Long-term effects of herbivory on tree growth are not consistent with browsing preferences

Herbivores affect plants via direct consumption, but also indirectly via changes in the vegetation. These indirect effects may only manifest after decades of exposure and, as a result, their impact is rarely accounted for. To better understand the direct and indirect impacts of a large herbivore, moose (Alces alces), on forests, we measured tree growth in areas that were both subjected to and excluded from herbivory for over 80 years. Growth data were gathered from five tree species, ranging from low to high palatability. We found that at small sizes, Betula papyrifera, a preferred species, benefited from herbivore exclusion. However, larger individuals grew more when exposed to herbivory, a response we attribute to lower competition in heavily browsed conditions. Populus tremuloides, a highly preferred tree, did not show any differences between levels of herbivory. Abies balsamea, a preferred winter browsed, was only marginally affected by browsing at smaller sizes. The two non-preferred species, Picea glauca and Picea mariana, did not show differential growth between herbivory levels. We conclude that herbivores can impact forests through both direct and indirect effects, that these effects are size specific, and that effects vary among species in ways not always predicted by consumption patterns.