In community ecology, interactions are pervasive across time and space. Separating the relative effects of individual factors involved in ecological interactions requires direct experimental manipulation of organisms, their environment, or both. In oak-hickory forests of central Pennsylvania, browsing by white-tailed deer (Odocoileus virginianus), acid deposition, and competition from dominant vegetation are all thought to reduce understory diversity and limit tree regeneration, however, to date no direct experimental manipulation has occurred to evaluate these factors simultaneously. To evaluate their effect on understory plant community composition, we established a network of 11 1/2500th ha microplots at 24 permanent sites in the Rothrock and Bald Eagle State Forests in 2014. Following establishment and full vegetation inventories at all sites, 1 of 7 full factorial vegetation treatment combinations of fence (to remove deer), lime (to improve soil chemistry) and herbicide (to remove competing vegetation) was randomly assigned and applied to subplots 5-11 at each location. Vegetation inventories were repeated at the same locations in 2016 (2 years post-treatment) and data for each treatment combination (including untreated controls) was compared for 4 biologically significant plant groups (deer-preferred taxa, ericaceous competing vegetation, seedlings considered acceptable regeneration, and all seedlings combined) using generalized linear mixed effects models.
Results/Conclusions
Control areas showed no change in abundance for any plant group from 2014 to 2016. For deer-preferred plants, the proportion of flowering individuals increased with liming application, indicating a potential relationship between soil chemistry and reproductive status. Deer exclusion did not change seedling abundance (desirable or otherwise) and did not increase the proportion of flowering individuals for deer-preferred taxa from 2014 to 2016. Fencing also had no effect on diversity of understory cover or seedling taxa. Herbicide application reduced the abundance of ericaceous competing vegetation across all herbicide treatments, supporting the potential for competitive release of surviving taxa. Despite the use of a broadcast non-targeted herbicide mix, deer-preferred plant abundance remained unchanged from 2014 to 2016 across all herbicide treatment combinations. Herbicide treatment alone reduced desirable seedling abundance, while combinations treatments did not.
Although results should be considered preliminary due to the short-term nature of the study, plant community responses to single factors were limited, minimizing support for deer herbivory, poor soil chemistry, or interspecific competition as independent drivers of community composition. Interactions between these factors were prevalent across most treatment combinations, and continued monitoring of these sites is needed to better understand long-term trends.