We summarize the information on the diet of roe deer Capreolus capreolus found in 33 European studies. After giving a short overview of the differences between the existing studies, we compare the information for each season. We submit the information, summarized in a matrix of 83 cases on 10 food groups, to a detrended correspondence analysis (DCA) and a two-way indicator species analysis (TWINSPAN). We calculate weighted averages grouping the information by season, habitat, research method and their crossproducts. The weighted averages are also used as input for a multivariate ratio analysis. Since the available food items dictate the possible diet composition we further investigate the in¯uence of the habitat on the reported food selection. The in¯uence of season on the diet composition is compared with the effect of the habitat, and other factors such as research method and geographical location of the study site. The review shows that there is relatively little seasonal variation in the diet composition, which is more closely correlated to the habitat than to the season.

We summarize the information on the diet of roe deer Capreolus capreolus found in 33 European studies. After giving a short overview of the differences between the existing studies, we compare the information for each season. We submit the information, summarized in a matrix of 83 cases on 10 food groups, to a detrended correspondence analysis (DCA) and a two-way indicator species analysis (TWINSPAN). We calculate weighted averages grouping the information by season, habitat, research method and their crossproducts. The weighted averages are also used as input for a multivariate ratio analysis. Since the available food items dictate the possible diet composition we further investigate the in¯uence of the habitat on the reported food selection. The in¯uence of season on the diet composition is compared with the effect of the habitat, and other factors such as research method and geographical location of the study site. The review shows that there is relatively little seasonal variation in the diet composition, which is more closely correlated to the habitat than to the season.

BackgroundSound decisions on control actions for established invasive alien species (IAS) require information on ecological as well as socio-economic impact of the species and of its management. Cost-benefit analysis provides part of this information, yet has received relatively little attention in the scientific literature on IAS.MethodsWe apply a bio-economic model in a cost-benefit analysis framework to greater Canada goose Branta canadensis, an IAS with documented social, economic and ecological impacts in Flanders (northern Belgium). We compared a business as usual (BAU) scenario which involved non-coordinated hunting and egg destruction with an enhanced scenario based on a continuation of these activities but supplemented with coordinated capture of moulting birds. To assess population growth under the BAU scenario we fitted a logistic growth model to the observed pre-moult capture population. Projected damage costs included water eutrophication and damage to cultivated grasslands and were calculated for all scenarios. Management costs of the moult captures were based on a representative average of the actual cost of planning and executing moult captures.ResultsComparing the scenarios with different capture rates, different costs for eutrophication and various discount rates, showed avoided damage costs were in the range of 21.15 M€ to 45.82 M€ under the moult capture scenario. The lowest value for the avoided costs applied to the scenario where we lowered the capture rate by 10%. The highest value occurred in the scenario where we lowered the real discount rate from 4% to 2.5%.DiscussionThe reduction in damage costs always outweighed the additional management costs of moult captures. Therefore, additional coordinated moult captures could be applied to limit the negative economic impact of greater Canada goose at a regional scale. We further discuss the strengths and weaknesses of our approach and its potential application to other IAS.