Deriving environmental quality standards in European surface waters: when are there too few data?

Recent technical guidance has been published by the European Commission that outlines methodologies for the derivation of Environmental Quality Standards (EQS) in European surface waters under the Water Framework Directive (WFD). The guidance allows the derivation of a long-term EQS from a small dataset. Specifically an EQS can be derived from just three acute data points, although the safety factors built into such an EQS are large (e.g. up to a factor of 1,000). Large safety factors make such EQS uncertain, and often difficult to achieve in practice. We examine dataset requirements for the derivation of EQS and specifically the minimum number of tests needed for setting EQS for long-term chemical exposures that result in reduced relative uncertainty, as assessed simply through the reduction in standard deviation of the means of the values derived. Using ecotoxicity datasets for four example chemicals, for which EQS have been derived in many jurisdictions, we show that variation in the EQS is greatest when using the minimum dataset allowable under the WFD guidance, but decreases rapidly when seven or more datapoints are available. Increasing the minimum number of ecotoxicity data in deriving an EQS results in a greater understanding of ecotoxicological effects. With this knowledge, the mitigating effects of water chemistry can be accounted for in deriving an EQS, even with relatively limited datasets. The new guidance suggests "simplistic" approaches to account for chemical availability, but does not detail how this might be undertaken. We provide examples of ways by which water chemistry effects can be included in deriving implementable EQS for metals with relatively few reliable and relevant data.