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

Royle, Karanth, Gopalaswamy, and Kumar developed a Bayesian implementation of spatially explicit capture-recapture tailored to camera-trap data, using data augmentation and Markov chain Monte Carlo to estimate density while modeling animal locations as latent activity centers. Applied to tiger camera-trap surveys, the approach naturally handled the spatial encounter process at fixed camera stations and produced density estimates with full posterior uncertainty without arbitrary buffer choices. The Bayesian framework made it straightforward to incorporate covariates and individual heterogeneity. This paper, together with its companion software, popularized SECR among camera-trap practitioners and underpinned tools such as SPACECAP. It is a foundational methodological reference that helped shift big-cat density estimation from buffered closed-population models to spatially explicit inference.