Two applications of hidden Markov models for animal movement data: the highs and lows of sampling resolution

Theoni Photopoulou (University of St Andrews)

Hidden Markov models are now commonly used for modelling time series of animal locations or other animal movement data. They are proving to be an extremely useful tool for understanding the evolution of various movement metrics through time. One of the strengths of HMMs are that they offer a core framework that is relatively easy to adapt and extend. I will present two applications that illustrate the diversity of problems that HMMs can be used to address within the field of animal movement. I will talk about a model for sparse location data from sharks, collected using passive acoustic telemetry, and a model for dive variables from a high spatial resolution dataset on grey seals and how HMMs can be used to make inferences about diving behaviour over longer time scales.