Moss-associated nitrogen (N) fixation is the largest source of new N to nutrient-limited high latitude ecosystems. Current methods of N fixation rate measurement can be indirect or expensive, yet it is critical to produce accurate regional scale predictions of N inputs to these rapidly changing ecosystems. The natural abundance of stable isotopes, particularly 15N, have been used as a proxy to indicate N fixation activity. Mosses utilizing N obtained via microbial fixation would be expected to have a δ15N signature near 0‰. In mosses, which lack stomata, δ13C enrichment increases along with moisture due to diffusion limitation created by water films. We measured both the N fixation rate and the natural abundance of stable isotopes associated with 11 moss genera across boreal and Arctic Alaska to test the hypotheses that (1) N fixation rates would increase with δ15N, approaching zero with higher N fixation rates and (2) that N fixation would increase with δ13C, as consistently high moisture would promote N fixation. Each of the 505 samples from 24 sites was split in half, with one half incubated for 24h with 15N2 gas and the other immediately dried for stable isotope analysis.
Results/Conclusions
Contrary to our first hypothesis, we saw no relationship between N fixation and δ15N, even after accounting for taxonomic differences between mosses (R2=0.01). The reasons for this may be that mosses are not utilizing N fixed by their associated microbes or that microbes were using alternative forms of nitrogenase that discriminate against 15N, thus producing a different natural abundance signature than expected from N fixation. Our second hypothesis was supported by a positive relationship between N fixation rate and δ13C for all mosses other than Sphagnum spp. (Model R2=0.41). A strong interaction (p<0.001) between moss genus and δ13C was driven both by significant differences in N fixation rates among mosses and by different levels of sensitivity to moisture which perhaps were driven by moss anatomy and community structure. We conclude that δ15N is not a reliable indicator of moss-associated N fixation rates at high latitudes, but that δ13C is related to N fixation rates.