Using Bayesian Multi-State Mark-Recapture Models to Understand Life History Dynamics of Atlantic Salmon in Freshwater

Atlantic salmon are a dynamic species that exhibit extreme variation in life history phenotypes.  Males can mature in freshwater at sizes that are an order of magnitude smaller than males that migrate to sea.  In addition, transition to the migratory phase is size-dependent and the probability of migrating at a particular age is highly growth rate dependent.  As a consequence, growth rate mediated life history decisions can strongly influence population dynamics.  Capture mark-recapture (CMR) studies of individually tagged animals offer an excellent method to assess the causes and consequences of variable life history decisions.  During freshwater phase, juvenile Atlantic salmon are particularly amenable to CMR studies as their territorial nature and small stream habitat allow for efficient sampling of multiple individuals.  Despite these characteristics, considerable effort is required to capture and assign large numbers of individuals to a proper life history state as many individuals are not captured when they are expressing a given life history phenotype.  In our analysis, we explore the use of Bayesian multi-state modeling approach to understand life history decisions in juvenile Atlantic salmon.  This approach has many advantages as an unobservable state can be specified and individual covariates such as size at age or growth rate can be used to estimate the probability of transitioning to an observable state and surviving on that state.  This approach can be used to investigate fitness cost associated with different life history decisions when information about life history states is incomplete.