Methods of Predicting Risks of Wheat Damage by White-Fronted Geese

Abstract: Accurately predicting occurrence of wildlife damage is crucial for effective management of problematic wildlife species, because accurate predication allows deterrence efforts to be focused at sites or times where damage is most likely. We explored methods to predict occurrence of white-fronted geese (Anser albifrons) grazing in wheat fields around Lake Miyajimanuma, Japan. Depletion of waste rice grains caused geese to forage on wheat leaves in spring, reducing wheat harvest in grazed fields. The cumulative number of goose-days per hectare of rice-planted area from the beginning of the staging period explained the variation in the proportion of geese foraging in wheat fields. A logistic regression model on the location of vulnerable fields showed that goose grazing was likely to occur in wheat fields far from roads and windbreaks and those close to (within 1,000–2,000 m of) previously grazed fields. Although probability of occurrence of goose grazing was initially low in wheat fields with scaring devices, effectiveness of such devices was lost over the 4 survey years. We recommend farmers in the study area prepare counter-damage measures when the cumulative number of goose-days per rice-planted area approaches a threshold above which some geese are predicted to start foraging on wheat (e.g., 199.46 goose-days/ha rice × 28.95 for 10% of geese foraging on wheat). Further, farmers should be aware that grazing on wheat is more likely to occur if wheat fields within 1,000–2,000 m have already been exploited during that particular season and should concentrate deterrence efforts to wheat fields that are far from roads and windbreaks. Systematic deployment of scaring devices over the entire habitat has a risk of accelerating the decline in effectiveness. Thus, we need methods to retard goose habituation to scaring devices, such as scaring with guns, providing alternative feeding sites, and preventing diet change by geese.