Investigations of endocrine disruption using Gambusia species - A review

Abstract

A number of model fish species have been proposed for the examination of the effects of exposure to endocrine disrupting compounds (EDCs) and standard protocols for this examination have been developed. Where diversity and abundance of native fish fauna is low and endemism is high, the value of an exotic surrogate model is considerable, particularly where a surrogate is present in similar habitats to native species of interest. Gambusia are considered useful for this purpose due to their wide-spread distribution, tolerance to a wide range of environmental conditions, suitable lifecycle for study across various life stages, high level of sexual dimorphism, and well understood endocrinology. This review assembles and examines existing scientific knowledge on the use, value and limitations of using mosquitofish (Gambusia spp.) for investigating the presence of EDCs in freshwater aquatic systems. The different methodologies and endpoints utilized in studies of EDC impacts in gambusia range from gene expression, biochemical analysis, sperm production, histology, gross morphology, behaviour, through to population effects.Somatic measures and morphological reproductive endpoints are reviewed with a discussion of appropriate study design, experimental method and data analysis methods. Research demonstrating the utility of increasingly popular biochemical endpoints of exposure in studies of EDC effects in gambusia is discussed. The expression of the egg yolk protein, vitellogenin, in male gambusia appears due to exposure to estrogenic compounds, however, the biological implication of this induction remains unclear. Despite the wide variation in steroid hormone levels within a population, deviations in distinct ratios of 17ß-estradiol to androgens for male and female fish have been shown to relate to exposure to EDCs. The usefulness of histological examination of the gonads and other organs in identifying exposure to EDCs is considered and evaluated. Finally, the difficulty in extrapolation of measured morphological, biochemical and histological endpoints to population level effects is discussed.