Nitrogen (N) deposition may homogenize plant communities and decrease beta-diversity, but how fine-scale landscape heterogeneity may control these changes is less well-understood. In California, urbanization and intensive agriculture expose adjacent oak savannas to enhanced N deposition. Within the oak savanna, isolated tree canopies are hotspots of N deposition. Does N reduce beta-diversity in the plant community a) with increasing fertilizer level b) between understory and open areas or c) across years of study? Fertilizer was applied in the 2011 and 2012 growing seasons to 80 two-meter-square plots in understory and open areas at four treatment levels, ranging from 5 to 200 kg/ha/yr, simulating a gradient in deposition rates across the region. The experiment was done during an intense drought. Percent cover of each species was measured in plots during the spring of 2011, 2012, and 2015. Using linear mixed effects modeling, we analyzed changes to the community based on N fertilizer, position in the landscape (understory versus open), and year. Partial Mantel tests were performed with a Bray-Curtis community dissimilarity matrix and environmental/spatial data to examine mechanisms controlling diversity across the landscape. We then performed a beta dispersion analysis to determine whether the community became more similar among treatments.
Results/Conclusions
Nitrogen fertilizer had a limited effect on the plant community. Fertilizer slightly increased productivity and soil N availability in the open relative to the understory, but this effect was significant only in the first year of the study. Partial Mantel tests showed community dissimilarity was significantly correlated with environmental differences between canopy and open areas, particularly gravimetric water content, reflecting the importance of the drought during the study. No changes in beta-diversity were associated with N fertilizer or understory versus open landscape positions. However, there was a significant interaction with year, in which beta-diversity was greater in open areas during Year 2, which corresponded to the first year of the drought. We did not find evidence that N deposition caused homogenization in the plant community. Rather, the plant community was resistant to N deposition and drought was the important driver. The impact of water stress across study years appeared to interact with the canopy to affect diversity patterns in the landscape. Specifically, while the open grassland exhibited compositional shifts, the understory community remained stable throughout a long and intense drought. Isolated savanna trees may help to maintain community structure in the face of environmental change.