International audience | More than 20 years after the Water Framework Directive was adopted, there are still major gaps in the sanitary status of small rivers and waterbodies at the head of basins. These small streams supply water to a large number of wetlands that support a rich biodiversity. Many of these waterbodies are fishponds whose production is destined for human consumption or for the restocking of other aquatic environments. However, these ecosystems are exposed to contaminants, including pesticides and their transformation products. This work aims to provide information on the distribution, diversity, and concentrations of agricultural contaminants in abiotic and biotic compartments from a fishpond located at the head of watersheds. A total of 20 pesticides and 20 transformation products were analyzed by HPLC-ESI-MS/MS in water and sediment sampled monthly throughout a fish production cycle, and in three fish species at the beginning and end of the cycle.The highest mean concentrations were found for metazachlor-OXA (519.48 +/- 56.52 ng.L-1) in water and benzamide (4.23 +/- 0.17 ng g(-1) dry wt.) in sediment. Up to 20 contaminants were detected per water sample and 26 per sediment sample. The transformation products of atrazine (banned in Europe since 2003 but still widely used in other parts of the world), flufenacet, imidacloprid (banned in France since 2018), metazachlor, and metolachlor were more concentrated than their parent compounds. Fewer contaminants were detected in fish and principally prosulfocarb accumulated in organisms during the cycle.Our work brings innovative data on the contamination of small waterbodies located at the head of a basin. The transformation products with the highest frequency of occurrence and concentrations should be prioritized for further environmental monitoring studies, and specific toxicity thresholds should be defined. Few contaminants were found in fish, but the results challenge the widely use of prosulfocarb.

More than 20 years after the Water Framework Directive was adopted, there are still major gaps in the sanitary status of small rivers and waterbodies at the head of basins. These small streams supply water to a large number of wetlands that support a rich biodiversity. Many of these waterbodies are fishponds whose production is destined for human consumption or for the restocking of other aquatic environments. However, these ecosystems are exposed to contaminants, including pesticides and their transformation products. This work aims to provide information on the distribution, diversity, and concentrations of agricultural contaminants in abiotic and biotic compartments from a fishpond located at the head of watersheds. A total of 20 pesticides and 20 transformation products were analyzed by HPLC-ESI-MS/MS in water and sediment sampled monthly throughout a fish production cycle, and in three fish species at the beginning and end of the cycle.The highest mean concentrations were found for metazachlor-OXA (519.48 ± 56.52 ng.L⁻¹) in water and benzamide (4.23 ± 0.17 ng g⁻¹ dry wt.) in sediment. Up to 20 contaminants were detected per water sample and 26 per sediment sample. The transformation products of atrazine (banned in Europe since 2003 but still widely used in other parts of the world), flufenacet, imidacloprid (banned in France since 2018), metazachlor, and metolachlor were more concentrated than their parent compounds. Fewer contaminants were detected in fish and principally prosulfocarb accumulated in organisms during the cycle.Our work brings innovative data on the contamination of small waterbodies located at the head of a basin. The transformation products with the highest frequency of occurrence and concentrations should be prioritized for further environmental monitoring studies, and specific toxicity thresholds should be defined. Few contaminants were found in fish, but the results challenge the widely use of prosulfocarb.

International audience | Glyphosate (N-phosphonomethylglycine; GLP) and its main metabolite AMPA (aminomethylphosphonic acid), are frequently detected in relatively high concentrations in European agricultural topsoils. Glyphosate has a high sorption affinity, yet it can be detected occasionally in groundwater. We hypothesized that shrinkage cracks occurring after dry periods could facilitate GLP transport to greater depths where subsoil conditions slow further microbial degradation. To test this hypothesis, we simulated a heavy rainfall event (HRE) on a clay-rich arable soil. We applied 2.1 kg ha−1 of 100% 13C3, 15N-labeled GLP one day before the simulated rainfall event. Microbial degradation of translocated GLP over a 21-day period was assessed by quantifying 13C incorporation into phospholipid fatty acids. Microbial degradation potential and activity were determined by quantifying the abundance and expression of functional genes involved in the two known degradation pathways of GLP; to AMPA (goxA) or sarcosine (sarc). We confirmed that goxA transcripts were elevated in the range of 4.23 x 105 copy numbers g−1 soil only one day after application. The increase in AMPA associated with a rise in goxA transcripts and goxA-harboring microorganisms indicated that the degradation pathway to AMPA dominated. Based on 13C-enrichment 3 h after the HRE, fungi appeared to initiate glyphosate degradation. At later time points, Gram+-bacteria proved to be the main degraders due to their higher 13C-incorporation. Once GLP reached the subsoil, degradation continued but more slowly. By comparing GLP distribution and its microbial degradation in macropores and in the bulk soil, we demonstrated different time- and depth-dependent GLP degradation dynamics in macropores. This indicates the need for field studies in which soil properties relevant to GLP degradation are related to limiting environmental conditions, providing a realistic assessment of GLP fate in soils.

Glyphosate (N-phosphonomethylglycine; GLP) and its main metabolite AMPA (aminomethylphosphonic acid), are frequently detected in relatively high concentrations in European agricultural topsoils. Glyphosate has a high sorption affinity, yet it can be detected occasionally in groundwater. We hypothesized that shrinkage cracks occurring after dry periods could facilitate GLP transport to greater depths where subsoil conditions slow further microbial degradation. To test this hypothesis, we simulated a heavy rainfall event (HRE) on a clay-rich arable soil. We applied 2.1 kg ha⁻¹ of 100% ¹³C₃, ¹⁵N-labeled GLP one day before the simulated rainfall event. Microbial degradation of translocated GLP over a 21-day period was assessed by quantifying ¹³C incorporation into phospholipid fatty acids. Microbial degradation potential and activity were determined by quantifying the abundance and expression of functional genes involved in the two known degradation pathways of GLP; to AMPA (goxA) or sarcosine (sarc). We confirmed that goxA transcripts were elevated in the range of 4.23 x 10⁵ copy numbers g⁻¹ soil only one day after application. The increase in AMPA associated with a rise in goxA transcripts and goxA-harboring microorganisms indicated that the degradation pathway to AMPA dominated. Based on ¹³C-enrichment 3 h after the HRE, fungi appeared to initiate glyphosate degradation. At later time points, Gram⁺-bacteria proved to be the main degraders due to their higher ¹³C-incorporation. Once GLP reached the subsoil, degradation continued but more slowly. By comparing GLP distribution and its microbial degradation in macropores and in the bulk soil, we demonstrated different time- and depth-dependent GLP degradation dynamics in macropores. This indicates the need for field studies in which soil properties relevant to GLP degradation are related to limiting environmental conditions, providing a realistic assessment of GLP fate in soils.

International audience | Accumulation of copper (Cu) in soils due to the application of fungicides may be toxic for organisms and hence affect winegrowing sustainability. Soil parameters such as pH and dissolved organic matter (DOM) are known to affect the availability of Cu. In this study, we investigated the contribution of chromophoric and fluorescent DOM properties to the prediction of Cu availability in 18 organic vineyard soils in the Bordeaux winegrowing area (France). The DOM parameters, assessed through absorbance and fluorescence analyses, and proxies for Cu availability (total soluble Cu and free ionic Cu2+) were measured in 0.01 M KCl extracts. Total soluble Cu (CuKCl) varied 23-fold while free ionic Cu2+ varied by a factor of 4600 among the soils. DOC concentrations were similar among the soils, but the samples differed in the quality of DOM as assessed by optical spectroscopy. Multilinear regression models with and without DOM quality parameters were investigated to predict Cu availability. The best model for CuKCl successfully explained 83% of variance and included pH, CuT, and two DOM fluorescence quality indices, the FI fluorescence index, which distinguishes between microbial and higher plant origins, and the HIX humification index. For the prediction of Cu2+, pH alone explained 88% of variance and adding DOM parameters did not improve modelling. The two Cu availability proxies were related to pH. This study confirms the prominent role of pH in Cu availability and underlines the importance of DOM quality to better predict Cu solubility

Accumulation of copper (Cu) in soils due to the application of fungicides may be toxic for organisms and hence affect winegrowing sustainability. Soil parameters such as pH and dissolved organic matter (DOM) are known to affect the availability of Cu. In this study, we investigated the contribution of chromophoric and fluorescent DOM properties to the prediction of Cu availability in 18 organic vineyard soils in the Bordeaux winegrowing area (France). The DOM parameters, assessed through absorbance and fluorescence analyses, and proxies for Cu availability (total soluble Cu and free ionic Cu²⁺) were measured in 0.01 M KCl extracts. Total soluble Cu (CuKCₗ) varied 23-fold while free ionic Cu²⁺ varied by a factor of 4600 among the soils. DOC concentrations were similar among the soils, but the samples differed in the quality of DOM as assessed by optical spectroscopy. Multilinear regression models with and without DOM quality parameters were investigated to predict Cu availability. The best model for CuKCₗ successfully explained 83% of variance and included pH, CuT, and two DOM fluorescence quality indices, the FI fluorescence index, which distinguishes between microbial and higher plant origins, and the HIX humification index. For the prediction of Cu²⁺, pH alone explained 88% of variance and adding DOM parameters did not improve modelling. The two Cu availability proxies were related to pH. This study confirms the prominent role of pH in Cu availability and underlines the importance of DOM quality to better predict Cu solubility.