Ecosystem metabolism controls nitrogen uptake in streams in Grand Teton National Park, Wyoming

Streams and rivers regulate nitrogen transport (N) to downstream ecosystems. Rates of N uptake can be high in streams, but controls on the variation in uptake rates of N among streams are not known. We measured ammonium (NH ) and nitrate (NO ) uptake velocities (Vf) and compared these with whole-reach estimates of gross primary 1 2 4 3 production (GPP) and community respiration (CR) in 11 low-nitrogen streams in Grand Teton National Park, Wyoming. We predicted that increased metabolism would positively relate to higher N demand because of stoichiometric N requirements associated with carbon fixation. Rates of GPP and CR explained 82% of variation in NH Vf. Nitrate Vf was controlled by GPP, not CR, with GPP explaining 75% of variation in NO Vf. Nitrate 1 2 4 3 concentrations did not increase downstream during NH addition in all streams, including streams with zero NO 1 2 4 3 uptake, suggesting low nitrification rates relative to NH uptake. Using a stoichiometric model, we show that areal 4 N uptake estimated from microbial and algal production was similar to measured areal N uptake. High primary production could be a prerequisite for streams exhibiting high NO uptake rates. 3 Streams and rivers are important avenues for N transport, yet they can also remove and transform dissolved N (Burns 1998; Peterson et al. 2001) and retain N in the terrestrial landscape (Howarth et al. 1996; Alexander et al. 2000). Hence, N cycling in streams has received much recent attention (Alexander et al. 2000; Sabater et al. 2000; Peterson et al. 2001). Ecologists have described nutrient dynamics in streams with the concept of spiraling (or uptake) length, which is the average distance traveled by dissolved nutrient before biotic uptake (Newbold et al. 1981). Using uptake length, areal uptake rates of N can be calculated from the water column to benthic biota (Newbold et al. 1981). Peterson et al. (2001) used a nutrient spiraling approach and found that headwater streams can remove and transform dissolved N inputs rapidly, presumably because of high rates of biological activity combined with high sediment–water contact time. Dissolved nitrogen uptake rates from the water column are especially high in shallow streams, thereby de1 Corresponding author ([email protected]). Acknowledgments We thank Mike Marshall, Kelly Gordon, Jamie Schaller, and Celine Louwers for help with field and laboratory work. Robert Schiller facilitated field studies in Grand Teton National park (GTNP) and Hank Harlow provided logistical support at the AMK Ranch. Patricia Colberg donated use of ion and gas chromatographs for anion and SF6 analyses. Mike Marshall, Jules Feck, Brad Taylor, and three anonymous reviewers commented on early drafts of the manuscript. Nate Nibbelink drew the map of GTNP. This research was supported by the University of Illinois Campus Research Board; the Department of Natural Resources and Environmental Sciences, University of Illinois; the University of Wyoming College of Arts and Sciences; and the University of Wyoming/National Park Service Research Station. This paper is a contribution to the University of Wyoming/National Park Service Research Station. creasing N loading to downstream ecosystems (Alexander et al. 2000). Although N uptake velocities and areal uptake rates can be high in streams (Tank et al. 2000; Peterson et al. 2001), no studies have quantitatively addressed what factors control variation in nitrogen uptake beyond geomorphic features of streams such as water velocity or depth (Alexander et al. 2000; Wollheim et al. 2001) or measures of transient storage (Valett et al. 1996; Hall et al. 2002). We expect that nutrient uptake by streams is high when net production is high (Grimm 1987). For example, a relationship between biological processes and nutrient uptake was implied when increased light from riparian vegetation removal was linked to higher ammonium (NH ) uptake in a stream in Spain (Sa4 bater et al. 2000). Physical attributes of streams can only indirectly control N uptake; it is the biotic demand by algae and microbes in attached biofilms that will ultimately determine N uptake and transformation in a stream. Given that many freshwater ecosystems are N limited (Elser et al. 1990; Francoeur 2001), we predict that variation in biological demand for dissolved N will explain most of the variation in stream uptake of N in low-nutrient streams. Streams with high autotrophic and heterotrophic production should remove more dissolved N because of simple stoichiometry: increased C fixation or heterotrophic C uptake will increase demand for N. Highly productive streams should transform N at higher rates, and are thereby more likely to control form and timing of watershed export of N. To test this hypothesis, we quantified NH and nitrate 4 (NO ) uptake in 11 low-nutrient, N-limited streams in north3 west Wyoming and related N uptake to whole-reach estimates of community respiration (CR) and gross primary production (GPP). Both N uptake and metabolism (GPP and CR) were measured at the segment scale (cf. Frissell et al. 1986), thereby integrating spatial variation over a 100–4001121 Stream metabolism and nitrogen uptake Fig. 1. The 11 study streams within Grand Teton National Park, Wyoming. MW, Moose-Wilson Creek; PB, Paintbrush Canyon tributary; NM, North Moran Bay Creek; GT, Glade Creek Tributary; LC, Lizard Creek; BC, Bailey Creek; PC1 and PC2, Pilgrim Creek channels 1 and 2; TO, Two Ocean lake outlet; SC, Spread Creek; DC, Ditch Creek. m length of stream. We define N uptake in this paper as gross removal of dissolved inorganic N from the water column, which is one component of overall retention (Newbold et al. 1981; Mulholland et al. 1997; Peterson et al. 2001). We express N uptake in three ways. First, we measured uptake length—the distance that a molecule travels before removal from the water column. However, this measure is strongly dependent on scale and is influenced by variance in stream discharge (Davis and Minshall 1999; Hall et al. 2002). We correct for the effect of stream size by calculating an uptake velocity that represents the biotic demand for a nutrient relative to its concentration in the water column. Finally, uptake velocity multiplied by stream water N concentration gives an areal N uptake rate from the water column to the benthos expressed as mass N per unit area per unit of time. Because stream size varied across our 11 study streams, we focus on uptake velocity and the areal N uptake rates to compare uptake among streams.