Failure to meet the exchangeability assumption in Bayesian multispecies occupancy models: Implications for study design

Land managers seek to understand whether treatments they implement
effectively achieve their goals to support, or at least not negatively
impact, species of concern. This power analysis identifies minimum data
collection needs to estimate treatment impacts when baseline species
community information is lacking, and when there is significant risk of
violating the species exchangeability assumption, such as when the
community is composed of multiple guilds.
This software simulates detection/non-detection community occupancy
datasets with several categorical treatment effect scenarios, fits
Bayesian hierarchical multispecies, single species, and hybrid occupancy
models, and calculates error metrics for species-specific treatment
effects.