Evaluating Threats in the Face of Uncertainty for Endangered Atlantic Salmon

To identify threats that should be the focus of recovery actions, it is necessary to characterize how survival rates are influenced by environmental change at a population level.  However, the ways in which data are analyzed have the potential to change our understanding of these effects and thus could influence the recovery recommendations to address threats.  These ideas were explored using juvenile survival data from three endangered Atlantic salmon populations in response to drought conditions, with the overall goal of identifying whether drought conditions should be focused on during recovery planning.  Explicit comparison of Generalized Linear Models and Robust Regression demonstrated that traditional analyses would have led to ambiguity in the hydrological predictors identified as significant as well as low overall confidence in the predicted relationships.  However, accounting for overdispersion and outliers in the analyses indicated that survival declined with the frequency of drought conditions on the St. Mary’s River, and declined with faster hydrological rise rates on the Nashwaak River.  The LaHave River population exhibited a strong decline in survival over time that was not explained by changes in drought conditions, indicating that other environmental predictors should be explored and recovery actions will likely be focused on other threats.