USGS Home
Contact USGS
Search USGS

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

Back to Table of Contents

Introduction

Nitrate (NO₃^-) ingestion may adversely affect human health (World Health Organization, 1996), and discharge of ground-water NO₃^- into surface-water bodies can have detrimental effects on ecosystem health (U.S. Environmental Protection Agency, 1998; Howarth and Marino, 2006). Shallow aquifers are particularly susceptible to NO₃^- contamination because aquifers often receive nitrogen (N) inputs from anthropogenic activities at or near land surface. Shallow aquifers frequently have a direct connection to rivers, and hence represent a pathway for widely dispersed chemical constituents to enter rivers.

Ground water near La Pine, Oregon (fig. 1) is vulnerable to NO₃^- contamination as a result of residential development using conventional (non-nitrogen-reducing) onsite wastewater treatment systems (septic tanks). In ground water near 
La Pine, NO₃^- is derived primarily from septic tank effluent and occurs in the shallow part of the aquifer (approximately the uppermost 20 ft of saturated sediments). Evidence for a septic tank effluent source includes (1) similarity between δ¹⁵N values (nitrogen isotopic content) of NO₃^- in shallow ground water and those in septic tank effluent, (2) relations between NO₃^- and chloride (Cl^-), (3) relations between NO₃^- and tracer-based ground-water ages, (4) hydraulic gradients, and (5) occurrence of NO₃^- in discrete plumes (Hinkle and others, 2007). In coming years, NO₃^- concentrations in ground water will increase in many parts of the aquifer as historically loaded NO₃^- is transported farther along flow paths and as additional NO₃^- from existing and new homes is added to the aquifer (Morgan and others, 2007).

Ammonium (NH₄⁺), as well as NO₃^-, occurs in the aquifer. However, the distribution of NH₄⁺ differs from that of NO₃^-. Ammonium at elevated concentrations (concentrations greater than 1 milligram nitrogen per liter [mg N/L]) occurs in deep ground water (100–400 ft below the water table), or in shallow ground water near rivers where deep, NH₄⁺-rich ground water discharges to rivers. This NH₄⁺ is derived from decomposition of sedimentary organic matter (not from reduction of NO₃^-), as indicated by (1) δ¹⁵N values of NH₄⁺, (2) relations between NH₄⁺ and Cl^-, (3) relations between NH₄⁺ and carbon, (4) occurrence of NH₄⁺ in old ground water, and (5) location of NH₄⁺ in deep and downgradient regions of the aquifer (Hinkle and others, 2007).

Ground water reduction/oxidation (redox) conditions in the aquifer near La Pine evolve from oxygen-reducing (oxic) conditions near the water table in recharge areas to increasingly reducing (suboxic, or post-oxygen-reducing) conditions farther downgradient. (A dissolved oxygen [O₂] concentration of 0.5 mg/L was used by Hinkle and others [2007] as a threshold between oxic and suboxic conditions; this threshold value is used in this report.) N is a redox-sensitive element that usually occurs as NO₃^- under oxic conditions. In redox processes, NO₃^- becomes reduced (typically, in ground water, through denitrification, with an end product of nitrogen gas [N₂]) once most O₂ is reduced. Denitrification is the dominant NO₃^- sink in ground water in the aquifer near La Pine (Hinkle and others, 2007). Following denitrification, manganese (Mn) and iron (Fe) reduction occur, followed by sulfate (SO₄^2-) reduction and later by methanogenesis. Mn and Fe reduction usually involves reduction of solid-phase (sediment) Mn and Fe to aqueous-phase Mn and Fe. SO₄^2- reduction results in formation of hydrogen sulfide (H₂S), and methane (CH₄) is formed during methanogenesis.

The typical redox progression in ground-water evolution is:

1. Aqueous O₂ concentrations decrease as a result of O₂ consumption;
2. NO₃^- concentrations decrease (if present in recharge water) due to denitrification;
3. Mn and Fe concentrations increase;
4. SO₄^2- concentrations decrease and are accompanied by increased H₂S concentrations; and then
5. CH₄ occurs.

Complications to this redox sequence can occur. Redox conditions can evolve beyond Fe reduction without large increases in aqueous Fe concentrations if iron sulfide solid phases such as pyrite (FeS₂) precipitate from solution, or if aquifer materials do not contain sufficient solid-phase Fe for Fe reduction. Nevertheless, this progression of redox state as inferred by the relative abundances of these redox-indicator species generally is observed during the evolution of ground water along flow paths through aquifers. An understanding of redox state inferred by the occurrence of redox-sensitive species in ground water can be used to infer zones of chemical (for example, NO₃^-) stability and instability in aquifers. Generally, NO₃^- is not reduced to NH₄⁺ in ground water, but patterns of NH₄⁺ occurrence in ground water near La Pine largely correspond with redox conditions, with elevated concentrations of NH₄⁺ in suboxic ground water, and negligible concentrations in oxic ground water (Hinkle and others, 2007).

Previous studies of ground water in the La Pine region focused on the overall aquifer and emphasized the effects of septic tank effluent on the ground-water resource (Morgan and others, 2007; Hinkle and others, 2007). In these previous works, a flow model was developed to evaluate alternatives for management of NO₃^- loads from septic tanks to the aquifer system. The model simulated the three-dimensional ground-water velocity distribution in the upper 120 ft of alluvial sediments (Morgan and others, 2007). Redox zonation was used to conceptualize the occurrence and fate of NO₃^- in the aquifer and to represent a reaction boundary for NO₃^- in the flow model. These geochemical and modeling investigations, including aquifer redox zonation efforts, were not specifically designed to address questions about NO₃^- transport to rivers such as the Deschutes and Little Deschutes Rivers. An understanding of NO₃^- transport to rivers has been hampered by uncertainty in redox zonation near the rivers. Redox zonation near these rivers was poorly constrained in this previous work because (1) few data were available on redox conditions near rivers, and (2) redox conditions tend to exhibit tremendous spatial variability in the near-river environment (Hill, 1996). The thickness of the oxic zone in the aquifer in the near-river environment was estimated to be 10 ft by Morgan and others (2007), based on available but limited data from domestic and monitoring wells (Hinkle and others, 2007). This estimate allowed for the possibility of shallow, oxic ground-water discharge to rivers, with the realization that the oxic zone in the near-river environment likely was thinner or nonexistent in some zones.

A refined understanding of redox conditions near rivers in the study area is needed before the link between ground-water and river-water conditions can be more fully addressed. This work was initiated by a cooperative study by the Deschutes County Community Development Department and the U.S. Geological Survey (USGS). The resulting understanding could be used to help guide management decisions for river protection. For example, if NO₃^- is loaded to ground water in the zone of contribution (beginning of ground-water flow paths) during the recharge process, that NO₃^- can be carried into rivers where oxic ground water enters rivers. Removal of that NO₃^- could be accomplished by reducing NO₃^- loads in the zones of contribution. Alternatively, reduction of NO₃^- loads to rivers could be accomplished by altering geochemical conditions at the end of ground-water flow paths through, for example, the creation of NO₃^‑‑removing artificial wetlands in vulnerable discharge areas (Martin and others, 1999).

Purpose and Scope

The purpose of this report is to provide a conceptual framework describing the distribution of redox conditions in aquifer sediments near the Deschutes and Little Deschutes Rivers. Suboxic ground water in the aquifer near La Pine, Oregon, represents a zone of NO₃^- instability in which denitrification generally can be expected (Hinkle and others, 2007). This framework will provide resource managers with a refined understanding of the NO₃^- attenuation capacity of part of the aquifer that was poorly constrained in previous assessments of the overall aquifer. The improved understanding of redox state in the near-river ground-water environment may assist in evaluating river vulnerability to NO₃^- from ground water or in identifying steps that can be taken to reduce that vulnerability.

This report characterizes redox conditions in the near-river environment by presenting data from ground-water samples collected from sub-river sediments at various points along transects in the Deschutes and Little Deschutes Rivers and analyzed for redox indicator species (primarily O₂). Multiple measurements were made at each of 10 transects, with sampling designed to characterize spatial variability in redox conditions along transects oriented perpendicular to the rivers. These data were used to relate ground water redox state to geomorphologic and landscape variables and features that can be used to extrapolate the results to the near-river ground‑water environments of the Deschutes and Little Deschutes Rivers near La Pine. Particle tracking with an existing ground‑water flow model demonstrated one way to relate potentially vulnerable river reaches to upgradient recharge areas.

The redox characterization work described in this report is limited to the near-river environment. Although study results were consistent with basic principles of ground-water geochemical evolution and ground-water/surface-water interactions, they were restricted by the limited number of transects installed for the study. Therefore, conclusions drawn from this work were inherently general and the patterns observed hold for many, but not all river segments.

Description of Study Area

The La Pine study area, as defined in previous investigations by Morgan and others (2007) and Hinkle and others (2005 and 2007), encompasses 247 mi² of the upper Deschutes Basin in central Oregon (fig. 1). About 30 river miles of the Deschutes River and 50 river miles of the Little Deschutes River are represented in the study area. The area is underlain by Quaternary alluvial and lacustrine deposits in a structural basin of Quaternary and Tertiary basalt, andesite, vent deposits, and pyroclastic rocks (Lite and Gannett, 2002). These sediments contain an aquifer that provides the primary source of drinking water to most local residents, numbering about 18,000 in 2005 and expected to number 26,000 by approximately 2019 (Rich and others, 2005; Morgan and others, 2007).

Rain and snowmelt are the primary sources of recharge. Precipitation ranges from 15 to 20 in/yr over most of the area (Taylor, 1993), falling primarily from November through March. In addition to recharge from rain and snow, ground water also flows into the study area from neighboring high-elevation areas (Gannett and others, 2001). Ground water discharges to the Deschutes and Little Deschutes Rivers, which meander across a flood-plain-and-wetland complex as much as one-quarter-mile or more wide.

The Deschutes and Little Deschutes Rivers are N limited (Jones, 2003), and are susceptible to increases in N loads. Increased N concentrations can increase primary productivity, which can increase the magnitudes of O₂ and pH swings in river water. Parts of the Deschutes River do not meet State of Oregon water-quality standards for O₂ and are included in the state 303(d) list of water-quality-impaired streams. Data currently are insufficient to determine the status of O₂ conditions in the Little Deschutes River (Anderson, 2000).

Back to Table of Contents