The Challenges of Fitting Movement Models to Data

MORALES, JUAN M.

Puerto Iguazú

Workshop; SATELLITE TELEMETRY WORKSHOP (Birds in Argentina and other areas of South America); 2011

Technological advances in different kinds of remote measurements (biotelemetry) are poised to revolutionize the way we do ecology. Developments in animal tracking technology have spawned a wealth of detailed movement paths for individuals of many species, from bees to whales and including crustaceans, molluscs, fish, birds, reptiles and mammals. Remote sensing technology is used increasingly to generate maps of environmental attributes with ever finer spatial and temporal resolution. In many ways, data collection technology has developed faster than the analytical methods available in the ecologists’ toolbox. The quantity and quality of movement data is now far beyond what can be handled by traditional movement analysis. Only a handful of quite recent papers attempt to explicitly model movement behaviour and the way it is affected by landscape heterogeneity and other factors. These approaches are possible thanks to the application of hierarchical Bayesian (HB) analysis coupled with Markov chain Monte Carlo Methods (MCMC). Our basic modelling premises are that the complexities of animal movement can be dissected into a few general movement strategies, and that animals modulate the switch among these strategies as they are affected by changes in the internal and external environment. This conceptual framework fits very nicely with the hierarchical structure of Bayesian mixture models. I show examples of how this can be achieved and highlight the methodological challenges behind it.