Assessing the importance of seepage and springs to nitrate flux in a stream network in the Wisconsin sand plains

Abstract

Evaluating the flow paths that contribute to solute flux in stream networks can lead to greater understanding of the linkages between biogeochemistry and hydrology. We compared the contributions of groundwater in spring brooks and in seepage through the streambed to nitrate flux in the Emmons Creek network in the Wisconsin sand plains. We predicted that spring brooks would contribute disproportionately to nitrate flux due to the presumed higher advection rates in springs and less opportunity for nitrate removal relative to seeps. Nitrate flux was measured in 15 spring brooks that entered Emmons Creek. Nitrate flux from seepage was measured at the locations of 30 piezometers, based on Darcy's Law, and by a reach-scale injection of Rhodamine water tracing (RWT). When seepage discharge was estimated from the RWT release, groundwater inputs from seepage and springs accounted for the discharge gain in the Emmons Creek channel. Springs brooks and seepage (based on the RWT release) contributed 37% and 63%, respectively, to nitrate flux inputs in the study reach. Contrary to our prediction, seeps contributed disproportionately to nitrate flux relative to their discharge. Relatively high rates of seepage discharge and higher than anticipated nitrate concentrations in the shallow pore water at seepage locations contributed to the unanticipated result.

Groundwater contributions of flow and nitrogen in a headwater agricultural watershed

“…Riparian seeps supply 45–217 m 3 /d of water and 0.6–4.2 kg/d of N. Comparatively, 0.8–13.2 m 3 /d of water and 0.02–0.25 kg/d of N are lost to groundwater. These results differ from work by Stelzer et al (2017) working in the sand plains of Wisconsin, where 37% of streamflow was derived from preferential flow in spring brooks (which were defined as outflows from riparian springs, similar to our definition of seeps), and 63% was made up of streambed groundwater matrix flow. Despite being in an agricultural watershed, nitrate concentrations were considerably lower in their study stream, approximately 2.5 mg N L −1 , and the stream reach was approximately 10 times longer, 1450 m. Differences in the ecoregion and underlying geology of the two sites also would affect groundwater flow paths, as the stream studied by Stelzer et al (2017) was underlain by an unconfined aquifer of well‐sorted sands whereas the site studied here is situated on fractured bedrock.…”

“…In contrast, the highly fractured nature of the shale bedrock underlying RS can facilitate larger amounts of preferential flow, while the low hydraulic conductivities of the streambed generally hinder matrix flow. Hydraulic conductivities reported by Stelzer et al (2017) ranged from 1.7 to 8.64 m/d, while this study found hydraulic conductivities ranging from 0.0009 to 8.27 m/d, with 7 of 14 measurements <1 m/d.…”

“…Stream seeps can often be the source of a large portion of flow delivered to streamflow and can have important implications for stream water chemistry (O'Driscoll & DeWalle, 2010). A growing body of literature exists on the role of seeps in groundwater‐based N inputs in streams (Burns, Murdoch, Lawrence, & Michel, 1998; O'Driscoll & DeWalle, 2010; Stelzer, Strauss, & Coulibaly, 2017; Williams, Buda, Elliott, Singha, & Hamlett, 2015), but few studies have explicitly partitioned components of matrix and preferential groundwater flow in fractured aquifers. In fractured aquifers, hydraulic conductivities are often elevated, allowing for quick, lateral movements of water with low residence times, which facilitates N‐rich waters to be transported to surface waters with little change to its chemistry (Gburek & Urban, 1990).…”

Macrophyte‐driven transient storage and phosphorus uptake in a western Wisconsin stream

“…Because M b accounted for most of the variation in F med 200 in the final predictive model, the primary mechanism of transient storage in Spring Coulee Creek was surface storage presumably caused by macrophyte beds. Others have likewise concluded that sand‐bed streams have very little hyporheic storage (Harvey & Wagner, ; Jin & Ward, ; Stelzer, Strauss, & Coulibaly, ; Stofleth et al, ). Values of A s / A in Spring Coulee Creek (0.15 ± 0.07) were similar to what Powers et al () found in other Wisconsin streams and were within range of the mean for sand‐bed streams (0.36 ± 0.22, as reviewed by Stofleth et al, ).…”

Determining the spatial distribution of percolation for winter wheat based on system dynamics and semi‐distributed modelling*

“…Percolation is the seepage of soil water out of the root zone and usually regarded as the water flux moving downward from the bottom of the root zone (Hare et al, 2017; Stelzer et al, 2017). Percolation is an important link in the water cycle (Xie et al, 2015) and water balance processes, especially for groundwater resources (Min et al, 2018).…”