Effects of data manipulation and statistical methods on species sensitivity distributions

Abstract

Species sensitivity distribution (SSD) methodology currently is used in environmental risk assessment to determine the predicted no-effect concentration (PNEC) of a substance in cases where a sufficient number of chronic ecotoxicological tests have been carried out on the substance, covering, for the aquatic environment with which we are concerned, three taxonomic groups: algae, invertebrates, and vertebrates. In particular, SSD methodology enables calculation of a hazardous concentration that is assumed to protect 95% of species (HC5). This approach is based on the hypothesis that the species for which results of ecotoxicological tests are known are representative, in terms of sensitivity, of the totality of the species in the environment, which raises a number of questions, both theoretical and practical. In this study, we compared various methods of constructing a species sensitivity-weighted distribution (SSWD). Each method is characterized by a different way of taking into account intraspecies variation and proportions of taxonomic groups (vertebrates, invertebrates, and algae), as well as by the statistical method of calculation of the HC5 and its confidence interval. Those methods are tested on 15 substances by using chronic no-observed-effect concentration data available in the literature. The choice of data (intraspecies variation and proportions between taxonomic groups) was found to have more effect on the value of the HC5 than the statistical method used to construct the distribution. The weight of each taxonomic group is the most important parameter for the result of the SSWD and letting literature references decide which proportions of data are used to construct it is not satisfactory.

Estimating hazardous concentrations by an informative Bayesian approach

“…Log-normal and related distributions (e.g., the log-logistic) have been tested in many articles that address SSD construction [8,10,12,14], and no theoretical justification for choosing other types of distributions have yet been identified. The SSD (after log transformation) can then be described as a normally distributed random variable: SSD$N m SSD ; s …”

“…Duboudin et al [8] described different weighting assumptions for deriving SSDs using frequentist approaches that can be used to correct disproportions in the number of data for the three main taxonomic groups. According to the alternatives proposed by Duboudin et al [8], two frequentist SSDs were then derived for each substance.…”

“…Second, because of the scarcity of available tests, the data used to generate SSDs often represent a mixture of end points (e.g., growth, reproduction, and mortality), each of them being more or less relevant for protecting the target [6]. Third, the trophic composition of taxa is heterogeneous within the data for a substance, which limits the relevance of comparative analyses [8]. The goal of this article is not to discuss these fundamental limitations of the SSD approach (even if they must be considered), especially those related to the quality of the ecotoxicological data, but to focus on the statistical method used to build SSDs.…”

“…Since the SSD concept was first developed, several frequentist statistical methods for estimating distributions and calculating HC5s have been suggested and compared [4,[8][9][10][11][12][13][14][15][16]. These methods differ in the choice of the underlying statistical distribution (empirical distribution, log-normal or log-logistic distribution, and others) and in the method used to estimate the confidence interval (bootstrap, asymptotic theory, and nonparametric statistics).…”

“…For example, Duboudin et al [8,20] calculated the chronic and acute SSDs of several substances for which sufficient data were available. If the standard deviation is relatively homogeneous among substances, prior information about the expected variance of the SSD can be defined for a new substance.…”