Bimodality of the combined removal and signs-of-activities estimator for sampling closed animal populations

摘要: The possibility of a bimodal log-likelihood function arises with certain data when the combined removal and signs-of-activities estimator is used. Bimodal log-likelihoods may, in turn, yield disjoint confidence intervals for levels. hypothesis that bimodality caused by violation equal catchability assumption model, leading to combination contradictory data/models set forth. Simulations exploring effect model assumptions on estimation inference showed unequal capture probability influenced frequency likelihoods; similarly, extreme parameter values number excessively large produced. A sex-specific developed as remedial tailored problem. simulations suggest both perform equally well over range presented, though easier implement.