Evaluation and comparison of the relationship between NOEC and EC10 or EC20 values in chronic Daphnia toxicity testing

Hypothesis-based no-effect-concentration (NOEC) and regression-based x% effect concentration (ECx) values are common statistical approaches used to summarize ecotoxicological effects. Controversy over the NOEC model has prompted a movement toward discontinuation of the NOEC in favor of ECx, but the best x% effect surrogate for NOEC has not yet been determined. Historically, 10% and 20% effect concentrations (EC10 and EC20) have been treated as NOEC analogs. Given these measurements' importance to ecotoxicology, further understanding of the relationships between NOEC and EC10 or EC20 is crucial. In the present study, a metadataset of daphnid chronic toxicity tests was compiled to analyze the strength and significance of NOEC:EC10 and NOEC:EC20 relationships. The impact of endpoint (e.g., mortality, reproduction) and test condition parameters (e.g., pH, temperature) on NOEC:EC10 and NOEC:EC20 was evaluated. Mortality endpoints were most sensitive 51% of the time, with growth and reproductive endpoints constituting the remainder, underscoring the value of using multiple endpoints to evaluate toxic effects rather than relying on reproduction as the a priori most sensitive endpoint. When test condition parameters were less restricted (e.g., pH, hardness), the NOEC:EC20 association was more robust, suggesting that variability introduced by test implementation increased variability in ECx calculation. The analysis revealed that, overall, EC10 was a more suitable analog than EC20 for NOEC. Recommendations include refinement and reporting of the test parameters pH and hardness to minimize variability in ECx calculation. Environ Toxicol Chem 2015;34:2378-2384. (c) 2015 SETAC