Assimilatory uptake rather than nitrification and denitrification determines nitrogen removal patterns in streams of varying land use

Agricultural and urban land use increase nitrogen (N) concentrations in streams, which can saturate biotic demand by plants, algae, and bacteria via assimilative uptake, and by nitrification and denitrification. We studied six streams per year in each of three land-use categories (agricultural, urban, and forested) for 3 yr (n 5 18 streams), and we compared whole-stream N uptake and microbial N transformation rates during spring, summer, and autumn. We measured whole-stream removal of added ammonium (NH z4 ) and nitrate (NO { 3 ) in the field and quantified sediment nitrification and denitrification rates in laboratory assays. Relative demand for NH z4 (as uptake velocity, Vf) was highest in spring and in streams with open canopies, implying a link with aquatic autotrophy. In agricultural and urban streams, whole-stream removal (as areal uptake, U) of NH z4 and NO { 3 , nitrification, and denitrification rates approached saturation at higher inorganic N concentrations. Nitrification and denitrification rates measured in redox-optimized laboratory assays were roughly equivalent, suggesting that in situ redox conditions will determine whether stream sediments are a net source or sink of NO 3 . Though nitrification and denitrification rates were measured under ideal redox conditions, they were always more than an order of magnitude lower than whole-stream NO 3 uptake, demonstrating their limited influence on wholestream NO 3 dynamics. Assimilatory processes, which temporarily store N removed from the water column, dominated whole-stream N demand and controlled downstream N flux. The ultimate fate of assimilated N remains unknown; in-channel storage cannot account for it, and thus a key question is what fraction may eventually be stored in downstream depositional zones or denitrified upon remineralization. Biological uptake and transformation of nutrients in headwater streams can regulate nutrient export to downstream ecosystems (Peterson et al. 2001), and this biological activity can be measured at the whole-stream level using nutrient spiraling techniques (Newbold et al. 1981; Stream Solute Workshop 1990). Nutrient spiraling couples nutrient uptake and release with downstream transport, and most studies of stream nutrient spiraling have been performed either in forested biomes (e.g., Tank et al. 2000; Ashkenas et al. 2004) or in stream systems minimally altered by human activities (e.g., Grimm and Fisher 1986; Dodds et al. 2000). Only recently have nutrient-spiraling studies investigated streams dominated by agricultural (Niyogi et al. 2004; Bernot et al. 2006) or urban (Grimm et al. 2005; Meyer et al. 2005) land uses, even though human-induced changes in land use influence most running waters in the United States (Meyer and Turner 1994). Agricultural and urban land uses typically increase dissolved inorganic nitrogen (DIN) concentrations in streams (Carpenter et al. 1998) and can also reduce riparian vegetation and increase light availability for aquatic autotrophs (Allan 2004). Therefore, knowledge of how land use mediates nutrient uptake and transformation in stream ecosystems is critical for understanding how streams regulate nutrient flux to downstream water bodies. Although land use can influence stream nutrient cycling by affecting DIN concentrations, seasonality is also an important driver of whole-stream nutrient uptake in midlatitude regions. For example, in temperate, forested streams, nutrient demand shows both a spring peak, when autotrophic activity increases concurrently with increasing light levels and temperature prior to deciduous forest leafout, and an autumn peak, when heterotrophs actively colonize and decompose allochthonous leaf litter inputs (Mulholland 1992, 2004). These seasonal peaks in nutrient demand have been confirmed in studies that examine nutrient uptake over an annual cycle (Simon et al. 2005; Hoellein et al. 2007; Roberts and Mulholland 2007), so spring and autumn may be considered biologically important periods for whole-stream nutrient demand in temperate streams. However, seasonality in nutrient uptake has rarely been studied in the context of streams that are 1 Present address: Department of Geography and Land Studies, Central Washington University, Ellensburg, Washington 98926. Acknowledgments We thank the many private landowners who granted us access to their property. The Drain Commissioners of Allegan, Barry, Calhoun, Eaton, and Kalamazoo Counties, the City of Kalamazoo, Allegan County Parks Department, and Fort Custer Training Center helped us coordinate site access. Jake Beaulieu, Denise Bruesewitz, Kathryn Docherty, Sally Entrekin, Natalie Griffiths, Jon Loftus, and Kris Premier provided assistance in the field and Rachel Clavers and Carrie DePalma provided assistance in the laboratory. Two anonymous reviewers provided comments that greatly improved this manuscript. C.P.A. was funded by the Arthur J. Schmidt Presidential Fellowship and Bayer Predoctoral Fellowship. Additional funding was provided by NSF-DEB 0111410. Limnol. Oceanogr., 53(6), 2008, 2558–2572 E 2008, by the American Society of Limnology and Oceanography, Inc.