Predicting stream nitrogen fluxes with patterns of watershed soil moisture

Background/Question/Methods

Precipitation events are becoming less frequent, but more intense. These shifts in precipitation patterns will likely result in changes to soil moisture, increasing the occurrence of drying and wetting cycles. Drying and rewetting cycles are known to cause sudden bursts of nitrogen mineralization, increasing the amount of soil nitrogen that is susceptible to leaching. Increases in leaching potential at small scales have the potential to increase nitrogen export at the watershed scale. Some evidence for this exists — monthly precipitation variability tends to decrease nitrogen retention when compared across North American sites. However, this evidence confounds ecosystem type with precipitation variability. How the N export of different ecosystems responds to shifts in precipitation patterns within an ecosystem is unknown. The goal of this research was to determine if there are specific frequencies of precipitation and drought, or soil drying and rewetting that result in the largest fluxes of N, and how the frequencies depend on ecosystem type. I tested for the effect of precipitation frequency on stream N fluxes. I used precipitation, temperature and stream flow data to model soil moisture across a number of experimental forests and LTER sites). I then determined the relationship between the local maxima of the wavelet transform of soil moisture and stream N fluxes. Additionally, I tested the ability of the slope of the local maxima chain, which represents the correlation structure of the soil moisture data to predict the intensity of stream N fluxes. This work was supported by NSF Grant No. DGE-0504195.

Results/Conclusions

I found that the local maxima found at daily frequencies of the wavelet transformed soil moisture do not correlate strongly with the largest total dissolved N fluxes at sites located at Andrews Experimental Forest or the Baltimore Ecosystem Study LTER. Local maxima at lower frequencies, such as those associated with synoptic weather patterns, seasonality, and multi-year trends may be more informative. Increased N flux out of watersheds at the global scale could have large negative effects on eutrophication of streams, rivers and coastal areas, contributing to dead zones and reducing biodiversity. Predicting the effects of rainfall on N fluxes could help manage watersheds, streams and coastal regions.