Quantifying phosphorus uptake using pulse and steady-state approaches in streams

Steady-state approaches to the study of stream nutrient processing have several limitations. Dynamic (time series) approaches are more flexible, and allow interpretation of nutrient additions introduced as unsteady slugs (pulses). We compared soluble reactive phosphorus (SRP) uptake metrics from experimental nutrient pulses modeled dynamically with those from continuous injections modeled with a steady-state approach. For six southern Wisconsin streams, uptake metrics from these two methods were similar despite low nutrient demand. Linear regression of paired-pulse versus steady-state estimates of the first-order uptake coefficient (λ; r2 = 0.84, slope = 1.21) and uptake velocity (vf; r 2 = 0.95, slope = 1.03) were highly significant. There was a tendency for slightly higher uptake with pulses, possibly due to P sorption. Sampling across five stations of one stream yielded a similar longitudinal pattern between experimental pulse SRP flux and steady-state (plateau) SRP concentration. Conservative transport parameters for pulse and continuously injected tracer data were also similar. These results suggest that unsteady nutrient amendments can provide usable nutrient uptake values, even in low-uptake situations for which uncertainty is high. The flexibility of dynamic approaches to nutrient spiraling facilitates research in poorly understood situations, including conditions of high water residence time, high discharge, and changing discharge or background chemistry. *Corresponding author: E-mail: [email protected] Acknowledgments We thank contributors to the River Ecology lab at the University of Wisconsin Center for Limnology, including manager Mark Lochner, Alex Bilgri, Nathan Braun, David Mark, and James Thoyre. Matt Diebel provided help with mapping. Special thanks to Richard Cates, Deborah Frosch of the Riverland Conservancy, and the Aldo Leopold Foundation for granting access to three study streams. This manuscript was greatly improved by comments and advice from Robert Runkel and Michael Gooseff. Support was provided by the National Research Initiative of the US Department of Agriculture (USDA) Cooperative State Research, Education, and Extension Service (CSREES grant 2004-35102-14793). Limnol. Oceanogr.: Methods 7, 2009, 498–508 © 2009, by the American Society of Limnology and Oceanography, Inc. LIMNOLOGY and OCEANOGRAPHY: METHODS Powers et al. Nutrient pulses and plateaus in streams 499 places and times at which nutrient removal mechanisms are meaningful to long-term dynamics, and would be valuable within the context of continuing global change. Therefore, methods and models are needed that expand the scope of nutrient spiraling to larger and more complex aquatic ecosystems. Dynamic models of lotic ecosystems are more flexible than steady-state models, and applicable to situations outside the scope of the steady-state approach to nutrient spiraling. For example, dynamic models allow interpretation of uptake time series from unsteady nutrient amendments (additions that do not achieve a stable experimental nutrient concentration), such as abbreviated continuous injections and instantaneous spikes/slugs (pulses). Although pulse methods present advantages in terms of scope and speed, it remains to be seen how nutrient uptake metrics derived from pulses relate to those derived from steady-state approaches. In this work, we used a dynamic transport model to quantify soluble reactive phosphorus (SRP) uptake velocity from phosphorus (P) pulses for multiple systems, and compared these values to those from paired continuous injections modeled with a steady-state approach. Our interest in this comparison was to evaluate pulse versus steady-state nutrient uptake metrics in situations where both were quantifiable and to facilitate the interpretation of experiments that necessitate the unsteady introduction of nutrients. The steady-state nutrient uptake approach—The quantification of stream nutrient dynamics using nutrient amendment experiments and steady-state modeling is supported by a rich methodological literature (e.g., Newbold et al. 1981, Stream Solute Workshop 1990, Payn et al. 2005). Although the general approach has been cited as laborious (Fisher et al. 2004), strengths relate to its methodological and mathematical elegance and simplicity. These characteristics have, fortuitously, facilitated widespread application of the model (Ensign and Doyle 2006). Conventional applications of the nutrient spiraling model involve the experimental introduction of ecologically relevant elements by continuous injection at the upstream end of a study reach. Nutrient uptake metrics for individual elements are then calculated from longitudinal declines in solute concentration along the reach during the enriched, stabilized period in the time series (the “plateau”). In its simplest form, the steady-state nutrient uptake approach yields the nutrient uptake length (Sw) defined by