Nitrous oxide is a potent greenhouse gas, with a global warming potential 298 times greater than that of CO2. In agricultural soils, N2O emissions are influenced by a large number of environmental characteristics and crop management techniques that are not systematically reported in experiments. Random Forest (RF) is a machine learning method that can handle missing data and ranks input variables on the basis of their importance. We aimed to predict N2O emission on the basis of local information, to rank environmental and crop management variables according to their influence on N2O emission, and to compare the performances of RI: with several regression models. RF outperformed the regression models for predictive purposes, and this approach led to the identification of three important input variables: N fertilization, type of crop, and experiment duration. This method could be used in the future for prediction of N2O emissions from local information. (c) 2013 Elsevier Ltd. All rights reserved.

Environ Pollut 859TF Times Cited:2 Cited References Count:50 | International audience | An investigation was undertaken to identify the most significant soil parameters that can be used to predict Cd, Pb, and Zn bioaccessibility in smelter-contaminated agricultural soils. A robust model was established from an extended database of soils by using : a training set of 280 samples to select the main soil parameters, to define the best population to be taken into account for the model elaboration, and to construct multivariate regression models, and a test set of 110 samples to validate the ability of the regression models. Total carbonate, organic matter, sand, P2O5, free Fe-Mn oxide, and pseudo total Aland trace element (TE) contents appeared as the main variables governing TE bioaccessibility. The statistical modeling approach was reasonably successful, indicating that the main soil factors influencing the bioaccessibility of TEs were taken into account and the predictions could be applicable for further risk evaluation in the studied area.

International audience | Background, aim and scope Agricultural landscapes comprise cultivated fields and semi-natural areas. Biological components of these compartments such as weeds, insect pests and pathogenic fungi can disperse sometimes over very large distances, colonise new habitats via insect flight, spores, pollen or seeds and are responsible for losses in crop yield (e.g. weeds, pathogens) and biodiversity (e.g. invasive weeds). The spatiotemporal dynamics of these biological components interact with crop locations, successions and management as well as the location and management of semi-natural areas such as roadverges. The objective of this investigation was to establish a modelling and simulation methodology for describing, analysing and predicting spatiotemporal dynamics and genetics of biological components of agricultural landscapes. The ultimate aim of the models was to evaluate and propose innovative cropping systems adapted to particular agricultural concerns. The method was applied to oilseed rape (OSR) volunteers playing a key role for the coexistence of genetically modified (GM) and non-GM oilseed rape crops, where the adventitious presence of GM seeds in non-GM harvests (AGMP) could result in financial losses for farmers and cooperatives. Material and methods A multi-year, spatially explicit model was built, using field patterns, climate, cropping systems and OSR varieties as input variables, focusing on processes and cultivation techniques crucial for plant densities and pollen flow. The sensitivity of the model to input variables was analysed to identify the major cropping factors. These should be modified first when searching for solutions limiting gene flow. The sensitivity to model processes and species life-traits were analysed to facilitate the future adaptation of the model to other species. The model was evaluated by comparing its simulations to independent field observations to determine its domain of validity and prediction error. Results The cropping system study determined contrasted farm types, simulated the current situation and tested a large range of modifications compatible with each farm to identify solutions for reducing the AGMP. The landscape study simulated gene flow in a large number of actual and virtual field patterns, four combinations of regional OSR and GM proportions and three contrasted cropping systems. The analysis of the AGMP rate at the landscape level determined a maximum acceptable GM OSR area for the different cropping systems, depending on the regional OSR volunteer infestation. The analysis at the field level determined minimum distances between GM and non-GM crops, again for different cropping systems and volunteer infestations. Discussion The main challenge in building spatially explicit models of the effects of cropping systems and landscape patterns on species dynamics and gene flow is to determine the spatial extent, the time scale, the major processes and the degree of mechanistic description to include in the model, depending on the species characteristics and the model objective. Conclusions These models can be used to study the effects of cropping systems and landscape patterns over a large range of situations. The interactions between the two aspects make it impossible to extrapolate conclusions from individual studies to other cases. The advantage of the present method was to produce conclusions for several contrasted farm types and to establish recommendations valid for a large range of situations by testing numerous landscapes with contrasted cropping systems. Depending on the level of investigation (region or field), these recommendations concern different decision-makers, either farmers and technical advisors or cooperatives and public decision-makers. Recommendations and perspectives The present simulation study showed that gene flow between coexisting GM and non-GM varieties is inevitable. The management of OSR volunteers is crucial for containing gene flow, and the cropping system study identified solutions for reducing these volunteers and ferals in and outside fields. Only if these are controlled can additional measures such as isolation distances between GM and non-GM crops or limiting the proportion of the region grown with GM OSR be efficient. In addition, particular OSR varieties contribute to limit gene flow. The technical, organisational and financial feasibility of the proposed measures remains to be evaluated by a multi-disciplinary team.

EASPEGESTADSUPDATINRACT1EJ2 INRA | Herbicide-tolerant (HT) crops such as those tolerant to glyphosate simplify weed management and make it more efficient, at least at short-term. Overreliance on the same herbicide though leads to the spread of resistant weeds. Here, the objective was to evaluate, with simulations, the impact on the advent of glyphosate resistance in weeds of modifications in agricultural practises resulting from introducing HT maize into cropping systems. First, we included a single-gene herbicide resistance submodel in the existing multispecific FLORSYS model. Then, we (1) simulated current conventional and probable HT cropping systems in two European regions, Aquitaine and Catalonia, (2) compared these systems in terms of glyphosate resistance, (3) identified pertinent cultural practises influencing glyphosate resistance, and (4) investigated correlations between cultural practises and species traits, using RLQ analyses. The simulation study showed that, during the analysed 28years, (1) glyphosate spraying only results in glyphosate resistance in weeds when combined with other cultural factors favouring weed infestation, particularly no till; (2) pre-sowing glyphosate applications select more for herbicide resistance than post-sowing applications on HT crops; and (3) glyphosate spraying selects more for species traits avoiding exposure to the herbicide (e.g. delayed early growth, small leaf area) or compensating for fitness costs (e.g. high harvest index) than for actual resistance to glyphosate, (4) actual resistance is most frequent in species that do not avoid glyphosate, either via plant size or timing, and/or in less competitive species, (5) in case of efficient weed control measures, actual resistance proliferates best in outcrossing species. An advice table was built, with the quantitative, synthetic ranking of the crop management effects in terms of glyphosate-resistance management, identifying the optimal choices for each management technique.

Measurement of solute-transport parameters through soils for a wide range of solute- and soil-types is time-consuming, laborious, expensive and practically impossible. So, indirect methods for estimating the transport parameters by pedo-transfer functions are now advancing. This study developed and evaluated an Artificial Neural Network (ANN) model for estimating the transport velocity (V), dispersion coefficient (D) and retardation factor (R) of NaAsO₂, Pb(NO₃)₂, Cd(NO₃)₂, C₉H₉N₃O₂ and CaCl₂ from the basic soil properties. Breakthrough data of the solutes were measured in 14 agricultural soils of Bangladesh by time-domain reflectometry (TDR) in repacked soil columns under unsaturated steady-state water-flow conditions. The transport parameters of the chemicals were determined by analyzing the solute breakthrough data. Bulk density (γ), organic carbon (OC), clay (C) content, pH, median grain diameter (D₅₀) and uniformity coefficient (Cᵤ) of the soils were determined. An ANN model for V, D and R was developed by using data of eight soils, validated/tested with the data of five soils and verified with the data of one soil. Clay content and bulk density of the soils were the most sensitive input variables to the ANN model followed by other soil properties (OC, C, pH, D₅₀ and Cᵤ). The model reliably predicted V, D and R with relative root-mean-square error (RRMSE) of 0.028–0.363, mean error (ME) of – 0.00004 to 0.0005, bias error (BOE%) of 0–0.003 and modeling efficiency (EF) of >0.99. Thus, the ANN model can significantly enhance prediction of pollution transport through soils in terms of cost and effort.

It has been hypothesised that, if ingested, plastic debris could act as vector for the transfer of chemical contaminants from seawater to organisms, yet modelling suggest that, in the natural environment, chemical transfer would be negligible compared to other routes of uptake. However, to date, the models have not incorporated consideration of the role of gut surfactants, or the influence of pH or temperature on desorption, whilst experimental work has shown that these factors can enhance desorption of sorbed contaminants several fold. Here, we modelled the transfer of sorbed organic contaminants dichlorodiphenyltrichloroethane (DDT), phenanthrene (Phe) and bis-2-ethylhexyl phthalate (DEHP) from microscopic particles of polyvinylchloride (PVC) and polyethylene (PE) to a benthic invertebrate, a fish and a seabird using a one-compartment model OMEGA (Optimal Modelling for EcotoxicoloGical Applications) with different conditions of pH, temperature and gut surfactants. Environmental concentrations of contaminants at the bottom and the top of published ranges were considered, in combination with ingestion of either 1 or 5% by weight of plastic. For all organisms, the combined intake from food and water was the main route of exposure for Phe, DEHP and DDT with a negligible input from plastic. For the benthic invertebrate, predictions including the presence of contaminated plastic resulted in very small increases in the internal concentrations of DDT and DEHP, while the net change in the transfer of Phe was negligible. While there may be scenarios in which the presence of plastic makes a more important contribution, our modelling study suggests that ingestion of microplastic does not provide a quantitatively important additional pathway for the transfer of adsorbed chemicals from seawater to biota via the gut.

Increasing heavy metal pollution in wetland ecosystems around the world pose significant health risks to waterbirds, especially the endangered species. We analyzed heavy metal pollution in bird foraging area of Caohai wetland in Guizhou Plateau (China), and used an integrated exposure risk model for assessing heavy metal (Cd, Pb, Cr, Ni, Zn, Sb) exposure risk in birds from the soil, water, plants and benthic invertebrates. There is considerable variation in the extent of heavy metal contamination across the different sampling sites, and Cd and Sb are the main contaminants. The mussel Anodonta showed greater heavy metal accumulation (except for Zn) compared to the snail species C. cathayensis. The different plant species also varied in terms of amount for accumulated heavy metals. The phytophagous together with omnivorous birds were exposed to Cd, Pb, Zn and Ni through plants rather than the soil, whereas the maximum Zn exposure in the omnivorous and carnivorous birds was through consumption of benthic invertebrates. Furthermore, the phytophagous black-necked cranes (Grus nigricollis) were less risk to heavy metal exposure compared to the omnivorous bar-headed goose (Anser indicus) and carnivorous ruddy shelducks (Tadorna ferruginea). The exposure risk of Cr (8.1) was highest, followed by Pb (5.1), Zn (3.8), Sb (1.0), Cd (0.33) and Ni (0.28). The heavy metal assessment heavy metal exposure risk for migratory birds should take into account the exposure from food and soil. Our findings provide new insights into developing measures to minimize heavy metal contamination in migratory birds.

Various sources supply PAHs that accumulate in soils. The methodology we developed provided an evaluation of the contribution of local sources (road traffic, local industries) versus remote sources (long range atmospheric transport, fallout and gaseous exchanges) to PAH stocks in two contrasting subcatchments (46–614 km²) of the Seine River basin (France). Soil samples (n = 336) were analysed to investigate the spatial pattern of soil contamination across the catchments and an original combination with radionuclide measurements provided new insights into the evolution of the contamination with depth. Relationships between PAH concentrations and the distance to the potential sources were modelled. Despite both subcatchments are mainly rural, roadside areas appeared to concentrate 20% of the contamination inside the catchment while a local industry was found to be responsible for up to 30% of the stocks. Those results have important implications for understanding and controlling PAH contamination in rural areas of early-industrialized regions.

Increases in dissolved organic carbon (DOC) fluxes may relate to changes in sulphur and nitrogen pollution. We integrated existing models of vegetation growth and soil organic matter turnover, acid–base dynamics, and organic matter mobility, to form the ‘MADOC’ model. After calibrating parameters governing interactions between pH and DOC dissolution using control treatments on two field experiments, MADOC reproduced responses of pH and DOC to additions of acidifying and alkalising solutions. Long-term trends in a range of acid waters were also reproduced. The model suggests that the sustained nature of observed DOC increases can best be explained by a continuously replenishing potentially-dissolved carbon pool, rather than dissolution of a large accumulated store. The simulations informed the development of hypotheses that: DOC increase is related to plant productivity increase as well as to pH change; DOC increases due to nitrogen pollution will become evident, and be sustained, after soil pH has stabilised.

An investigation was undertaken to evaluate the empirical model developed by Pelfrêne et al. (2012), predicting the human bioaccessibility of Cd and Pb in smelter-contaminated agricultural topsoils, by including other soil uses: 50 urban and 65 woody habitat topsoils collected in the same area. The results showed that land use significantly affected the pseudototal metal concentrations and their oral bioaccessibility. However, whatever the soil's physicochemical parameters and degree of contamination, the ‘agricultural’ model can be used to simulate metal gastric bioaccessibility in urban and woody habitat soils. To simulate gastrointestinal bioaccessibility, this model can be used directly if the pseudototal metal concentrations are on the same order of magnitude as those usually recorded in the agricultural soils studied or after the use of a correction factor if these concentrations are greater. These results showed that the oral bioaccessibility predictions could be applicable for further environmental risk evaluation.