Confronting spatial capture-recapture models with realistic animal movement simulations.

Area of Science:

• Ecology
• Wildlife Biology
• Statistical Modeling

Background:

• Spatial capture-recapture (SCR) models are vital for estimating mobile animal population densities.
• Previous evaluations of SCR models relied on simplified simulated animal movement data.
• Realistic movement patterns, driven by individual responses to environmental factors, are crucial for accurate ecological modeling.

Purpose of the Study:

• To evaluate the robustness of various SCR models using data simulated from a mechanistic individual-based movement model.
• To assess how well SCR models capture density and detectability influenced by resource availability and individual movement.
• To investigate the challenges in inferring home-range size and fitting transience models with realistic movement data.

Main Methods:

• Generated individual movement trajectories using a mechanistic individual-based model incorporating resource memory and territoriality.
• Simulated detection data from these movement trajectories.
• Fitted basic and variant SCR models (resource selection, transience) to the simulated data, including those with covariates for density and detectability.

Main Results:

• Abundance estimates from SCR models were robust to low-degree violations of movement assumptions.
• SCR models successfully identified the positive relationship between resource quality and population density.
• Covariate models struggled to detect fine-scale effects of resource quality on detectability and space use, possibly due to data resolution or model misspecification.

Conclusions:

• Integrating complex movement into SCR models may not be essential for accurate abundance estimates if individual heterogeneity is low.
• More complex movement processes are warranted for understanding finer-scale ecological and movement patterns.
• The study provides a framework for generating realistic SCR data to develop and evaluate advanced movement models.