On continuous-time capture-recapture in closed populations.

Area of Science:

• Ecology
• Biometrics
• Statistical Modeling

Background:

• Continuous-time capture-recapture models are crucial for population size estimation.
• Discretizing continuous data for traditional models can lead to information loss.
• Previous work by Schofield et al. (2018) introduced Poisson process-based models.

Purpose of the Study:

• To clarify the importance of the Poisson model in Schofield et al.'s approach.
• To demonstrate that data discretization with a Poisson model is effective.
• To compare discrete-time Poisson models with continuous-time methods.

Main Methods:

• Applying Poisson models to capture counts for individuals.
• Utilizing data discretization techniques for capture-recapture data.
• Comparing population size estimates from different modeling approaches.

Main Results:

• The core of Schofield et al.'s method is the Poisson assumption, not continuous time.
• Discretizing data and applying a Poisson model yields results comparable to continuous-time methods.
• Using a Bernoulli model on discretized data leads to information loss.

Conclusions:

• Poisson models are key for accurate capture-recapture analysis, regardless of time scale.
• Discretization is a viable strategy when combined with appropriate Poisson modeling.
• Avoid traditional Bernoulli models on discretized data for robust population estimates.