Defining ecoregions based on soil invertebrates for defining pesticide exposure scenarios

Abstract

Environmental Risk Assessment (ERA) is a process of identifying and evaluating the adverse effects on the environment caused by a chemical substance. Modeling environmental relevant concentrations in soil (ERCsoil) requires a different approach than the standard exposure scenario. Ecologically relevant scenarios must calculate exposure according to the habitats of soil organisms’ communities, their role in supporting soil functions and allow modeling ERC in different soil layers all around Europe. The aim of this study is to contribute in the definition of a EU-wide ecoregion-based map to improve the ecological relevance of soil exposure scenarios for collembola and isopods. These organisms were selected based on their importance ecological role in European soils, presence in a wide geographical scale, different morphological and ecological characteristics and data availability. Finland, Germany and Portugal were selected as model countries. The European Food Safety Authority (EFSA) databases used for this study compile information from published and some unpublished articles, species catalogs, and regional inventories. European Joint Research Center (JRC) maps provided the missing environmental variables for the spatial analysis. Soil organisms groups were classified by life form: euedaphic, hemiedaphic and epigeic for collembola; soil dwellers and litter dwellers for isopods; and then classified by dominance classes. Life form raw richness was used to create a generalized linear model (GLM) to describe the soil organisms’ distribution and class dominance. The software STELLA was employed to design a Stochastic Dynamic Methodology (StDM) model to predict distribution of the target soil groups. The results of the GLM and StDM model simulations were incorporated in ArcView 9.2 using the spatial analyst and geostatistical analysis extensions. The raster calculator and Ordinary Kriging were chosen to produce raw richness distribution maps for all life forms of collembola and isopods and to map class dominance. The models were not very successful at predicting low frequencies of dominance classes. Regardless, they were in line with ecological and biogeographic information for the considered groups. For collembola, Finland was dominated by epigeic species, while Portugal showed a dominance of epigeic and hemiedaphic species. In the case of Germany, the analysis methods reached different conclusions and patterns, the raster calculator analysis showed clear epigeic dominance while the ordinary kriging map displayed epigeic and hemiedaphic dominance. For isopods, both methodologies produced similar values for the two life forms in all countries, on average from 0 to 50% for soil dwellers and from 50 to 100% for litter dwellers. The only worst-case scenario predicted for pesticide assessment in all three countries was litter to 1 cm. Overall, the results obtained from the spatial and the geostatistical analysts were not helpful to define ecoregions for pesticide risk assessment given the available data and the selected GLM variables, as they do not provide enough discrimination between worstcase scenarios. Future studies should consider including only site data with complete environmental variables information and a specified geographical location. Abundance would also be a welcome improvement to the model.