Spatially explicit capture-recapture models for relative density

Area of Science:

• Ecology
• Wildlife Biology
• Statistical Modeling

Background:

• Spatially explicit capture-recapture (SECR) is crucial for estimating animal population parameters.
• Maximum likelihood is applied to SECR models, including full and conditional models.
• The conditional model assumes uniform density, while the full model allows spatially varying density.

Purpose of the Study:

• To extend the conditional SECR model for estimating relative density as a function of spatial variables.
• To demonstrate the robustness of relative density coefficients to heterogeneous detection.
• To highlight the advantages of the conditional model, including covariate inclusion and computational efficiency.

Main Methods:

• Utilizing maximum likelihood for two SECR model flavors: full and conditional.
• Extending the conditional model to incorporate habitat and spatial variables for relative density estimation.
• Demonstrating model performance through examples and simulation studies.

Main Results:

• Conditional SECR models can estimate relative density, which is robust to detection heterogeneity.
• The conditional model efficiently integrates individual detection covariates.
• Absolute density surfaces and population growth rates can be derived from conditional model fits.

Conclusions:

• Conditional likelihood models are versatile and efficient for SECR analyses.
• These models provide a flexible framework for estimating density, relative density, and population growth rates.
• Conditional models are recommended for a central role in advancing SECR methodologies.