Bayesian inference in camera trapping studies for a class of spatial capture-recapture models

Area of Science:

• Ecology
• Wildlife Biology
• Statistical Modeling

Background:

• Spatial capture-recapture (SCR) methods are crucial for estimating wildlife abundance and density.
• Camera trapping is a widely used non-invasive technique for wildlife monitoring.
• Existing SCR models often struggle to account for complex spatial processes and trap configurations.

Purpose of the Study:

• To develop a flexible hierarchical model for spatial capture-recapture data.
• To improve abundance and density estimations in camera trapping studies.
• To explicitly incorporate individual movement and trap-specific capture probabilities.

Main Methods:

• Developed a hierarchical model with a point process for individual spatial distribution and an observation model for trap encounters.
• Formulated the model as a generalized linear mixed model with individual home range centers as random effects.
• Employed a Bayesian inference framework using data augmentation.

Main Results:

• The proposed models provide a robust framework for analyzing spatial capture-recapture data.
• Demonstrated the model's applicability using camera trapping data of tigers in India.
• Successfully addressed the complexities of trap movement and varying camera operationality.

Conclusions:

• The developed hierarchical models offer enhanced accuracy for estimating wildlife abundance and density.
• The Bayesian data augmentation approach provides a powerful tool for inference in complex SCR studies.
• This methodology is valuable for wildlife management and conservation efforts, particularly in camera trapping scenarios.