Summary
In forests of eastern North America, white‐tailed deer (Odocoileus virginianus) can directly affect, via herbivory, the presence, abundance and reproductive success of many plant species. In addition, deer indirectly influence understorey communities by altering environmental conditions.
To examine how deer indirectly influence understorey plants via environmental modification, we sampled vegetation and environmental variables in‐ and outside deer exclosures (10–20 years old) located in temperate forests in northern Wisconsin and the Upper Peninsula of Michigan, USA. We assessed how excluding deer affected understorey community composition and structure, the soil and light environment, and relationships between direct and indirect effects, using non‐metric multidimensional scaling (NMDS), mixed linear models and nonparametric multiplicative regression (NPMR).
Excluding deer altered sapling communities and several aspects of the understorey environment. Excluding deer from plots with lower overstory basal area increased sapling abundance, decreasing the amount of light available to groundlayer plants. Exclusion also reduced soil compaction and the thickness of the soil E horizon.
The composition of understorey communities covaried in apparent response to the environmental factors affected by exclusion. In several common species and groups, E horizon thickness, compaction, openness, and/or total (sapling and overstory) basal area were significant predictors of plant frequency.
Complementary analyses revealed that deer exclusion also altered the frequency distributions of several species and groups across environmental space.
Synthesis. Deer alter many facets of the understorey environment, such as light availability, soil compaction and thickness of the soil E horizon, which, in turn, appear to mediate variation in plant communities. Those environmental modifications likely compound direct impacts of herbivory as drivers of understorey community change. Our results provide evidence that deer effects on the environment have important implications for forest composition. Thus, we suggest a re‐examination of the common assumption that understorey community shifts stem primarily from tissue removal.