Reconciling potential nitrification rates and site characteristics across the landscape

Background/Question/Methods

Intrinsic soil properties, as well as the microbial community therein, have been identified as potential mechanisms driving nitrification rates across landscapes. Recent literature indicates that a range of local soil properties including moisture, temperature, texture and pH drive differences in nitrogen (N) export among sites. While edaphic characteristics may partially describe landscape N patterns, there is also evidence for a relationship between soil carbon content and nitrogen transformations. In this study, we examined an extensive set of site characteristics across sub-watersheds within the Coweeta basin, and sought to identify which mechanisms drive nitrification patterns. The Coweeta Long-Term Ecological Research (LTER) site in North Carolina is a 2,185 hectares watershed basin ranging in elevation from 679 to 1,592 meters. Soils are classified as immature Inceptisols or developed Ultisols, but physical and chemical properties vary throughout the basin. We established high- and low-elevation transects within 10 sub-watersheds of differing aspect, elevation and treatment history. Along a 50 m transect, we sampled riparian, mid-slope and upslope soils in conjunction with potential nitrification rates.

Results/Conclusions

We reduced the dataset to four main fixed effects, and used linear models to evaluate those effects on potential nitrification. Model selection was determined using Akaike’s information criterion (AIC) values. An ANOVA was run using the reduced model, indicating a strong overall interaction between N mineralization and treatment (p < 0.0001).  Site selection for the project included paired N-S watersheds of similar elevation and treatment (e.g. planted white pine). While previous research has demonstrated that land conversion affects N mineralization and nitrification rates, all treatments within the basin are at least 30 years old. However, legacies of this land-management history are apparent in contemporary nitrification rates. There were significant interactions between soil moisture, N mineralization and treatment (p < 0.001) and soil temperature, N mineralization and treatment (p < 0.01). Using the output from the linear model, it appears that decreasing soil moisture in a previously disturbed watershed helps explain lower net potential nitrification and N mineralization. In addition, in a high elevation, previously thinned watershed, soil temperature is tied to higher net potential nitrification and N mineralization. Collectively, we find evidence that historical operations and N mineralization strongly explain variation in nitrification across the landscape.