Stochastic process models for animal trajectories

Speaker Name: 
Mevin Hooten
Speaker Title: 
Speaker Organization: 
Department of Statistics, Colorado State University
Start Time: 
Monday, February 11, 2019 - 10:00am
End Time: 
Monday, February 11, 2019 - 11:15am
Simularium, E2-180
Bruno Sanso


Advances in animal telemetry data collection techniques have been a catalyst for the creation of statistical methodology for analyzing animal movement data.  Such data and methodology have provided a wealth of information about animal space use and the investigation of individual-based animal-environment relationships.  While the technology for data collection is improving dramatically over time, we also have massive archives of historical animal telemetry data of varying quality, both forms of which can lead to new ecological insights.  I provide a brief overview of the history of statistical models for animal movement at both the individual and population levels, and present a unified framework for modeling animal trajectories in continuous time.  This framework reconciles previous approaches to animal movement modeling based on stochastic differential equations and process convolutions commonly used in spatial statistics.  I show how to accommodate temporal heterogeneity in the trajectories using deformation methods and by accounting for interactions among individuals.   I also show how to extend these continuous-time stochastic process models to formally incorporate mechanisms based on animal physiology and decision making.  Throughout, I demonstrate several examples of statistical models for animal trajectories by analyzing data from migratory birds, ungulates, and large predators in a variety of regions spanning multiple continents.    


Mevin Hooten is a Professor at Colorado State University, a Fellow of the American Statistical Association, and also serves as the Assistant Unit Leader of the Colorado Cooperative Fish and Wildlife Research Unit.  He received his PhD in Statistics from the University of Missouri (2006) under the guidance of Prof. C.K. Wikle.  Mevin’s research focuses on the development of statistical methods and hierarchical models for analyzing spatial and spatio-temporal data to answer challenging ecological and environmental questions.