Animal dispersal is an important spatial process yet we lack a good mechanistic understanding and the ability to accurately quantify and predict dispersal patterns. We present a new approach to develop population redistribution kernels starting from individual movement data, test potential drivers of individual differences in dispersal and investigate long-term survival consequences. The approach is based on the use of a single time-dependent distance statistics, the net square displacement, which encapsulates key statistical properties of animal movements. Different movement behaviors lead to different displacement patterns and the functional forms can be predicted from theory. Using nonlinear mixed effects models the most adequate model can be identified within a model selection framework, population redistribution kernel as well as individual differences in dispersal can be quantified, and hypothesized drivers of the latter can be tested. We first tested the approach using simulated movement data, showing that population and individual movement parameters are correctly estimated. Second, we applied the approach on a dataset of 234 female elk radiotracked up to seven years after release in four contrasting areas in
Results/Conclusions
A multiphasic dispersal model (pre-dispersal, transience, settlement – modeled using a sigmoid function) best captured observed displacement patterns and was able to predict with near certainty displacement patterns over time and space, including different individuals and data on other populations obtained from the literature. Individual differences and the presence of conspecifics mostly determined differences in dispersal, together with the effects of release conditions, whereas effects of habitat differences were negligible. Finally, using Cox proportional hazard models it was shown that both the timing and distance of dispersal affected mortality risk 2 ~ 7 years after release. The approach provides a quantitative, predictive framework for dispersal ecology and reintroduction biology.
