PS 28-159 - Invasive grass has greater effect on ecosystem processes in forests with dissimilar nutrient economies

Tuesday, August 8, 2017
Exhibit Hall, Oregon Convention Center
Laura Y. Podzikowski1, Marissa R. Lee2, Catherine Fahey3, Justin P. Wright4, S. Luke Flory5 and Richard P. Phillips1, (1)Biology, Indiana University, Bloomington, IN, (2)University Program in Ecology, Duke University, Durham, NC, (3)Interdisciplinary Ecology, University of Florida, Gainesville, FL, (4)Biology, Duke University, Durham, NC, (5)Agronomy Department, University of Florida, Gainesville, FL
Background/Question/Methods:  We tested the hypothesis that the biogeochemical consequences of an invasive grass, Microstegium vimineum (M.v.), are greatest in plots where the dominant native plants possess nutrient use traits most dissimilar to M.v. Commonly known as Japanese stiltgrass, M.v. is a persistent understory invader in forests of the eastern and midwestern US that has fast cycling nitrogen (N) use traits (e.g., fast-decaying N-rich litter, promotion of microbial nitrification and rapid N uptake). To test the trait dissimilarity hypothesis, we established paired M.v.-invaded and non-invaded plots in forest stands that differed in relative abundance of trees with fast-cycling N use traits (e.g., trees associating with arbuscular mycorrhizal fungi; “AM”) and slow cycling N use traits (e.g., trees associating with ectomycorrhizal fungi; “ECM”). The forest stands were at three locations (Georgia, Indiana, and North Carolina) across a gradient of trait dissimilarity. We sampled surface soils (0-5cm) and aboveground vegetation at peak biomass from 2014-2016 to quantify soil N and C cycling.

Results/Conclusions:  Non-invaded, ECM-dominated stands were characterized by low net nitrification rates, low soil pH, and high soil C:N, whereas non-invaded AM-dominated stands were characterized by high net nitrification rates, high soil pH, and low soil C:N. In support of our hypothesis, we observed the largest effect of M.v. in ECM-dominated stands. Averaging normalized effect sizes (g) across sample year for all three sites, we found M.v. enhanced net nitrification rates (1.93±0.40SE), soil pH (1.81±0.27SE), and reduced soil C:N ratios (1.27±0.38SE) in ECM-dominated stands. In contrast, the effects of invasion were near zero (ranging 0.27±0.14 to 0.59±0.09SE) in stands dominated by AM trees. Consistent with our initial hypothesis, we observed negative relationships between net nitrification (r2=0.43, p<0.0001), soil pH (r2=0.32, p<0.0001) and the relative abundance of ECM trees, and a positive relationship between soil C:N (r2=0.32, p<0.0001) and the relative abundance of ECM trees. However, in plots invaded by M.v., these relationships flattened, suggesting invasion may homogenize forests, reducing biogeochemical heterogeneity across the landscape. Collectively, our results illuminate how M.v. disproportionately alters biogeochemical conditions in environments where resident trees have the most dissimilar traits, and provides a conceptual framework for assessing invasive species impacts on ecosystem processes.