In this paper, we develop a model to evaluate the capacity of accommodation areas for overwintering waterfowl, at a large spatial scale. Each day geese are distributed over roosting sites. Based on the energy minimization principle, the birds daily decide which surrounding fields to exploit within the reserve boundaries. Energy expenditure depends on distance to the roost and weather conditions. Food intake rate is determined by functional responses, and declines with consumption. A shortage occurs when birds cannot fulfil their daily energy requirement. Most foraging takes place on pasture, with complementary feeding for some of the species on cereals and harvest remains. We applied the model to five waterfowl species overwintering in the Netherlands. From a comparison with field data, the model appears to produce realistic grazing pressures on pasture, especially for geese, and a realistic decline in sward height, but the use of arable fields is less in agreement with observations. For current goose and wigeon numbers, hardly any shortages are expected, but extrapolating the population increase observed during the last decade, considerable shortages are expected in the near future (2015). However, we find that several uncertainties may contribute to more severe shortages: a probabilistic (and therefore less optimal) choice of foraging location, a shorter maximum distance to the roost, and a lower effective availability of resources due to disturbances and other edge effects. Between species we find both competition and facilitation. Both type of interactions, as well as the spatial pattern of resource exploitation, are explained from functional responses and energetic costs of the species.