Disentangling density‐dependent dynamics using full annual cycle models and Bayesian model weight updating

Abstract

Summary Density dependence regulates populations of many species across all taxonomic groups. Understanding density dependence is vital for predicting the effects of climate, habitat loss and/or management actions on wild populations. Migratory species likely experience seasonal changes in the relative influence of density dependence on population processes such as survival and recruitment throughout the annual cycle. These effects must be accounted for when characterizing migratory populations via population models. To evaluate effects of density on seasonal survival and recruitment of a migratory species, we used an existing full annual cycle model framework for American black ducks Anas rubripes, and tested different density effects (including no effects) on survival and recruitment. We then used a Bayesian model weight updating routine to determine which population model best fit observed breeding population survey data between 1990 and 2014. The models that best fit the survey data suggested that survival and recruitment were affected by density dependence and that density effects were stronger on adult survival during the breeding season than during the non‐breeding season. Analysis also suggests that regulation of survival and recruitment by density varied over time. Our results showed that different characterizations of density regulations changed every 8–12 years (three times in the 25‐year period) for our population. Synthesis and applications. Using a full annual cycle, modelling framework and model weighting routine will be helpful in evaluating density dependence for migratory species in both the short and long term. We used this method to disentangle the seasonal effects of density on the continental American black duck population which will allow managers to better evaluate the effects of habitat loss and potential habitat management actions throughout the annual cycle. The method here may allow researchers to hone in on the proper form and/or strength of density dependence for use in models for conservation recommendations.