Capture-recapture studies are widely used by population biologists and ecologists to study the dynamics of various wildlife populations. Fairly recent innovations in statistical methodology allow us to study movement as well as important factors such as survival, recapture and recovery rates. Our paper is motivated by a series of data on a population of mouflons (Ovis gmelini) in the Caroux-Espinouse massif and focuses upon discriminating between competing biological hypotheses corresponding to the dependence of any or all of the population parameters upon either sex, location or age. We consider the Arnason-Schwarz model and use a closed-form expression for the likelihood with given sufficient statistics to facilitate computation. Efficiently exploring model space through a Bayesian analysis using reversible jump Markov chain Monte Carlo methodology, we discuss the utility of model averaging for properly accounting for uncertainty in parameter estimation. We highlight the increase in understanding of the population obtained through our analysis.
Appeared as King, R and Brooks, S.P. (2003) "Survival and Spatial Fidelity of Mouflon: the Effect of Location, Age and Sex". Journal of Agricultural, Biological and Environmental Statistics 8 pp 486-513