Root endophyte communities are affected by nitrogen differentially in two co-dominant alpine tundra plants from Niwot, LTER

Background/Question/Methods

Due to pollution from the cities of Boulder and Denver, CO, Nitrogen (N) deposition has been increasing at Niwot Ridge, LTER. Over the last decade, one of the co-dominant alpine tundra plants, Geum rossii, has experienced substantial dieback, while the other co-dominant, Deschampsia caespitosa, gains dominance. G. rossii dieback has been attributed to elevated soil N rather than inter-species competition, as it dies in N addition plot even when D. caespitosa is excluded. The mechanism by which N kills G. rossii is still unknown. We hypothesize elevated soil N causes shifts in endophytic fungal communities, and that these shifts are responsible for G. rossii death.  We expect G. rossii communities to become more pathogenic and parasitic in response to N. G. rossii and D. caespitosa roots were sampled from N-addition and control plots. Samples were surface sterilized prior to DNA extraction and 454 titanium pyrosequencing of fungal ITS ribosomal DNA. Sequences were quality filtered, parsed into OTUs (97% similarity), assigned taxonomy and rarified using Qiime. The effects of N treatment on beta diversity were assessed by RDA, and on OTU abundance by ANOVA using the vegan and nlme packages in R, respectively.

Results/Conclusions

In G. rossii roots, endophyte communities were dominated by fungi in the order Helotiales (73%). G. rossii endophyte communities did not cluster by N treatment at any taxonomic level except species. Twenty OTUs correlated with N, but shifts in relative abundance occurred among closely related species. Most sensitive taxa responded to N positively, including taxa that belong to pathogenic groups. D. caespitosa roots were dominated by Helotiales (29%) in control plots, but by Xylaria (34%) and Rhytismitales (27%) in N plots. D. caespitosa communities clustered by N at all taxonomic levels. Similarly, though only three D. caespitosa OTU’s were correlated with N, shifts in abundance occurred between distantly related species, so that whole fungal Classes responded to N. Most affected taxa responded negatively. These data indicate that endophyte communities of D. caespitosa and G. rossii respond substantially differently to N addition. D. caespitosa endophyte communities appear to change more fundamentally. These changes may be important to D. caespitosa’s ability to adapt to changing nutrients. If so, these data may provide insight into the mechanisms by which abiotic soil factors alter above ground vegetation dynamics.