USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2007–5239

Back to Table of Contents

Results

The locations of the 10 ground-water transect sites are shown in figure 1. Well locations and screen depths are listed in table 1. Geochemical data are listed in table 2.

Transect Configurations

Transects generally consisted of a well installed in the center of the river, and one or more wells toward each edge of the river (lateral wells) (fig. 3). Lateral wells generally were installed near the river edge in coarse sediments and not in sloughed bank material. These wells tapped sediments likely to have a relatively direct connection to aquifer sediments, rather than collapsed-bank sediments that may behave more as a barrier than a conduit for aquifer discharge to the river. Most wells were screened at 1.0 ft below the riverbed. Transect configurations that deviated from these guidelines are discussed below.

Ground-Water Transects on Deschutes River

Site D194.1 consisted of two wells (table 1). River depth prevented installation of a well in the center.

Site D198.5 consisted of four wells: one in the center, one at each bank, and one toward the left edge of water. In attempting to install a well near (but not at) the left edge of water to complement the well at the left edge, multiple attempts at increasingly greater distances away from the left edge failed to yield a well that would produce water, until a well installed 18.8 ft from the left edge produced water. Sediment on the riverbed at this point was silt with macrophytes—conditions that generally were avoided but were accepted in this instance due to difficulty in finding a suitable location closer to the left edge of water. Sub-riverbed sediments between this well and the left edge of water probably are heterogeneous and have little correlation to lithologies on the riverbed itself.

Ground-Water Transects on Little Deschutes River

Site LD16.1 consisted of five wells. One nest of two wells (depths of 1.0 and 2.0 ft) was installed in the center of the river, another nest (depths of 1.0 and 2.0 ft) toward the left edge, and a single well toward the right edge. The shallow (1.0 ft) ground water in the center and near the left edge contained oxic, dilute ground water. Values of specific conductance (related to dissolved-ion content) were 53 and 58 μS/cm at 25°C, respectively (table 2), similar to the value of 54 μS/cm at 25°C measured in the river water during well installation. The presence of oxic ground water with specific conductance values similar to the specific conductance of river water suggested the presence of hyporheic water. The hyporheic zone is a place of interaction between river and ground water, commonly occurring in near-river sediments. Physical processes occurring at small spatial scales can promote hyporheic exchanges against a backdrop of unidirectional hydraulic gradients at larger spatial scales. For example, hyporheic exchanges can occur in response to localized hydraulic gradients arising from pool-and-riffle sequences in rivers (Harvey and Bencala, 1993). The apparent presence of hyporheic water prompted the decision to install deeper wells because the presence of oxic, hyporheic water in sediments under a river does not provide information about the redox state of ground water moving from terrestrial recharge areas to the river. The deeper (2.0 ft) ground water in the center and near the left edge contained suboxic ground water with slightly greater values of specific conductance (63 and 71 μS/cm at 25°C, respectively). The deeper ground water was more geochemically evolved than the shallower ground water, indicating that either hyporheic water had undergone a longer geochemical evolution, or ground water had entered the sediments from recharge areas away from the river. Regardless of the origin of the deeper ground water, the presence of suboxic ground water at a depth of 2.0 ft at these two points represents a NO₃^--reducing zone between the river at these locations and terrestrial recharge areas.

Site LD24.9, with nine wells, was more densely instrumented than the other sites in an attempt to provide a more detailed characterization of the nature and geometry of the oxic/suboxic boundary in the near-river environment (table 1). Seven of the nine wells were installed along one transect, and the other two wells were installed in the river thalweg (deepest part of the channel) 16 and 69 ft downstream of the primary transect. One nest of two wells (depths of 1.0 and 2.0 ft) was installed in the center of the river. One well was installed at the left edge of water, another well between the center of the river and the right edge of the river, and a nest of three wells at the right (east) edge of water (depths of 1.0, 2.0, and 3.0 ft).

The shallow well in the center of the river (LD24.9.1) yielded oxic water with a specific conductance of 
52 μS/cm at 25°C, similar to the specific conductance of the river water during well installation (46 μS/cm at 25°C). The deeper well in the center (LD24.9.4) was suboxic, with a specific conductance of 94 μS/cm at 25°C. As was observed at site LD16.1, a layer of oxic hyporheic water appears to overlie a deeper layer of suboxic ground water in part of the transect at LD24.9.

The nest of three wells at the right edge of water at site LD24.9 (fig. 3) was installed to provide increased definition of the geochemical nature of ground water entering the river from the east—the presumed source area for NO₃^- previously measured in ground water near this site (table A2, appendix A). The nest of wells was installed at the transect edge where the riverbed was composed of silt. However, there was no obvious bank slough, and the river reach at this site was primarily sand and gravel, indicating an overall coarse-textured sedimentary geologic framework.

The two off-transect wells at site LD24.9 were installed to provide information about the longitudinal scale of the NO₃^-–bearing ground-water plume at this site. These wells also helped characterize the spatial heterogeneity of oxic and suboxic ground water occurrence in near-river ground water.

At site LD32.8, an island caused most water to flow in a right channel, and a small amount of water to flow in a left channel. Site LD32.8 consisted of three wells in the right channel, including one in the center of that channel, plus one well near the left edge of the left channel. The three wells in the right channel were installed at a depth of 2.0 ft because of concern that the well-sorted, coarse gravel in the bed of the right channel at this site could promote hyporheic flow. The left channel, however, was sandy, and the well depth there was set at 1.0 ft.

Site LD37.8 consisted of four wells. The riverbed at site LD37.8 consisted primarily of well-sorted, coarse gravel (some gravely sand near the right edge), so the wells were installed at depths of 2.0 ft in an attempt to sample below possible hyporheic water.

Redox Conditions, Nitrogen, and Chloride in Ground Water from Transects

The previous section described well nests at transects LD16.1 (two nests) and LD24.9 (one nest) that yielded oxic ground water in shallower wells and suboxic ground water in deeper wells. Other wells yielding oxic ground water were (1) the well at the left edge of water at D198.5, (2) all three wells in the nest at the right edge of water at LD24.9, and (3) the two wells near the left edge of water at LD37.8. (The O₂ concentration in ground water near the right edge of water at D194.1.2, 0.5 mg/L was borderline oxic/suboxic. This ground water had a H₂S odor and an Fe concentration of 12,700 μg/L—one of the highest Fe concentrations measured in this study. These data indicate strongly reducing conditions. Thus, this sample is considered to be suboxic.)

Origin of water and solutes, availability of nutrients, and additional characterization of redox conditions can be inferred by evaluating concentrations of and relations between O₂, NO₃^-, NH₄⁺, Fe, and Cl^-. Relations between Fe and O₂ concentrations (fig. 4A) reflect the suppression of Fe dissolution in the presence of O₂ and in the absence of strongly acidic (for example, acid-mine drainage) conditions. This pattern is consistent with basic principles of redox progression in aquifers.

Relations between Fe and NO₃^- concentrations (fig. 4B) also are consistent with the expected sequence of redox progression in aquifers. Concentrations of NO₃^- are low or less than the MRL in Fe-rich ground water.

Nitrate and NH₄⁺ appear to be mutually exclusive (fig. 4C). NH₄⁺ is a reduced form of N, but reduction from NO₃^- to NH₄⁺ generally is for assimilation (Brock and Madigan, 1988), and most NO₃^- reduction in ground water is by denitrification, usually to N₂ (Freeze and Cherry, 1979). However, NO₃^- or NH₄⁺ occurrence does reflect environmental redox state—NO₃^- is stable and mobile in oxic ground water, whereas elevated concentrations of NH₄⁺ tend to occur in suboxic ground water (fig. 4D).

Patterns of NO₃^- and NH₄⁺ in ground water near La Pine also reflect different sources of NO₃^- and NH₄⁺. Concentrations of Cl^- also reflect NO₃^- and NH₄⁺ origin; Cl^- is a particularly useful tracer because it is nonreactive in most ground water systems. Septic tank effluent contains several tens of milligrams of N and Cl^- per liter, and NO₃^- and Cl^- co-occur in ground water receiving septic tank effluent. Other potential Cl^- sources, such as road salt applications, agricultural fertilizers, and dissolution of evaporites, generally are negligible in the study area (Hinkle and others, 2007). Ground water in the La Pine study area with background concentrations of Cl^- (on the order of a few milligrams per liter or less) indicate that ground water is unimpacted or minimally impacted by septic tank effluent, whereas concentrations of more than several milligrams per liter frequently indicate a septic tank effect. Mineralization of sedimentary organic matter, however, yields ground water characterized by elevated concentrations of NH₄⁺ but low concentrations of Cl^-. Patterns of NH₄⁺ and Cl^- (fig. 4E) reflect this source of NH₄⁺.

Redox data presented in figure 4 demonstrate understandable patterns of solute occurrence within a redox framework. Such a framework can be used to infer the sources of some solutes and to predict the occurrence of redox‑sensitive solutes within a given redox zone.

Spring and Slough Samples

Ground water enters rivers through sediments beneath the rivers; this part of the resource was the component intercepted by transect wells. Ground water also enters rivers by discharge from springs and seeps. Data collection activities often relied on canoe access, an activity that, along with additional scouting from roads and trails, provided opportunities for identification of springs and seeps. Numerous seeps and minor springs were observed, but only one flowing spring was found (river mile 209.2 on the Deschutes River) that yielded sufficient water volume for meaningful measurement of O₂ and for collection of water samples with a peristaltic pump (fig. 1, tables 1 and 2). The spring water was suboxic.

Samples were collected from one slough, a cut-off meander bend of the Deschutes River at river mile 194.8 (fig. 1, tables 1 and 2). At the time of sampling, the water elevation in the slough was greater than the elevation of the Deschutes River, and drainage occurred through a narrow, flowing stream connecting the slough to the river. The slough water was oxic, as would be expected for such a surface‑water body. However, other data from the slough indicated the presence of complex geochemistry.

Back to Table of Contents