Increase in crop damage caused by wild boar (Sus scrofa L.): the "refuge effect"

The occurrence of crop damage by wild boars raised dramatically in the last decades, implying an increase in social conflicts, expenditures for compensation and a risk to natural ecosystems. Many researchers have explained this phenomenon by considering wild boar biology, behaviour and abundance. Little or no attention has been devoted to wildlife management and the agricultural mosaic. We hypothesised that the agricultural structure of the landscape and wildlife management planning, including hunting, can play a relevant role in causing crop damage. We studied a Mediterranean area in central Italy that is characterised by a patchy agriculture, dividing the surface into hexagons. A large number of terrain parameters were calculated at the large (hexagons) and local (buffer) scale, including the topography, land use and agricultural management. We also considered wildlife variables such as the number of wild boar shot down, hunting management and the legal status of the wild boar. The terrain and management data for each hexagon were submitted to a generalised binomial stepwise multiple logistic regression. The resultant model demonstrated an accuracy of 0.76, a misclassification rate of 0.24 and an odds ratio of 10.41. The most important variables selected by the regression were the woods in the area where hunting was banned (P < 0.001), a 1-km buffer of intensively cultivated farmland close to the woods where hunting was banned (P < 0.001), a 1-km buffer of intensively cultivated farmland along the river (P < 0.05), the forest edge (P < 0.001), and the mean number of wild boar that were shot (P < 0.05). In this study, we proved that an important factor in explaining crop damage is the "refuge effect" (a buffer close to the wooded areas where hunting was banned) and the 1-km buffer along possible dispersion routes.