Assessment of bias and precision among simple closed population mark-recapture estimators

Mark-recapture methods have been heavily studied and employed in fisheries and other wildlife sciences over the past century to approximate population sizes for animal species of interest. This paper focuses on the compar-ative statistical performance through simulation of common closed population mark-recapture estimators, including those of Lincoln-Petersen, Chapman, Chao, Schnabel, and Schumacher-Eschmeyer. A new bias-adjusted version of the Schumacher-Eschmeyer estimator is proposed and is shown to exhibit superior performance at small sample sizes in comparison to the original estimator. Simulation results indicate that Chapman's method outperforms all other two-visit methods and that bias-adjusted versions of Schnabel and Schumacher-Eschmeyer differ slightly depending on bias or precision, but both perform well. Minimum sample sizes such that resulting estimates are approximately unbiased are proposed to advise practitioners on the most appropriate use of these estimators for simple closed population mark-recapture data.

Automated summary

This paper compares the statistical performance of common closed population mark-recapture estimators through simulation, finding that a new bias-adjusted version of the Schumacher-Eschmeyer estimator outperforms the original estimator at small sample sizes. The authors propose minimum sample sizes to achieve approximately unbiased estimates, providing guidance for practitioners using these estimators for simple closed population mark-recapture data.