Fish Population Ecology and Ecological Risk Assessment

Abstract

Density-dependent processes are crucial in the regulation of fish populations and strongly influence their resilience to exploitation and exposure to toxic chemicals. Multiple density-dependent processes occur at different stages in the life-cycle of fish, and a general pattern of such processes in the ontogeny of fish has been suggested but not clearly demonstrated in natural populations. This thesis aimed to provide a detailed experimental assessment of density-dependent processes through the entire life-cycle, using laboratory and semi-natural populations of zebrafish, Danio rerio, and to explore the implications of these processes for the ecological risk assessment of endocrine disrupting chemicals using an individual-based population model. Results clearly demonstrate the importance of density-dependent mortality in the early juvenile life-stage and density-dependent growth in the late juvenile and adult life-stages consistent with evidence from wild populations of much larger wild species, suggesting the existence of general ontogenetic patterns of density dependence that are invariant to maximum size. Patterns of density dependence found in populations of zebrafish under semi-natural conditions in Bangladesh were similar to those observed in the laboratory, except that the absolute strength of density dependence was higher and consequently, carrying capacity lower, by about two orders of magnitude in the semi-natural populations. A conclusion from these studies is that these patterns of density dependence are applicable generally across the teleost taxa due to developmental similarities. The population model incorporating these patterns of density dependence showed that density dependence compensated for reasonably high levels of disruption for many individual-level endpoints currently used in risk assessment, including fecundity and sex ratio. This indicates that current risk assessment practices are highly conservative and the inclusion of population models such as developed here for zebrafish, could enhance the scientific basis and ecological realism of laboratory derived data used in risk analysis.