Several hardwood-dominated watersheds, including the long-term reference watershed (WS4), at Fernow Experimental Forest (FEF), West Virginia, have become N saturated from chronically-elevated N deposition. WS4 also displays a considerable degree spatial heterogeneity in soil N dynamics. Long-term data for in situ (buried bag) net nitrification in mineral soil reveal a gradient of three sites within WS4 varying in rates of nitrate production: LN, MN, and HN (low, medium, and high nitrification rates, respectively). Field-based gradients of net nitrification were essentially identical to those found under the controlled conditions in the laboratory, suggesting that field-based variability has arisen from variation in soil microbial communities, rather than from transient ambient factors, e.g., moisture and temperature. The best predictor of spatial pattern of net nitrification in WS4 is clay content, suggesting weathering as a potentially ultimate control on soil N dynamics. This study characterized soil microbial communities along this weathering gradient and assessed factors important in explaining microbial composition. In summer 2009, we sampled mineral soil from each of the three sites to a depth of 5 cm, and analyzed these soils for pH, organic matter, extractable N, and net nitrification potential. We assessed microbial communities in these samples with PLFA analysis.

Results/Conclusions

Consistent with field gradient patterns, HN soil had the highest nitrification potential, LN soil exhibited no net nitrification, and MN was intermediate. Canonical correspondence analysis indicated that microbial community composition varied along the weathering gradient. We found the predominance of fungal groups (18:2n6 and 18:1n9c) at the most weathered LN site and gram negative bacteria (18:1n7c—which generally include nitrifier species) at the least weathered HN site. Indeed, one of the more important vectors in CCA was fungi:bacteria ratio, which increased in the direction of LN plots in ordination space. Other factors important in explaining variation in microbial community composition were pH/net nitrification (both increasing with HN) and organic matter/extractable ammonium (both increasing with LN). Interestingly, the degree of spatial heterogeneity of the microbial community, as indicated by the area of site triangles in ordination space, varied with N availability in ways that are similar to that predicted by the N homogeneity hypothesis (which predicts that community diversity declines in N-impacted terrestrial ecosystems because of reductions in spatial heterogeneity of available N across the landscape). LN sites displayed the greatest variability in soil microbial community composition, whereas HN had the lowest.