Assessment of multi-chemical pollution in aquatic ecosystems using toxic units: compound prioritization, mixture characterization and relationships with biological descriptors

Abstract

Chemical pollution is typically characterized by exposure to multiple rather than to single or a limited number of compounds. Parent compounds, transformation products and other non-targeted compounds yield mixtures whose composition can only be partially identified by monitoring, while a substantial proportion remains unknown. In this context, risk assessment based on the application of additive ecotoxicity models, such as concentration addition (CA), is rendered somewhat misleading. Here, we show that ecotoxicity risk information can be better understood upon consideration of the probabilistic distribution of risk among the different compounds. Toxic units of the compounds identified in a sample fit a lognormal probability distribution. The parameters characterizing this distribution (mean and standard deviation) provide information which can be tentatively interpreted as a measure of the toxic load and its apportionment among the constituents in the mixture (here interpreted as mixture complexity). Furthermore, they provide information for compound prioritization tailored to each site and enable prediction of some of the functional and structural biological variables associated with the receiving ecosystem. The proposed approach was tested in the Llobregat River basin (NE Spain) using exposure and toxicity data (algae and Daphnia) corresponding to 29 pharmaceuticals and 22 pesticides, and 5 structural and functional biological descriptors related to benthic macroinvertebrates (diversity, biomass) and biofilm metrics (diatom quality, chlorophyll-a content and photosynthetic capacity). Aggregated toxic units based on Daphnia and algae bioassays provided a good indication of the pollution pattern of the Llobregat River basin. Relative contribution of pesticides and pharmaceuticals to total toxic load was variable and highly site dependent, the latter group tending to increase its contribution in urban areas. Contaminated sites' toxic load was typically dominated by fewer compounds as compared to cleaner sites where more compounds contribute.

Keywords: Biological descriptors; Compound prioritization; Lognormal distribution; Mixture toxicity; Toxic units.