R Code for occupancy simulations using R Shiny app featured in Methods in Ecology and Evolution review (Kitzes et al. 2025)

This app supports the exploration of how different stages of the the sensors + AI pipeline affect bias and precision in single-season static occupancy models. Select parameters using the sliders, then click the button to simulate 10 data sets and fitted models. The parameter ranges below correspond roughly to those for a typical temperate bird survey in which recorders are deployed at many points for multiple days, although the same simulations apply to other vocalizing taxa

NOTES: The models may fail to estimate upper and lower confidence interval bounds for psi (returning NA) under certain combinations of parameters. This can occur when the simulated detection history consists either of all 0's or all 1's for every site (i.e., there is no variability in detection of presences within any site). Try adjusting parameters to lower detectability and/or the false positive rate to address this issue.