and its dependence on system design is not well known until now. In this study, the behavior of a full scale (length: 45 m) reactive swale treating drainage water of an agricultural watershed in Brittany (France) with high nitrate concentrations in the receiving river, was monitored for one season (6 months). As flow in this field scale system is usually restricted to winter and spring months (December – May), it usually operates at low water temperatures of 5°C - 10°C. Tracer tests revealed shorter than designed retention times due to high inflows and preferential flow in the swale. Results show a correlation between residence time and nitrate reduction with low removal (<10%) at short residence times (<0.1 d), increasing to >25% at residence times >10h (0.4 d). Performance was compared to results of two technical scale reactive swales (length: 8 m) operated for 1.5 years at two different residence times (0.4 and 2.5 days), situated at a test site of the German Federal Environmental Agency (UBA) in Berlin (Germany). Similar nitrate reduction was observed for comparable temperature and residence time, showing that up-scaling is a suitable approach to transfer knowledge gathered from technical scale experiments to field conditions. For the design of new mitigation systems, one recommendation is to investigate carefully expected inflow volumes in advance to ensure a sufficient residence time for effective nitrate reduction at low temperatures.

Diffuse nitrate (NO3-) contamination from intense agriculture adversely impacts freshwater ecosystems, and can also pose a risk to human health if receiving surface waters are used for drinking water production. Implementation of near-natural mitigation zones such as reactive swales or wetlands have been proven to be promising measures to reduce nitrate loads in agricultural drainage waters. However, the behaviour of these systems at low temperatures and its dependence on system design is not well known until now. In this part of the Aquisafe project, the behaviour of a full scale (length: 45 m) infiltration ditch and two parallel wetlands (surface flow wetland and infiltration wetland) treating drainage water of two agricultural watersheds in Brittany (France) with high nitrate concentrations in the receiving river, were constructed and monitored for 3 flow seasons in 2011, 2012 and 2013 to evaluate field scale performance of these systems. As the flow in both sites is usually restricted to winter and spring months (December – May), systems usually operate at low water temperatures of 5°C - 10°C. Tracer tests revealed shorter than designed retention times (average values for whole flow season 2013: 1.1 h for infiltration ditch, 4.3 h for infiltration wetland and 8.4 h for surface wetland) due to high inflows and preferential flow. This likely is the main reason for observed low average retention of nitrate loads of 1.5-3% during the whole flow season. However, increase of relative nitrate retention to up to 80% during low flow conditions at the end of flow season in May with higher HRT and increasing temperatures show that investigated systems generally work. Results show a stronger correlation between residence time and nitrate reduction for all three systems compared to correlation with temperature. Retention times necessary in existing systems to achieve nitrate retention >30% were 1 day for infiltration ditch and 3 days for wetlands. Performance was compared to results of two technical scale reactive swales (length: 8 m) operated for 1.5 years at two different residence times (0.4 and 2.5 days), situated at a test site of the German Federal Environmental Agency (UBA) in Berlin (Germany). Similar nitrate reduction was observed for comparable temperature and HRT values (during low flow conditions at end of flow season 2013), showing that up-scaling is a suitable approach to transfer knowledge gathered from technical scale experiments to field conditions. For the design of new mitigation systems, expected inflow volumes have to be investigated carefully in advance to ensure a sufficient residence time for effective nitrate reduction at low temperatures.

This report compiles the results of three consecutive work packages that have been worked on during the Aquisafe II project. The approach developed is based on the previous Aquisafe I project where the Soil Water Assessment Tool (SWAT) was used as an analytical instrument to develop mitigation strategies for N loads and concentrations in the Ic catchment. During Aquisage I we concluded that SWAT should include a wetland function with which the effect of artificially, constructed wetlands on solute N fluxes can be evaluated. Chapter 1 compiles results of an extensive literature review that was made to identify potential wetland routines and processes that can be included in SWAT. The SWAT add-on to be developed should allow to individually test the effect on single wetlands (e.g. in a given hydrological response unit or subcatchment) as well as the effect of multiple wetlands on the landscape scale. We therefore implemented a stand alone version of the new wetland module which is described in Chapter 2. Here we show the general functionality and individual components of the wetland module. The chapter ends with a virtual application of the modules using SWAT outputs copied from the Ic results. Additionally, a Monte Carlo based sensitivity analyses of the wetland module input parameters showed that the denitrification rate seems to be the most constrained parameter for the simulation of N turnover in the new wetland module. A full implementation of the new wetland module is described in chapter 3. Here, the structural embedment of the wetland module in the SWAT architecture is described. To proof the functionality of the SWAT wetland module model runs were compared to the stand alone version to make sure that the module was correctly implemented. We conclude that the SWAT wetland extension is ready to be tested in real world catchments. Such a full test of the SWAT wetland model was planned towards the end of Aquisafe II. However, as data from the wetlands constructed within Aquisafe II were not available in due time, this last test of the SWAT module was possible.

The export of agricultural contaminants from agricultural landscapes of the US Midwest has contributed to the impairment of surface waters throughout the Mississippi River Basin and has been linked to various human health concerns. Natural treatment systems (wetlands, bioswales, bioreactors) can capture agricultural runoff and significantly reduce nutrient loading to downstream waters but there is a paucity of data on the effectiveness of these treatment systems to attenuate the suite of pollutants (nutrients and synthetic organics) typically found in agricultural runoff. This understanding is important given that the degradation of different pollutants involves metabolic pathways that often require different redox environments. As part of the Aquisafe-2 project, a bioretention swale comprising two treatment cells (a subsurface cell in series with a surface cell) was monitored, and its performance evaluated over a three-year period (2011 - 2013). Results showed that the bioswale was moderately efficient with regard to nitrate (NO3-; retention range: 16-58 %). N removal averaging 30 % was measured during a series of wetting events during which the bioswale operated at an estimated average hydraulic retention time (HRT) of 0.97 day. Spatial analysis of the data showed that almost all the NO3- removal occurred in the subsurface cell; however, N removal was also measured in the surface cell under low flow conditions (estimated HRT: 2.5 days). The highest rates of N removal (~ 58 %) were measured when the bioswale stayed wet for several days probably due to the development of a more optimum environment for denitrifying microbes. Nitrate removal capacity was limited by NO3- availability, short retention times during high flows, and the frequent fluctuation between oxic and anoxic conditions, but not by water temperature (8.3-16.6 oC) and dissolved organic carbon (DOC; 1.9 - 29.2 mg C L-1). The bioswale performance with regard to soluble reactive phosphorus (SRP) and atrazine was more variable, with net retention during some periods and net release at other times. The bioswale was a net source of P during most sampling periods with an average SRP release corresponding to 13 % of input, probably due to desorption of water soluble P from the topsoil applied during construction. This interpretation is supported by the progressive decline in P release observed between the first and third year of monitoring. The subsurface and the surface cells contributed almost equally to the fate of P in the bioswale. Likewise, the bioswale was at times a small/moderate sink (13-31 % retention) for atrazine, and a net source (-38 % to -15 %) during periods when the bioswale received overland runoff from the adjacent crop field which bypassed the subsurface cell. Results suggested that competition between atrazine and DOC for sorption sites is a possible mechanism affecting atrazine removal efficiency. Additional work is needed to compare the efficiency of the subsurface and surface cells with regard to atrazine, and elucidate the biogeochemical factors controlling its fate in the bioswale.

The present laboratory study tests the hypothesis that straw-bark mulch bioreactors are capable of concurrently retaining nitrate (NO3-) and the herbicides atrazine or bentazone at short hydraulic residence times (HRT). In a 1 year column experiment at HRT of ~4h three organic carbon sources, straw of common wheat (Triticum aestivum L.), bark mulch of pine tree (Pinus sp.) and a mixture of both materials, showed high reduction of continuously dosed NO3- (100mgL-1) with average denitrification rates of 23.4g-Nd-1m-3, 8.4g-Nd-1m-3 and 20.5g-Nd-1m-3, respectively. Under denitrifying conditions, fast and substantial retention of continuously dosed atrazine (20µgL-1) was observed with estimated dissipation times (DT50) between 0.12 and 0.49 days in the straw-bark mulch bioreactor. In parallel batch experiments, it could be confirmed that atrazine retention is based on adsorption to bark mulch and on degradation processes supplied by the organic materials as continual sources of carbon. In contrast, bentazone was not significantly reduced under the experimental conditions. While aging of materials was clearly observed in a reduction of denitrification by 60-70% during the experiment, systems still worked very well until the end of the experiment. The results indicate that the combined use of straw and bark mulch could increase the efficiency of mitigation systems, which are installed to improve the quality of drainage water before its release to surface waters. Further, the addition of these materials has the potential of parallel retention of NO3- and less mobile herbicides like atrazine, even during high flow events, as expected at the outlet of agricultural drainage systems. High removal is expected for mitigation system designed to remain saturated most of the time, whereas bioreactors that run periodically dry are not covered by this study. However, further experiments with the tested materials at technical or field scale under more realistic climate conditions need to be carried out.

In agricultural watersheds affected by diffuse pollution, limitation of fertilizer and pesticide application may not be sufficient to achieve good river water quality. After waterworks had to be closed in Brittany due to elevated nitrate concentrations in the river Ic (> 50 mg-NO3 L-1), the project Aquisafe has been initiated. The objective of Aquisafe is to reduce pollutant loads (nitrate and pesticides) from agricultural fields by implementation of near-natural mitigation zones at diffuse pollution hotspots at the head of watersheds. Simple and small solutions have to be designed in order to more efficiently reduce nitrate and pesticide concentrations in receiving rivers. In addition, a planning tool has to be developed to determine optimal locations to construct these systems. Finally, a tool to assess the effectiveness of these reactive zones on watershed water quality will be implemented. In order to reach the first objective, design features are tested on three scales: 1) laboratory scale, 2) technical scale and 3) field scale. 1) In the laboratory, column experiments were conducted with different organic substrates at short hydraulic residence times (HRT). The efficiency for parallel reduction of nitrate and two common herbicides in Europe, Bentazon and Isoproturon, was explored (Krause Camilo, 2012). 2) In technical scale, two parallel swales were filled with the most suitable material determined in (1) for a one year test. The influence of HRT and temperature was investigated. For nitrate, high reduction could be achieved at short HRT; results for herbicides still have to be confirmed. 3) One infiltration ditch and two simple wetlands were constructed in Brittany (France), taking into account experiences from other scales. These systems are now monitored to investigate the effects of upscaling. Site locations were chosen based on a validated and repeatable GIS-based overlay method that prioritises zones of potential contribution to nitrate pollution (Orlikowski et al, 2011). Additionally, a new wetland module is being developed for the Soil and Water Assessment Tool (SWAT). It allows to predict impacts of wetland constructions on nitrate concentrations in receiving rivers; the module is now implemented but still has to be calibrated with in situ monitoring results. The presentation will focus on results of the up-scaling approach, and will show how the tools of Aquisafe can be used for supporting the development of strategies at catchment scale.

The present study aimed at developing a universal method for the localization of critical source areas (CSAs) of diffuse nitrate (NO3-) pollution in rural catchments with low data availability. Based on existing methods, land use, soil, slope, riparian buffer strips and distance to surface waters were identified as the most relevant indicator parameters for diffuse agricultural NO3- parameters were averaged in a GIS-overlay to localize areas with low, medium and high risk of NO3- pollution. The five parameters were averaged in a GIS-overlay to localize areas with low, medium and high risk of NO3- pollution. A first application of the GIS approach to the Ic catchment in France, showed that identified CSAs were in good agreement with results from river monitoring and numerical modelling. Additionally, the GIS approach showed low sensitivity to single parameters, which makes it robust to varying data availability. As a result, the tested GIS-approach provides a promising, easy-to-use CSA identification concept, applicable for a wide range of rural catchments.

Rural watersheds often face diffuse pollution problems due to agricultural activities. In the Ic watershed in Brittany (France), nitrate concentrations in rivers frequently exceed the EUthreshold of 50 mg-NO3 L-1, despite various actions to reduce the impact from agriculture. As a result, other solutions are considered, such as mitigation systems that can prevent transfer of agricultural pollutants from cropland to the streams. Constructed wetlands have been shown to fit this aim, because they can reach significant N removal for water residence times above ~12 hours, can be implemented decentrally within rural watersheds, while meeting cost and policy requirements. However, constructed wetlands require space, which is particularly scarce and costly in intensively used agricultural watersheds. As a consequence, it was decided to test a more area-effective solution in three pilot systems. On the one hand land-use itself was optimized (i) at site 1 by placing two wetlands with same inflow and dimension on an area of minor agricultural value adjacent to a stream (one surface and one subsurface-flow, both 20 x 10 meters) and (ii) at site 2 by building an elongated infiltration wetland (45 x 2 meters) directly in an existing drainage ditch, thus preventing any use of agricultural surface. In both cases farmers agreed to the placement of the wetlands free of charge. On the other hand it was attempted to raise the areal removal efficiency, with a focus on denitrification, since nitrate is of most concern with inflow concentrations to the sites ranging between 30 and 66 mg-NO3 L-1. This increase in denitrification is attempted (a) by increasing the range of anoxic zones within the wetlands and (b) by adding carbon sources. For (a) one wetland at each site is filled with gravel with bottom outlets to enforce underground passage. Moreover saturation level within the infiltration wetlands and thus hydraulic retention time, can be controlled at drain outlets. For (b) organically rich soil is added to both wetlands at site 1 and carbon sources are mixed with the gravel at site 2. The three wetlands have been constructed in 2010 and are currently monitored for flow and water quality at inlets, as well as at surface and subsurface outlets. The monitoring will allow the calculation of substance mass balances for the entire rain season, expected from December 2010 to May 2011.