Predicting winter abundance of woodcock Scolopax rusticola using weather data: implications for hunting management

The sustainable use of natural resources such as game animals requires adjusting extraction to changes in population abundance. Population abundance monitoring is thus necessary to ensure an adaptive management, but this can be difficult in the case of migratory species where breeding areas are in remote places without local monitoring programs. Predictive models of the winter abundance based in the relation between climate and reproduction success or survival could be a useful alternative to monitoring networks in the breeding areas. In this paper, we evaluate the role of weather variables as indicators of winter abundance estimates. We used Game Abundance Indices (total number of woodcock observed during hunting days, divided by the number of hunting hours), collected by volunteer hunters during 21 seasons, and temperature, rainfall and number of days with snow, calculated in May, June and July in the breeding areas and December to January in the winter areas. The best models explaining variations in winter abundance included number of rainy days in May and June and temperature in July as explanatory variables. All variables were positively correlated with abundance except temperature in July. The predictive quality of the best model based on a leave-one-out cross-validation procedure (i.e. the Pearson correlation coefficient between observed values and LOO-predicted values) was 0.76. We discuss the applications of this predictive model to develop an adaptive hunting management scheme for the species.