Modeling Migration of Silver Eels to Forecast Critical Intervals of Risk

Populations of eels are in decline worldwide. One potentially significant source of mortality may occur at hydroelectric facilities during the downstream migration of adult silver eels. To help mitigate this mortality, a tool that can forecast migratory events could be of great use to managers that aim to reduce dam mortality while minimizing disruption of power generation at hydroelectric facilities. A number of studies have demonstrated relationships between downstream movements of silver eels and environmental events; however, there have been few attempts to summarize this available information into a predictive framework. Herein, we assess the utility of Bayesian hierarchical modeling methods to forecast the probability of migratory events of silver eels. We have compiled data on daily counts of migrating eels from 19 river systems which include one long term dataset on European eels (Anguilla anguilla), one long term dataset on American eels (Anguilla rostrata) from the Delaware and multiple short term datasets on American eels from rivers in Maine. We use local environmental information on moon phase, water temperature, and discharge to predict migration. The hierarchical framework we adopt combines this information across sites and species to provide robust estimates of the probability of eel migration.