REM: performance on a high-density fallow deer (Dama dama) population

Abstract

We present an application of the random encounter model (REM) to estimate population density of a closed population of fallow deer (Dama dama). REM promises to be a powerful tool for providing density estimates for species which are not individually recognisable, but it still requires thorough testing to assess its limits and performance. In our study area, effective fencing prevents animals from migrating in or out of the area thus assuring a closed population; still the application of REM presented several challenges, including a very high density (more than 600 ind/km2) and a constant fission/fusion of groups of the study species. We applied both stratified and unstratified analysis approach, and we estimated density over a range of daily range values. The REM approach underestimates the true density value of about 31% and 28% the CV of the estimates was 0.39 and 0.52, for stratified and unstratified approach, respectively. Still it provides evidence in support of REM as a method for providing density estimates of free-ranging unmarked individuals with a fission–fusion social structure.