Land use is a key factor determining conditions in streams. In particular, streams impacted by surrounding agriculture receive increased nutrient loads as excess fertilizers drain from the agricultural land. Consequently, concentrations of nutrients, such as nitrogen, may be significantly higher in these streams than in mixed-use or forested streams. The microbial process of denitrification is one way to remove this nitrogen; therefore, knowledge of a stream’s denitrification rate is essential in determining the stream’s ability to reduce transport of nitrogen downstream where it may contribute to eutrophic conditions. This study examined how total bacterial abundance, denitrifier community abundance and diversity, denitrification rates, algal biomass, and benthic invertebrates vary longitudinally in agricultural and mixed-use streams, and how observed patterns correlate with physical and chemical factors, such as nutrient and DOC concentrations. Two headwater streams from intensive agricultural areas (Leary Weber ditch, Sugar Shirley, Indiana) and one with mixed land use (Sycamore Creek, Indiana) were sampled under base flow in summer and fall (2009). Total bacterial abundance was measured using DAPI staining.  Denitrification rates were determined using the chloramphenicol-amended acetylene-block technique. Denitrifier community composition was determined with quantitative PCR and terminal restriction fragment length polymorphism (TRFLP) analysis of the nosZ (nitrous oxide reductase) gene.

Results/Conclusions

Overall, intra-stream variation was lower than inter-stream. Denitrification rates varied greatly with surrounding land use and season, ranging from 28 to 14,575 µg N/m²/h.  Similar variation was found in chemical factors such as nitrate concentration (from 1.7 to 6.30 mg/L), and community factors such as bacterial abundance (from 1.60 x 10⁶ to 3.69 x 10⁷ cells/g DM). PCR confirmed nosZ presence at every site and TRFLP revealed 20 to 30 different peaks per site. Redundancy analysis of TRFLP profiles revealed that a significant portion of variation in denitrifier community structure was attributed to land use; 64% of variation was accounted for by inter-stream differences when comparing Sycamore Creek and Leary Weber. Agricultural impacts explained many patterns, such as higher concentrations of nitrate, total nitrogen and soluble reactive phosphorus, as well as significantly higher denitrification rates. Overall, denitrification rates were strongly correlated with concentrations of dissolved oxygen (R= 0.79), nitrate (R= 0.69), and total nitrogen (R= 0.57). We suggest that stream denitrification rate and denitrifier community composition are strongly linked to surrounding land use type, related to significantly increased nutrient loading at agriculturally impacted sites.