Bayesian inference was used to estimate
parameters. We specified flat prior distributions for both the detection and occupancy processes, and a Gamma (0.5, 0.0005)
distribution for the spatial process (Johnson et al., 2013). After a burn of 10 000 iterations, parameter posterior distributions
were computed from 40 000 iterations and used to find parameter mean, SD, and 90% Bayesian credible interval (BCI) values.
Occupancy models were performed using the package ‘‘stocc’’ in program R (R Development Core Team, 2009).