Herbaceous plant species diversity relationships with nitrogen deposition: Soil and climate contingencies across the USA

Background/Question/Methods

Herbaceous plant species that are rooted in soil and have moderate to short lifespan are indicators of nitrogen (N) deposition that complement the benefits of a) lichen indicators that are unbuffered by soil and b) tree indicators that typically have longer generation times. Herbaceous plant species are likely to provide the first indication of changes in vascular plant species richness and composition in response to N deposition. To investigate relationships between herbaceous plant species richness and N deposition, and to tease apart N deposition from climate and soil influences,  we began by assembling herbaceous plant species data from 12,970 forested, shrubland, and grassland sites from existing datasets distributed across the continental USA. We then extracted corresponding estimates of N deposition (from modeled CMAQ dry deposition and interpolated NADP wet deposition), soil pH (from SSURGO), and climate (temperature and precipitation from PRISM). Finally, we regressed herbaceous plant species richness again the predictors described above, as well as their interactions, at the following two scales: 1) national-scale, without regard to vegetation type, and 2) gradient-scale, with separate analyses for each vegetation type by source dataset combination that had an adequate N deposition gradient and number of sites.

Results/Conclusions

In the national-scale analyses, total herbaceous plant species richness had a hump-shaped relationship with N deposition that was most pronounced above 10 kg ha^-1 yr^-1 and at low soil pH. The contingency of species richness declines upon both N deposition and soil pH is consistent with a soil acidification mechanism of species loss, but has not previously been demonstrated at such a broad scale. In gradient-scale analyses, total richness declined in 35% of gradients and increased in 17% of gradients, while in contrast the introduced invasive fraction of species increased in 56% of gradients and declined in only 16% of gradients. Collectively, increases in invasive species may be partially masking losses of native species, requiring additional analyses of species composition and evenness. Precipitation moderated total species richness relationships with N deposition, but unlike soil pH the direction of influence varied from the national-scale to the gradient-scale. In conclusion, we successfully identified some of the ecosystems where herbaceous plant species are most vulnerable to N deposition, demonstrated that N deposition influences on herbaceous species are partially contingent on soil pH and precipitation, and began to unravel the complexities of disparate plant functional groups.