Assessing, mapping and validating site-specific ecotoxicological risk for pesticide mixtures: A case study for small scale hot spots in aquatic and terrestrial environments Claudia Vaj a , Stefania Barmaz a , Peter Borgen Sørensen b , David Spurgeon c , Marco Vighi a,n a Department of Environmental Sciences, University of Milano-Bicocca, Piazza della Scienza, 1, 20126 Milano, Italy b National Environmental Research Institute, Department of Terrestrial Ecology, ˚ Arhus University, Vejlsøvej 25, 8600 Silkeborg, Denmark c Centre for Ecology and Hydrology, Maclean Building, Benson Lane, Crowmarsh Gifford, Wallingford, Oxon, OX10 8BB, UK article info Article history: Received 4 March 2011 Received in revised form 6 June 2011 Accepted 17 July 2011 Available online 25 August 2011 Keywords: Pesticides Risk assessment PNEC GIS Mixtures Stress factors abstract Mixture toxicity is a real world problem and as such requires risk assessment solutions that can be applied within different geographic regions, across different spatial scales and in situations where the quantity of data available for the assessment varies. Moreover, the need for site specific procedures for assessing ecotoxicological risk for non-target species in non-target ecosystems also has to be recognised. The work presented in the paper addresses the real world effects of pesticide mixtures on natural communities. Initially, the location of risk hotspots is theoretically estimated through exposure modelling and the use of available toxicity data to predict potential community effects. The concept of Concentration Addition (CA) is applied to describe responses resulting from exposure of multiple pesticides The developed and refined exposure models are georeferenced (GIS-based) and include environmental and physico-chemical parameters, and site specific information on pesticide usage and land use. As a test of the risk assessment framework, the procedures have been applied on a suitable study areas, notably the River Meolo basin (Northern Italy), a catchment characterised by intensive agriculture, as well as comparative area for some assessments. Within the studied areas, the risks for individual chemicals and complex mixtures have been assessed on aquatic and terrestrial aboveground and belowground communities. Results from ecological surveys have been used to validate these risk assessment model predictions. Value and limitation of the approaches are described and the possibilities for larger scale applications in risk assessment are also discussed. & 2011 Elsevier Inc. All rights reserved. 1. Introduction The procedures currently used for assessing ecotoxicological risk to fulfil the regulatory requirements for toxic substances are usually based on standard approaches, and require a comparison between exposure values, such as a Predicted Environmental Concentration (PEC) generally estimated via the application of predictive models, and effect parameters, such as a Predicted No Effect Concentration (PNEC), usually derived from laboratory toxicity data. A notable exception is the standard procedure used for pollinators, based on the Hazard Quotient (HQ) calculated as the ratio between pesticide application rate (g/ha) and a toxicological endpoint (mg/bee). The application rate replaces exposure in the procedure for assessing pollinator effects and thus HQ does not have any quantitative meaning of exposure evaluation. The standard procedures developed for regulatory risk assess- ment classifies both potentially hazardous chemicals and the relative risk. However, the results are not truly representative of actual site-specific conditions and thus difficult to relate to the risk posed to real ecosystems. Laboratory toxicity tests cannot account for the true complexity of community level responses to chemical exposure. Moreover, current procedures are mainly focused on individual chemicals and not on the effects of environmentally relevant mixtures as would be expected in many receiving ecosystems. The toxicological response to a mixture of chemical substances can be predicted using two different models: the Concentration Addition (CA) model, applicable to chemicals with the same mode of action, and the Independent Action (IA) model, applicable to chemicals with different modes of action (Greco et al., 1992). Selecting a suitable model of combined approach to joint effect prediction requires a knowledge of the toxicological mode of Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/ecoenv Ecotoxicology and Environmental Safety 0147-6513/$ - see front matter & 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ecoenv.2011.07.011 n Corresponding author. Fax: þ39 02 6448 2795. E-mail address: marco.vighi@unimib.it (M. Vighi). Ecotoxicology and Environmental Safety 74 (2011) 2156–2166