Denitrification represents a permanent pathway for removal of nitrate from streams.  Elevated nitrate concentrations in agricultural streams can support high denitrification rates, which often vary annually with seasonal changes in nitrate concentrations.  Changes in the dissolved organic matter (DOM) pool (quantity or quality) may provide a greater/lesser supply of electron donors with varying degrees of suitability for denitrifying bacteria, resulting in an altered rate of denitrification, either through changing the abundance of denitrifying bacteria and/or the efficiency of the denitrifying community.  Previous research suggests stream bacterial community composition responds to changes in DOM quality, which may shift the structure of the denitrifying community and impact denitrification rates.  Thus, two mechanisms can operate to determine variability in denitrification rates.  The first is based on a metabolic response to changing resources, whereas the second is based on an assemblage-level genetic shift in response to changing resources.  Our study was designed to investigate the relative importance of, and interactions between, these two mechanisms.  Research was conducted in an agricultural headwater stream in central Indiana, U.S.  Denitrification rates were measured using a combination of the chloramphenicol-amended acetylene-block technique and whole-stream ¹⁵N-nitrate releases.  Denitrifier community composition was analyzed using the nosZ gene (nitrous oxide reductase) via quantitative PCR and terminal restriction fragment length polymorphism.  

Results/Conclusions

Results using acetylene block denitrification assays suggest strong temporal variation in denitrification rates associated with variation in nitrate concentrations.  Denitrification rates in November 2008 were undetectable or very low (<0.18 mg N/m²/h) corresponding to low nitrate concentrations (0.02 mg N/L) whereas rates in January 2009 were much higher (6.3 mg N/m²/h ±1.7 SE) because of high nitrate concentrations (5.1 mg N/L).  Although DOM concentrations in January were 2.6 mg C/L, denitrification was significantly higher when bottles were amended with labile DOM (as glucose; two-way ANOVA, P=0.02) likely because only 20% of the ambient DOM was bioavailable.  Interestingly, PCR recovered nosZ in all sediment samples examined suggesting denitrifiers were present even when conditions for denitrification were not optimal.  Overall, our results indicate that denitrification rates were seasonally limited by the quantity of labile DOM and that denitrifiers were present even during periods of low denitrification.  We suggest a full understanding of the controls on the downstream export of nitrogen must consider the direct influences of DOM on denitrification rates and also the indirect influences of DOM as a mediator of denitrifier community composition.