Transport of N and P in U.S. streams and rivers differs with land use and between dissolved and particulate forms

Abstract

We used a recently published, open‐access data set of U.S. streamwater nitrogen (N) and phosphorus (P) concentrations to test whether watershed land use differentially influences N and P concentrations, including the relative availability of dissolved and particulate nutrient fractions. We tested the hypothesis that N and P concentrations and molar ratios in streams and rivers of the United States reflect differing nutrient inputs from three dominant land‐use types (agricultural, urban and forested). We also tested for differences between dissolved inorganic nutrients and suspended particulate nutrient fractions to infer sources and potential processing mechanisms across spatial and temporal scales. Observed total N and P concentrations often exceeded reported thresholds for structural changes to benthic algae (58, 57% of reported values, respectively), macroinvertebrates (39% for TN and TP), and fish (41, 37%, respectively). The majority of dissolved N and P concentrations exceeded threshold concentrations known to stimulate benthic algal growth (85, 87%, respectively), and organic matter breakdown rates (94, 58%, respectively). Concentrations of both N and P, and total and dissolved N:P ratios, were higher in streams and rivers with more agricultural and urban than forested land cover. The pattern of elevated nutrient concentrations with agricultural and urban land use was weaker for particulate fractions. The % N contained in particles decreased slightly with higher agriculture and urbanization, whereas % P in particles was unrelated to land use. Particulate N:P was relatively constant (interquartile range = 2–7) and independent of variation in DIN:DIP (interquartile range = 22–152). Dissolved, but not particulate, N:P ratios were temporally variable. Constant particulate N:P across steep DIN:DIP gradients in both space and time suggests that the stoichiometry of particulates across U.S. watersheds is most likely controlled either by external or by physicochemical instream factors, rather than by biological processing within streams. Our findings suggest that most U.S. streams and rivers have concentrations of N and P exceeding those considered protective of ecological integrity, retain dissolved N less efficiently than P, which is retained proportionally more in particles, and thus transport and export high N:P streamwater to downstream ecosystems on a continental scale.

Water availability and seasonality shape elemental stoichiometry across space and time

“…These results indicate a decoupling of particulate versus dissolved stoichiometry; during high flows enhanced floodplain connectivity results in higher quality particulate fluxes (low C:N) and high DOC fluxes and the dilution of groundwater N, resulting in lower C:N soluble fluxes. Further, there is much greater variability in C:N of soluble fluxes versus particulate fluxes (SE = 15.27 versus SE = 0.86, soluble versus particulate, respectively, particulate fluxes from Atkinson et al, 2009) similar to findings by Shousha et al (2021) and Manning et al (2020). As particulate forms of C, N, and P are more likely to be retained, and exhibit greater retention times than dissolved solutes (Hall et al, 2009; Minshall et al, 2000), stoichiometric constancy of particulate nutrients could result in enhanced delivery of dissolved C and N to downstream ecosystems, and relatively stronger retention of particulate P. Further work should understand how the decoupling of stoichiometric fluxes of nutrients and basal resources impact stream food webs across different functional feeding groups and levels.…”

“…Material processing and transport to river networks impact a suite of ecological processes by altering absolute nutrient concentrations and the relative availability, or stoichiometric ratios, that strongly influence stream ecosystem productivity and food webs (Cross et al, 2006; Hall et al, 2005; Welti et al, 2017). However, few studies have considered the stoichiometric balance of watershed fluxes (but see Kincaid et al, 2020; Manning et al, 2020; McDowell et al, 2019), especially over longer time scales. Ecological stoichiometry traditionally focuses on the ratio of carbon to nitrogen to phosphorus (C:N:P; Elser & Hamilton, 2007; Frost et al, 2006), frequently comparing the availability of nutrients to the Redfield ratio (106 C:16 N:1 P; Redfield, 1958).…”

Addressing conservation practice limitations and trade‐offs for reducing phosphorus loss from agricultural fields

“…Trade‐offs arising from CP implementation, such as the form of P (sediment‐bound vs. soluble P) and mode of delivery (surface vs. subsurface, acute versus chronic), can affect downstream water quality and ecological status (Manning et al., 2020). In the western Lake Erie basin, for example, a convergence of conservation tillage, broadcast fertilizer application, and tile drainage resulted in an increased proportion of highly bioavailable soluble P fractions transmitted via subsurface preferential flow pathways during storm runoff (Jarvie et al., 2017; Meals et al., 2012).…”

Different forms of carbon, nitrogen, and phosphorus influence ecosystem stoichiometry in a north temperate river across seasons and land uses

“…For example, concentrations of highly bioavailable ammonium (NH 4 + ) and soluble reactive phosphorus (SRP) have been found to increase from 10-100 times and 50 times, respectively, downstream of WWTPs in US rivers (Haggard et al 2005), and overall loads are higher with larger human populations (Withers and Jarvie 2008;Carey and Migliaccio 2009). The dominant N form entering rivers from agricultural nonpoint sources is typically NO 3 À , which is highly mobile in the soil matrix (Caraco and Cole 1999;Stanley and Maxted 2008), whereas for P, high dissolved concentrations are typically observed, but precipitation promotes increased particulate mobilization as a function of soil erosion (Vanni et al 2001;Goyette et al 2019;Manning et al 2020). In terms of C, there is little consensus on the net effect of human activity on riverine concentrations (Stanley et al 2012;Xenopoulos et al 2021).…”