Estimating the Size of a Population Through Repeated Sampling: A Survey Sampling View on Capture-Recapture...

Area of Science:

• Ecology
• Statistics
• Population Dynamics

Background:

• Capture-recapture methods are crucial for estimating elusive population sizes using repeated observations.
• Current methods often model capture frequencies or histories, with less focus on sampling design.
• Latent heterogeneity and variable sample sizes present ongoing challenges in population estimation.

Purpose of the Study:

• To introduce a novel survey sampling framework for capture-recapture modeling.
• To develop population size estimators adaptable to specific experimental data collection methods.
• To address limitations in current methods, including latent heterogeneity and variable sample sizes.

Main Methods:

• Framing capture occasions as sample draws within a survey sampling approach.
• Utilizing probability proportional to size (PPS) sampling as a specific design assumption.
• Modeling individual inclusion probabilities to account for heterogeneity and variable sample sizes.
• Developing a novel simulation framework for evaluating estimator performance under unequal probability sampling.

Main Results:

• The proposed estimator demonstrates competitive performance, matching the accuracy of established methods.
• Simulations reveal insights into estimator behavior under varying sampling effort and heterogeneity.
• Applications to 14 benchmark datasets validate the method's effectiveness.

Conclusions:

• The survey sampling approach offers a flexible and effective alternative for capture-recapture modeling.
• The new method successfully accommodates latent heterogeneity and variable sample sizes.
• This framework provides a valuable tool for accurate population size estimation in ecological and biological studies.