Selectivity by moose vs the spatial distribution of aspen:: a natural experiment

Patch use theory predicts that herbivores perceive food as patches and spend more time in high quality patches, i.e. feeding sites providing a high net rate of intake of energy ana::or limiting nutrients. The herbivores should accordingly not discriminate among food items in such high quality patches, and food items should thus be eaten in proportion to availability. In contrast, classical diet theory assumes food selection to take place at the level of individual plants. and predicts that the forager should concentrate on the most profitable item until availability drops below some critical threshold. Here we address how the spatial distribution of European aspen Popidus tremula, a highly preferred browse species, affects the selectivity by moose Alces alces at the patch and the tree level. The study was performed in a managed boreal forest landscape in coastal northern Sweden. where aspen occurs highly aggregated almost exclusively in discrete patches. We compared moose selectivity for aspen and browsing intensity on aspen ramets and other browse species in aspen patches versus at randomly located sites. Random sites and aspen stands were utilised equally by moose in terms of overall use of forage. There was no difference in total coverage of forage species and relative moose density. Selectivity for aspen was stronger at random sites than at aspen sites, and the browsing intensity on aspen was similar, We conclude that moose did not perceive aspen stands as discrete patches, and used aspen ramets more in accordance with diet theory. These findings agree with the idea that large generalist herbivores strive to maintain a mixed and balanced diet, causing rare species to be over-utilised (negative frequency-dependent food selection). By such selective feeding. moose may reinforce the aggregated distribution of aspen in the managed boreal forest landscape.