Assessing distance sampling estimation under nonhomogeneous

Abstract

We conduct a simulation study to assess the performance of conventional distance sampling estimators when the spatial distribution of individuals in the study area follows a non-homogeneous Poisson process. The following two approaches are considered: 1) objects in the study area are fixed, but different transects are randomly allocated in every simulation; 2) transects are fixed, but objects are randomly placed in every simulation. Different intensity functions are used to simulate non-homogeneous Poisson processes and six different scenarios are considered for objects location in study area. Transects are positioned randomly or systematically with a random start.