N-Mixture Models

In this chapter, we will introduce N-mixture models (Wyatt 2002; Royle 2004) and extensions that use count data to estimate true abundance of species. These models, in comparison to occupancy models, further move away from classic metapopulation models in the sense that they can work without 0’s (or changes between 0’s and non-0’s in dynamic versions; Dail and Madsen 2011, Hostetler and Chandler 2015) in true abundance. The dynamic versions of N-mixture models also provide some demographic information, as the change in true abundance is exactly due to the BIDE processes. However, we will see that dynamic N-mixture models’ ability to provide demographic information is still limited, due to stringent model assumptions.