A comparison of abundance estimates from extended batch-marking and Jolly-Seber-type experiments

Area of Science:

• Ecology
• Population Dynamics
• Statistical Modeling

Background:

• Traditional capture-recapture methods often require individual capture histories for accurate population estimates.
• Multi-sample batch-marking studies have been underutilized due to assumptions about data requirements.
• Previous work by Huggins et al. (2010) introduced a pseudo-likelihood approach for batch-marking data.

Purpose of the Study:

• To develop and maximize a likelihood function for multi-sample batch-marking studies.
• To compare the efficiency and accuracy of abundance estimation methods in batch-marking designs.
• To evaluate the performance of the Crosbie-Manly-Arnason-Schwarz (CMAS) model against other estimators.

Main Methods:

• Developed and maximized a novel likelihood function for batch-marking data.
• Utilized simulated data from a Jolly-Seber-type study, adapted for batch-marking scenarios.
• Compared abundance estimates from the CMAS model, Huggins' pseudo-likelihood, and the new likelihood function.

Main Results:

• Abundance estimates were comparable across the CMAS, Huggins, and the newly developed likelihood estimators.
• The new likelihood function generally exhibited lower mean square error than Huggins' pseudo-likelihood method.
• Employing unique identifiers and the CMAS model led to improved precision in abundance estimates.

Conclusions:

• Researchers can confidently use batch-marking studies to obtain unbiased abundance estimators.
• The developed likelihood function offers an improvement in precision and reduced mean square error.
• Study design elements, such as capture probabilities and sample size, can be manipulated to minimize mean square error.