Assessing the precision of wild boar density estimations

Abstract

Effective management of wild boar (Sus scrofa) populations has to be based on precise estimates of local densities. The development of an effective and cost-efficient technique to cope with this need has always represented a challenge for wildlife managers and researchers. Drive counts, hunting bags, and Random Encounter Model (REM) are among the most frequently used techniques, with the latter recently gaining wide recognition. We sought to compare the 3 methods in terms of their suitability for management, precision, and effort required. Moreover, we evaluated the uncertainty of REM results when all sources of error were considered. In our study, the 3 methods were applied to a wild boar population of the Italian Apennines in 2013. We used the delta method to assess the total uncertainty of REM density estimates on the basis of the errors associated to all the parameters involved. Notably, the 3 methods tested showed consistent mean density estimates, though none of them reached fully satisfying levels of precision for management purposes. Since the low precision of REM was mostly due to the high variability of the group-size parameter, we propose simple technical improvements aimed at reducing the variability of this parameter and, thus, of REM. Although all the methods tested still need to be further developed to be effective for wild boar management, REM seems to be the most promising one in terms of both potential precision and effort required. The limited effort required by REM is particularly relevant in the current wildlife management scenario, where funds are often lacking and the number of hunters acting as volunteers is decreasing.