Explicit Integrated Population Modeling: Escaping the Conventional Assumption of Independence

• Activity Number: 702
• Type: Contributed
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Audrey Beliveau* and Roger Pradel and Michael Schaub and Richard Lockhart and Carl J. Schwarz
• Companies: Simon Fraser University and Centre d'écologie fonctionnelle et évolutive and Swiss Ornithological Institute and Simon Fraser University and Simon Fraser University
• Keywords: capture-recapture ; integrated population modelling ; state-space modelling ; Bayesian modelling

Abstract

Integrated population modeling is a relatively new development in statistical ecology that permits the joint analysis of different sources of data. Typically, the joint likelihood is obtained using an independence assumption so that it can be conveniently expressed as a product of the likelihoods of the respective datasets (Schaub and Abadi, 2011). In this paper, I present a new Bayesian model that can be used when the independence assumption is not suitable. A key aspect of the model is that it makes use of latent variables that keep track of the states of the marked and unmarked individuals separately while allowing unmarked individuals to become marked when so. I present the results of a simulation study that compares, under various scenarios, the new model to the usual model that assumes independence. Finally, I use the methods to analyze capture-recapture data and count data collected for 15 years on a colony of Greater horseshoe bats (Rhinolophus ferrumequinum) in the Valais, Switzerland.