COS 37-7
A national synthesis of atmospheric nitrogen deposition relationships with herbaceous plant diversity
Broad-scale and spatially heterogeneous patterns of atmospheric nitrogen deposition within the US could influence herbaceous plant diversity, but would only be detectable if plant responses seen in site-specific experimental nitrogen additions to plains, forest, desert, and alpine ecosystems are not obscured by other environmental gradients such as elevation, soil, and climate. To address this synthetic question we assembled herbaceous plant species composition and abundance data from 24,196 vegetation plots in 20 US datasets from individual investigators, state Natural Heritage Programs, the FIA Phase 3 vegetation indicator database, the NPS I&M program, and the ESA VegBank archive. Where multiple subplots per plot were available, we seamlessly rarefied or extrapolated plant diversity to a common species completeness of 90% to mitigate variable rates of species detection. We calculated total atmospheric nitrogen deposition from modeled CMAQ dry deposition and interpolated NADP wet deposition grids, and then obtained elevation, soil, and climate data from NED, SSURGO, and PRISM, respectively. Our analyses to date have focused on the frequency of positive, negative, and neutral plant diversity vs. nitrogen deposition relationships (alone and interacting with the other environmental covariates), after separating by source dataset and vegetation community (National Vegetation Classification).
Results/Conclusions
Atmospheric nitrogen deposition reduced herbaceous species diversity along some gradients but not others. Relationships were dependent on vegetation community type, and were modulated by strong interactions with climate and elevation. Divergent diversity responses are not surprising, given expected variability in baseline nitrogen availability and soil acidity, degree of nitrogen limitation relative to phosphorus and water limitations, differences in the pre-existing species pools, and steepness of gradient. We emphasize that our approach is well suited for vegetation communities that are distributed along moderate to large existing gradients of nitrogen deposition. Rare community types not distributed across nitrogen gradients will require other approaches. Likewise, some vegetation communities may exist entirely below or above their critical load of nitrogen deposition, with the former not yet showing a response and the latter potentially having already lost sensitive species before they were sampled. Some of our other analyses in progress include evaluating nitrogen deposition relationships with the diversity and abundance of plant functional groups (e.g., graminoids vs. forbs and exotics vs. natives), investigating response curves of individual species, and including a more comprehensive set of environmental covariates in those more localized regions where such data exist.