International audience | In this study, the PAH-degrading bacteria of a constructed wetland collecting road runoff has been studied through DNA stable isotope probing. Microcosms were spiked with (13)C-phenanthrene at 34 or 337 ppm, and bacterial diversity was monitored over a 14-day period. At 337 ppm, PAH degraders became dominated after 5 days by Betaproteobacteria, including novel Acidovorax, Rhodoferax and Hydrogenophaga members, and unknown bacteria related to Rhodocyclaceae. The prevalence of Betaproteobacteria was further demonstrated by phylum-specific quantitative PCR, and was correlated with a burst of phenanthrene mineralization. Striking shifts in the population of degraders were observed after most of the phenanthrene had been removed. Soil exposed to 34 ppm phenanthrene showed a similar population of degraders, albeit only after 14 days. Results demonstrate that specific Betaproteobacteria are involved in the main response to soil PAH contamination, and illustrate the potential of SIP approaches to investigate PAH biodegradation in soil.

Organochlorine compounds (OC) were determined in Arctic bivalves (Mya truncata, Serripes groenlandicus, Hiatella arctica, Chlamys islandica) from Svalbard with regard to differences in geographic location, species and variations related to their size and age. Higher chlorinated polychlorinated biphenyls (PCB 101- PCB 194), chlordanes and α-hexachlorocyclohexane (α-HCH) were consistently detected in the bivalves and PCBs dominated the OC load in the organisms. OC concentrations were highest in Mya truncata and the lowest in Serripes groenlandicus. Species-specific OC levels were likely related to differences in the species' food source, as indicated by the δ13C results, rather than size and age. Higher OC concentrations were observed in bivalves from Kongsfjorden compared to the northern sampling locations Liefdefjorden and Sjuøyane. The spatial differences might be related to different water masses influencing Kongsfjorden (Atlantic) and the northern locations (Arctic), with differing phytoplankton bloom situations.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]BELCA | International audience | This study evaluated embryotoxicity and genotoxicity of two dissolved metals copper and cadmium (Cu and Cd) and two pesticides (metolachlor and irgarol) occurring in Arcachon Bay (SW France) in Pacific oyster (Crassostrea gigas) larvae and investigated the relationship between those two endpoints. Embryotoxicity was measured by calculating the percentage of abnormal D-shaped larvae and genotoxicity was evaluated with DNA strand breaks using the comet assay. After 24 h exposure, significant increases of the percentage of abnormal D-larvae and the DNA strand breaks were observed from 0.1µg/L for Cu, 10 µg/L for Cd and 0.01 µg/L for both irgarol and metolachlor in comparison with the controls. A strong positive relationship between embryotoxicity and genotoxicity was recorded for Cu, Cd and metolachlor. The current study suggests that copper, irgarol and metolachlor can induce larval abnormalities and DNA damage in a population of exposed oysters at environmentally relevant concentrations.

International audience | In this study, the PAH-degrading bacteria of a constructed wetland collecting road runoff has been studied through DNA stable isotope probing. Microcosms were spiked with (13)C-phenanthrene at 34 or 337 ppm, and bacterial diversity was monitored over a 14-day period. At 337 ppm, PAH degraders became dominated after 5 days by Betaproteobacteria, including novel Acidovorax, Rhodoferax and Hydrogenophaga members, and unknown bacteria related to Rhodocyclaceae. The prevalence of Betaproteobacteria was further demonstrated by phylum-specific quantitative PCR, and was correlated with a burst of phenanthrene mineralization. Striking shifts in the population of degraders were observed after most of the phenanthrene had been removed. Soil exposed to 34 ppm phenanthrene showed a similar population of degraders, albeit only after 14 days. Results demonstrate that specific Betaproteobacteria are involved in the main response to soil PAH contamination, and illustrate the potential of SIP approaches to investigate PAH biodegradation in soil.

Modeling water flow in a VFCW is a prerequisite to model wastewater treatment using process based filtering models. As for soils, when the vertical structure varies in different material types, it has a significant impact on water flow passing through it. The heterogeneous filtering material is composed of a mix of mineral porous material, and organic matter which makes its hydraulic characterization a difficult task. Indeed, the porosity may serve as preferential flow paths through which water can bypass most of the soil porous matrix in a largely unpredictable way. Consequently, non-equilibrium conditions in pressure heads are created between preferential flow paths and the soil matrix pore region. Preferential flows limit the applicability of standard models for water flow that are commonly based on Richards’ equation. Even if it is possible to simulate water content variations within a VFCW, we can not correctly model outflow with the standard van Genuchten-Mualem function. A number of various model approaches have been proposed to overcome this problem. These models mostly try to separately describe flow and transport in preferred flow paths and slow or stagnant pore regions. The objective of this study was to compare the various existing models simulating the preferential flows within the French VFCWs. Moreover, by assuming that several layers hydraulically different compose the VFCW, we tested at which layer(s) it is necessary to apply the non-equilibrium models. A tracer experiment was performed to evaluate the non-equilibrium degree. It was conducted on a 100 p.e. plant in operation since 2004. Monitoring consisted in measuring inlet and outlet flows, infiltration rates and water content at a time interval of 1 minute. We used the HYDRUS-1D software package containing various non-equilibrium flow modeling approaches. The physical non-equilibrium transport models were used to simulate outflow, the tracer breakthrough curve as well as water contents within a French-type VFCW. Physical non-equilibrium models include the dual-porosity model (mobile-immobile water model, with water content mass transfer or head mass transfer), and the dual-permeability model (matrix and fracture pore regions). We also applied a bimodal single-domain approach (Durner model) in order to see if it is actually necessary to use non-equilibrium models to effectively simulate VFCW outflow. Performance of the various non-equilibrium models (accuracy and limitation) was assessed by comparing the simulated and measured tracer fluxes using the mean square relative error (MSRE) of prediction. The comparison between measured and simulated tracer breakthrough curves indicates that the non-equilibrium (dual-porosity or dual-permeability model) approach seem to be the most appropriate for simulating preferential flow paths. In addition, simulations reveal that all layers participate in the preferential flow path process. These preferential flow paths are mainly due to the sludge layer that has been developed on the surface of VFCW since its start-up (swelling/shrinking during the feeding/rest periods) and to the network of roots and rhizomes present in it.

The Hydrus-2D software package allows to display the interaction of processes in constructed wetlands. Whereas several types of constructed wetlands were simulated successfully, the results for combined sewer overflow treatment wetlands are less satisfying so far. Main reasons can be found in the high variations of feeding characteristics as well as in extended dry periods of several weeks. In preparation of model improvements by considering organic filtrates as nutrient source in dry periods it seems to be essential to display hydraulics and solute transports correctly by tracer test studies. This was done in column experiments in the past, but not for full-scale constructed wetlands due to operational challenges. The given study shows tracer test results in pilot-scale operation in two different ways. Obtained measurement results were simulated with 1D and 2D approaches. Simulation results show the importance of considering inlet and outlet placement inside the filter beds correctly.

International audience | Organochlorinated compounds including PolyChloroBiphenyles, Dichloro-DiphenylTrichloroethan and metabolites are determinated in Stenella coeruleoalba (n = 37) stranded on the french Mediterranean coasts from 2007 till 2009. Studies are carried out on lung, muscle, kidney, liver, and blubber. The sought-after compounds are all detected to variable levels in each tissue and organ. In general, total PCBs are the most abondant, followed by total DDTs. The concentration (in ng g(-1) of lipid weight) in blubber of S. coeruleoalba, varied from 2,052 to 158,992 for PCBs and from 1,120 to 45,779 for DDTs. The ratios DDE/tDDTs are higher than 80% in almost all samples. The overall results of this work, compared to previous studies concerning the Mediterranean Sea, seems to confirm the tendency to a decrease of the contamination by organics compounds for the cetaceans in the Western Mediterranean Sea. (C) 2012 Elsevier Ltd. All rights reserved.

In the study, we establish centennial records of anthropogenic lead pollution at different locations in the North Atlantic (Iceland, USA, and Europe) by means of lead deposited in shells of the long-lived bivalve Arctica islandica. Due to local oceanographic and geological conditions we conclude that the lead concentrations in the Icelandic shell reflect natural influxes of lead into Icelandic waters. In comparison, the lead profile of the US shell is clearly driven by anthropogenic lead emissions transported from the continent to the ocean by westerly surface winds. Lead concentrations in the European North Sea shell, in contrast, are dominantly driven by local lead sources resulting in a much less conspicuous 1970s gasoline lead peak. In conclusion, the lead profiles of the three shells are driven by different influxes of lead, and yet, all support the applicability of Pb/Ca analyses of A. islandica shells to reconstruct location specific anthropogenic lead pollution.

The migration of uranium from polluted soil has been investigated in the field, and through modelling of thermodynamics and kinetics of uranium-water-rock interactions. Field monitoring following surface contamination by uranium deposits revealed up to 5 m deep uranium migration in soil and chalk substrate, as well as uranium concentrations in groundwater significantly higher than the geochemical background. Such observations can hardly be explained by a pure reactive transport dominated by reversible adsorption of uranium onto mineral phases. Therefore, a reactive transport model using the HYTEC code has been developed to better assess uranium migration through soil to the carbonate aquifer. Reactive transport modelling shows that adsorption of U (VI) at equilibrium on goethite at pH 7 is responsible for strong immobilization of uranium in the soil and carbonate matrix, matching uranium concentration profiles observed in boreholes. Simulations considering highly mobile ternary complex Ca2UO2(CO3)3(aq) in the aqueous phase cannot account alone for the rapid migration of uranium through the unsaturated zone. Without a mobile colloidal phase, the model clearly underestimates the concentration of aqueous U(VI) that reached groundwater underneath polluted soils.