Estimating the Population Size for Capture-Recapture Data with Unequal Catchability

TLDR

The study proposes a point estimator and confidence interval for closed population size that accounts for heterogeneous capture probabilities. The estimator is illustrated with real wildlife data and evaluated through Monte Carlo simulations to assess its performance. The estimator performs well when most individuals are captured only once or twice, outperforming the jackknife, and its confidence interval reliably maintains nominal coverage.

Abstract

A point estimator and its associated confidence interval for the size of a closed population are proposed under models that incorporate heterogeneity of capture probability. Real data sets provided in Edwards and Eberhardt (1967, Journal of Wildlife Management 31, 87-96) and Carothers (1973, Journal of Animal Ecology 42, 125-146) are used to illustrate this method and to compare it with other estimates. The performance of the proposed procedure is also investigated by means of Monte Carlo experiments. The method is especially useful when most of the captured individuals are caught once or twice in the sample, for which case the jackknife estimator usually does not work well. Numerical results also show that the proposed confidence interval performs satisfactorily in maintaining the nominal levels.