Relationships between intertidal clam population and health status of the soft-shell clam Mya arenaria in the St. Lawrence Estuary and Saguenay Fjord (Québec, Canada).

The purpose of this study was to examine the impacts of anthropogenic activity on the health status of intertidal clam populations of the Saguenay Fjord and the St. Lawrence Estuary (Québec, Canada). Clams were collected during low tide at sites subject to direct contamination and at sites far from human activity. Clams were analyzed for tributyltin and dibutyltin total levels and toxic stress (glutathione S-transferase, gonadal lipid peroxidation and DNA strand breaks), immunocompetence (phagocytic activity, hemocyte count and viability), reproduction (gonado-somatic index, gamete maturation, and vitellogenin-like proteins), energy status (temperature-dependent mitochondrial electron transport, and gonad lipids), and individual status (age, condition factor, and growth index). These responses were compared against population characteristics such as live clam density, number of empty shells, and sex ratio. The results show that clam density decreased with distance from the estuary (high salinity level) to upstream of the fjord (low salinity). There was no clear relationship between the number of empty shells and distance or site quality. Clam density values corrected against distance were significantly correlated with hemocyte viability, phagocytic activity, mitochondrial electron transport (MET), DNA damage in gonad, and temperature-dependent mitochondrial electron transport activity. A canonical analysis of the various groups of biomarkers revealed that population metrics were more strongly related with immunocompetence, followed by energy status and temperature-dependent mitochondrial electron transport activity. However, toxic stress biomarkers were strongly associated with energy status and reproduction. This was further confirmed by non-linear modeling using adaptive artificial neural networks (genetic selection and back propagation learning paradigms), where the following parameters were able to predict population parameters with <20% error: gonad maturation and somatic index, MET (at 4 degrees C), gonad LPO, DNA damage, and phagocytic capacity. Intertidal clam populations were influenced by a distance gradient effect (salinity), where immunocompetence, in addition to energy status, was the strongest physiological parameter related to clam population metrics.