Clark Rushing

I am trying to fit a N-mixture model that uses the correlation amongst adjacent spatially replicated counts (i.e., stops) to estimate probablity of detection and the total number of individuals present at a site. Following Royle & Kery (2015), the model uses a multinomial likelihood, with the stop level counts conditioned on the observed "total" count. The model works fine when there are no 0 counts in the observed data. However, when there are 0 counts, JAGS returns the following error: Error in...