Ungulate urea-nitrogen deposition and spatial patterning of herbaceous-layer vegetation in a temperate forest ecosystem

Background/Question/Methods

Increased atmospheric nitrogen deposition from human activities tends to decrease spatial heterogeneity in soil nitrogen availability. Ungulate-facilitated nitrogen excretion (i.e. urea nitrogen) could increase soil nitrogen heterogeneity by facilitating biogeochemical hotspots at the neighborhood scale (10-20 m). We used a set of three fenced white-tailed deer (Odocoileus virginianus) exclosures (4-30 ha area) and three unfenced reference sites to examine if deer can affect the spatial structure of herbaceous-layer plant communities via nitrogen hotspots. In spring and summer 2011, we measured percent cover of herbaceous-layer vegetation by species, density of woody seedlings and saplings, nitrogen ion availability (ammonium and nitrate) of mineral soil, and under-canopy radiation. Deer-source urea nitrogen deposition at unfenced sites was estimated from fecal pellet group counts using an algorithm relating fecal nitrogen to urinary nitrogen content. Within-site spatial structure was sampled using 1 m² quadrants placed in a repeating cyclical pattern along two intersecting 52 x 1 m belt transects. We analyzed the data non-spatially by comparing site-level means, and spatially by modeling semivariance as a function of distance. We hypothesized that unfenced sites would have more complex spatial structure in nitrogen availability and vegetation attributes compared to fenced deer exclosures.

Results/Conclusions

Non-spatial analysis did not reveal a difference between site-level means, with the exception of saplings which were absent or nearly absent at all unfenced sites. Semivariance modeling indicated more complex (more parameters in the best model) spatial structure in fenced sites than unfenced sites for ammonium and nitrate availability, vegetation percent cover, and under-canopy radiation. Species richness and Shannon’s diversity index did not differ in complexity of spatial structure. This finding suggests that deer affect the spatial patterning of nitrogen availability and vegetation percent cover by decreasing rather than increasing spatial structure at neighborhood scales. For nitrogen availability, deposition of urea-nitrogen in random locations within preferred deer habitat is likely breaking up larger scale spatial patterns in such a way that nitrogen availability appears random at a sampling grain size of 1 m². At our study sites, vegetation percent cover is likely less spatially complex at unfenced sites through the combination of two indirect effects of deer: removal of light competition with the sapling layer, and homogenization of nitrogen availability. Future studies of ungulate effects on nitrogen availability should sample at a grain size smaller than 1 m² to capture fine-scale heterogeneity.