International audience | Medicago sativa was cultivated at a former harbor facility near Bordeaux (France) to phytomanage a soil contaminated by trace elements (TE) and polycyclic aromatic hydrocarbons (PAH). In parallel, a biotest with Phaseolus vulgaris was carried out on potted soils from 18 sub-sites to assess their phytotoxicity. Total soil TE and PAH concentrations, TE concentrations in the soil pore water, the foliar ionome of M. sativa (at the end of the first growth season) and of Populus nigra growing in situ, the root and shoot biomass and the foliar ionome of P. vulgaris were determined. Despite high total soil TE, soluble TE concentrations were generally low, mainly due to alkaline soil pH (7.8–8.6). Shoot dry weight (DW) yield and foliar ionome of P. vulgaris did not reflect the soil contamination, but its root DW yield decreased at highest soil TE and/or PAH concentrations. Foliar ionomes of M. sativa and P. nigra growing in situ were generally similar to the ones at uncontaminated sites. M. sativa contributed to bioavailable TE stripping by shoot removal (in g ha−1 harvest−1): As 0.9, Cd 0.3, Cr 0.4, Cu 16.1, Ni 2.6, Pb 4, and Zn 134. After 1 year, 72 plant species were identified in the plant community across three subsets: (I) plant community developed on bare soil sowed with M. sativa; (II) plant community developed in unharvested plots dominated by grasses; and (III) plant community developed on unsowed bare soil. The shoot DW yield (in mg ha−1 harvest−1) varied from 1.1 (subset I) to 6.9 (subset II). For subset III, the specific richness was the lowest in plots with the highest phytotoxicity for P. vulgaris.

Medicago sativa was cultivated at a former harbor facility near Bordeaux (France) to phytomanage a soil contaminated by trace elements (TE) and polycyclic aromatic hydrocarbons (PAH). In parallel, a biotest with Phaseolus vulgaris was carried out on potted soils from 18 sub-sites to assess their phytotoxicity. Total soil TE and PAH concentrations, TE concentrations in the soil pore water, the foliar ionome of M. sativa (at the end of the first growth season) and of Populus nigra growing in situ, the root and shoot biomass and the foliar ionome of P. vulgaris were determined. Despite high total soil TE, soluble TE concentrations were generally low, mainly due to alkaline soil pH (7.8–8.6). Shoot dry weight (DW) yield and foliar ionome of P. vulgaris did not reflect the soil contamination, but its root DW yield decreased at highest soil TE and/or PAH concentrations. Foliar ionomes of M. sativa and P. nigra growing in situ were generally similar to the ones at uncontaminated sites. M. sativa contributed to bioavailable TE stripping by shoot removal (in g ha⁻¹ harvest⁻¹): As 0.9, Cd 0.3, Cr 0.4, Cu 16.1, Ni 2.6, Pb 4, and Zn 134. After 1 year, 72 plant species were identified in the plant community across three subsets: (I) plant community developed on bare soil sowed with M. sativa; (II) plant community developed in unharvested plots dominated by grasses; and (III) plant community developed on unsowed bare soil. The shoot DW yield (in mg ha⁻¹ harvest⁻¹) varied from 1.1 (subset I) to 6.9 (subset II). For subset III, the specific richness was the lowest in plots with the highest phytotoxicity for P. vulgaris.

International audience | Impacts of subsurface biogeochemical processes over time have always been a concern for the long-term performance of zero valent iron (Fe0) based Permeable Reactive Barriers (PRB’s). To evaluate the biogeochemical impacts, laboratory experiments were performed using flow-through glass columns for 210 days at controlled temperature (20 °C). Two different particle size of Fe0 were used in the columns, and to simulate indigenous microbial activity, extra carbon source was provided in two columns (biotic columns) and remaining two columns were kept abiotic. Heavy metals (Zn, As) were removed efficiently in all the columns, and no exhaustion of treatment capability or clogging were observed during our experimental duration. Newly formed Fe- mineral phases and precipitates were characterized using XRD, SEM-EDX and micro-XRF techniques in solid phase at the end of the experiment. In addition, 16S rRNA gene extraction was used for microbial community identification in biotic columns. During the incubation, microbial population shifted in favour of Desulforsporosinus species (sulfate-reducing bacteria) from initial dominance of Acidithiobacillus ferrooxidans in sediments. Dominant mineral phases detected in biotic columns were mackinawite (FeS) and sulfate green-rust while in abiotic columns magnetite/maghemite phases were more prevalent.

Impacts of subsurface biogeochemical processes over time have always been a concern for the long-term performance of zero valent iron (Fe⁰)-based permeable reactive barriers (PRBs). To evaluate the biogeochemical impacts, laboratory experiments were performed using flow-through glass columns for 210 days at controlled temperature (20 °C). Two different particle sizes of Fe⁰ were used in the columns, and to simulate indigenous microbial activity, extra carbon source was provided in the two columns (biotic columns) and the remaining two columns were kept abiotic using gamma radiations. Heavy metals (Zn, As) were removed efficiently in all the columns, and no exhaustion of treatment capability or clogging was observed during our experimental duration. Newly formed Fe mineral phases and precipitates were characterized using X-ray diffraction (XRD), scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM-EDX), and micro-XRF techniques in solid phase at the end of the experiment. In addition, 16S rRNA gene extraction was used for microbial community identification in biotic columns. During the incubation, microbial population shifted in favor of Desulfosporosinus species (sulfate-reducing bacteria) from initial dominance of Acidithiobacillus ferrooxidans in sediments. Dominant mineral phases detected in biotic columns were mackinawite (FeS) and sulfate green rust, while in abiotic columns, magnetite/maghemite phases were more prevalent.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]BELCA | International audience | Arsenic (As) pollution in water has important impacts for human and ecosystem health. In freshwaters, arsenate (AsV) can be taken up by microalgae due to its similarity with phosphate molecules, its toxicity being aggravated under phosphate depletion. An experiment combining ecological and ecotoxicological descriptors was conducted to investigate the effects of AsV (130 microg L-1 over 13 days) on the structure and function of fluvial biofilm under phosphate-limiting conditions. We further incorporated fish (Gambusia holbrooki) into our experimental system, expecting fish to provide more available phosphate for algae and, consequently, protecting algae against As toxicity. However, this protection role was not fully achieved. Arsenic inhibited algal growth and productivity but not bacteria. The diatom community was clearly affected showing a strong reduction in cell biovolume; selection for tolerant species, in particular Achnanthidium minutissimum; and a reduction in species richness. Our results have important implications for risk assessment, as the experimental As concentration used was lower than acute toxicity criteria established by the USEPA.

Arsenic (As) pollution in water has important impacts for human and ecosystem health. In freshwaters, arsenate (Asⱽ) can be taken up by microalgae due to its similarity with phosphate molecules, its toxicity being aggravated under phosphate depletion. An experiment combining ecological and ecotoxicological descriptors was conducted to investigate the effects of Asⱽ (130 μg L⁻¹ over 13 days) on the structure and function of fluvial biofilm under phosphate-limiting conditions. We further incorporated fish (Gambusia holbrooki) into our experimental system, expecting fish to provide more available phosphate for algae and, consequently, protecting algae against As toxicity. However, this protection role was not fully achieved. Arsenic inhibited algal growth and productivity but not bacteria. The diatom community was clearly affected showing a strong reduction in cell biovolume; selection for tolerant species, in particular Achnanthidium minutissimum; and a reduction in species richness. Our results have important implications for risk assessment, as the experimental As concentration used was lower than acute toxicity criteria established by the USEPA.

International audience | Chlordecone (CLD) is an organochlorine pesticide whose extended use led to the contamination of at least 20 % of agricultural soils from the French West Indies. Livestock reared on polluted areas are involuntary contaminated by CLD and their level of contamination may exceed the threshold values set by the European Union. Thus, characterizing the CLD behaviour in farm animals appear as a real issue in terms of food safety for local populations. The aim of this experiment was (i) to characterize the CLD disappearance in various tissues after exposure cessation and (ii) to evaluate the potential effect of body fatness on this process. Two groups of eight growing goats were submitted to either a basal diet or a high energy diet for 50 days before being intravenously contaminated with 1 mg CLD kg(-1) body weight. Two days after CLD contamination, half of the kids of each experimental group were slaughtered in order to determine pollutant levels in the serum, liver, adipose tissues, and empty carcass. The remaining animals were submitted to a 30-day decontamination period before slaughtering and measurements as described above. The implemented nutritional plan resulted in both groups of kids with significant differences in terms of body fatness. CLD was mainly concentrated in the liver of animals as described in the literature. It was found also in kids' empty carcass and adipose tissues; however its levels in the empty carcass (muscles and bones) were unexpected since they were higher than in fat. These results indicate that the lipophilic pollutant CLD is found mainly in liver but also in muscles and fat. Concerning the animals' depuration, a 30-d decontamination period was sufficient to observe a decrease of CLD levels by more than 75 % in both experimental groups and neither CLD concentrations nor CLD amounts were significantly affected by kids' body fatness.

Chlordecone (CLD) is an organochlorine pesticide whose extended use led to the contamination of at least 20 % of agricultural soils from the French West Indies. Livestock reared on polluted areas are involuntary contaminated by CLD and their level of contamination may exceed the threshold values set by the European Union. Thus, characterizing the CLD behaviour in farm animals appear as a real issue in terms of food safety for local populations. The aim of this experiment was (i) to characterize the CLD disappearance in various tissues after exposure cessation and (ii) to evaluate the potential effect of body fatness on this process. Two groups of eight growing goats were submitted to either a basal diet or a high energy diet for 50 days before being intravenously contaminated with 1 mg CLD kg⁻¹ body weight. Two days after CLD contamination, half of the kids of each experimental group were slaughtered in order to determine pollutant levels in the serum, liver, adipose tissues, and empty carcass. The remaining animals were submitted to a 30-day decontamination period before slaughtering and measurements as described above. The implemented nutritional plan resulted in both groups of kids with significant differences in terms of body fatness. CLD was mainly concentrated in the liver of animals as described in the literature. It was found also in kids’ empty carcass and adipose tissues; however its levels in the empty carcass (muscles and bones) were unexpected since they were higher than in fat. These results indicate that the lipophilic pollutant CLD is found mainly in liver but also in muscles and fat. Concerning the animals’ depuration, a 30-d decontamination period was sufficient to observe a decrease of CLD levels by more than 75 % in both experimental groups and neither CLD concentrations nor CLD amounts were significantly affected by kids’ body fatness.