Quantifying the relative contributions of diazotrophic community composition and soil characteristics on nitrogen fixation rates

Background/Question/Methods

In general, the role of microbial community composition as a quantitative and qualitative determinant of ecological processes remains actively debated. As might be expected, the relationship between diazotrophic community structure and N-fixation rates in soil remain largely uncharacterized. In order to evaluate the relationship between diazotrophic community structure and N-fixation rates in soil, we used cultivation independent techniques nifH terminal restriction fragment length polymorphisms (TRFLP) analysis characterizing soil diazotrophic community composition in relation to soil N-fixation measurements so we could detect both cultivated and uncultivated diazotrophs in soil. The nifH gene is the most common biomarker used for cultivation-independent analysis of diazotrophic communities. TRFLP is a molecular fingerprinting technique which is commonly used for studying relationships between microbial community composition and environmental factors. A long term agricultural experiment site in Chazy, Clinton County, New York was used for this study. We examined variation in nifH TRFLP, soil characteristics, and N-fixation rates with respect to tillage and biomass management practices at the site at 6 times over a period of 2 years. A general linear model (GLM) was used to evaluate the degree to which variation in community composition, soil characteristics, treatments and time points were related to rates of N-fixation in soil. The AMMI model was used to visualize the interaction between the environmental factors and the diazotrophic community.

Results/Conclusions

Time was the largest source of variation in the diazotroph community but treatments also influenced community composition. In the first GLM model (p <0.0001), time was the most important factor in the model explaining 24.64% of the variation in N-fixation rates (p < 0.0001). Variation in the diazotrophic community structure explained 23.4% of the variation in N-fixation rates (p = 0.0103). Soil characteristics and experimental treatment were each unable to explain significant variation in N-fixation rates in this model. A 2^nd GLM model (p <0.0001), which excluded treatments and time as factors, indicated that variation in diazotroph community composition explained 30.3% of the variation in N-fixation rates (p < 0.0001), and was the most important independent variable in this model. Variation in soil characteristics explained only 12.34% of the variation in N-fixation rates (p = 0.005). These results suggest that changes that occur in the diazotrophic community over time are major drivers of N-fixation rates. The diazotrophic community composition is in turn sensitive to treatment effects leading to interactions between treatment and time, which could influence N-fixation rates.