A Comparative Study of Bayes Estimators of Closed Population Size from Capture-Recapture Data

摘要

Capture-Recapture models estimate the unknown sizes of animal populations. For closed populations, with constant size N during the study, eight standard models exist for estimating N. These models allow for variation in animal capture probabilities due to time effects, heterogeneity among animals, and behavioral effects after the first capture. We present Bayesian versions of these eight models. Through simulation, we compare performance of estimation of N under each model. Each model is fit to data sets generated under each of the eight model assumptions, allowing assessment of model robustness. In our simulation conditions, the most robust model in estimating N is found to be the model with behavioral and time effects. Finally, we illustrate our methods by applying them to a population of cottontail rabbits.