Can Benchmarking Increase the Accuracy of Predicting Biodegradation Rates across Aquatic Ecosystems?

Describing and dealing with the large temporal and spatial variability in biodegradation rate constants is a requirement for robust persistence and exposure assessment. Chemical benchmarking uses a well-characterized reference chemical in a manner analogous to an internal standard in analytical chemistry; its measured biodegradation rate constant captures environmental-specific information that is used to predict the variability of the rate constants of other chemicals. We compiled 1656 biodegradation rate constants for 97 chemicals in European and Australian aquatic ecosystems, all measured with the same modified OECD 309 test protocol. Two benchmarking approaches were assessed for their ability to reduce the spatiotemporal variability in the data: (i) normalizing all chemicals to a single benchmark chemical (universal benchmarking); and (ii) grouping chemicals and normalizing the chemicals within each group to a benchmark chemical chosen from within that group (group-specific benchmarking). Universal benchmarking did not reduce the measured variability, while group-specific benchmarking did when the grouping of chemicals was optimized using the data. However, when chemical grouping was predicted based on the molecular fingerprint MACCS (Molecular ACCess System) or initial biotransformation rules, there was no reduction in variability for most chemicals. Group-specific benchmarking has promise as a tool to predict spatiotemporal variability in biodegradation rate constants when an appropriate chemical grouping is possible, but our current understanding of the key chemical features associated with biodegradability is insufficient to reliably group chemicals a priori.