How to Characterize Chemical Exposure to Predict Ecologic Effects on Aquatic Communities?

R Schäfer, Nadine Gerner, Ben Kefford, Jes Rasmussen, M Beketov, Dick de Zwart, M Liess, Peter von der Ohe

Research output: Contribution to journalArticlepeer-review

70 Citations (Scopus)

Abstract

Reliable characterization of exposure is indispensable for ecological risk assessment of chemicals. To deal with mixtures, several approaches have been developed, but their relevance for predicting ecological effects on communities in the field has not been elucidated. In the present study, we compared nine metrics designed for estimating the total toxicity of mixtures regarding their relationship with an effect metric for stream macroinvertebrates. This was done using monitoring data of biota and organic chemicals, mainly pesticides, from five studies comprising 102 streams in several regions of Europe and South-East Australia. Mixtures of less than 10 pesticides per water sample were most common for concurrent exposure. Exposure metrics based on the 5% fraction of a species sensitivity distribution performed best, closely followed by metrics based on the most sensitive species and Daphnia magna as benchmark. Considering only the compound with the highest toxicity and ignoring mixture toxicity was sufficient to estimate toxicity in predominantly agricultural regions with pesticide exposure. The multisubstance Potentially Affected Fraction (msPAF) that combines concentration and response addition was advantageous in the study where further organic toxicants occurred. We give recommendations on exposure metric selection depending on data availability and the involved compounds.
Original languageEnglish
Pages (from-to)7996-8004
Number of pages9
JournalEnvironmental Science Technology (Washington)
Volume47
DOIs
Publication statusPublished - 2013

Fingerprint

Dive into the research topics of 'How to Characterize Chemical Exposure to Predict Ecologic Effects on Aquatic Communities?'. Together they form a unique fingerprint.

Cite this