| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a NHEERL-Western Ecology Division, USEPA, 200 SW 35th St., Corvallis, OR 97333
b USDA Agricultural Research Service, Corvallis, OR 97331
c Oregon State University, Corvallis, OR 97331
d CA 95616
* Corresponding author (Wigington.Jim{at}epa.gov)
Received for publication September 27, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: CR, cultivated riparian zone • FLD, ryegrass seed field • LCE, Lake Creek East study site • LCW, Lake Creek West study site • NCR, noncultivated riparian zone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The ability of vegetated riparian areas to influence the quality of water draining from agricultural lands has been demonstrated by a number of studies (Lowrance et al., 1984; Peterjohn and Correll, 1984; Gilliam, 1994; Correll, 2000). The hydrologic flowpath of water as it moves from agricultural fields to streams is a critical determinant of the effectiveness of riparian areas in processing nutrients and chemicals in agricultural waters (Phillips et al., 1993; Lowrance et al., 1997). The quality of water that moves through the biologically active rooting zone in riparian areas is more likely to be improved than water that by-passes the riparian zone by movement as deep ground water or channelized surface runoff.
The Willamette Valley of Oregon has extensive areas of poorly drained lands that frequently are used for commercial grass seed production because the land is marginally productive for most other agricultural systems (Griffith et al., 1997). Unfortunately, few riparian studies have examined the water quality functions of riparian areas in poorly drained landscapes. In 1994, we initiated a study to examine the ability of riparian areas in poorly drained landscapes to process nutrients and chemicals in water draining from grass seed fields. In this paper, we evaluate the effectiveness of a grass–forb riparian buffer in removing nitrate from water draining from grass seed agricultural fields. We hypothesized that for our study site, shallow ground water draining from the grass seed fields was the predominant water flowpath from field to streams and that nitrate concentrations would be reduced as subsurface water moves through the rooting zone of the grass–forb, noncultivated, riparian area.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
We established two study sites along or near an intermittent tributary of Lake Creek (referred to hereafter as Lake Creek) that drains perennial ryegrass fields in the southern Willamette Valley of western Oregon (44°32' N, 123°03' W). The two sites, Lake Creek West (LCW) and Lake Creek East (LCE), represented perennial ryegrass seed fields with similar soils and hydrology but contrasting riparian vegetation (Fig. 1) . Neither site had subsurface drainage tiles. The drainage area for LCW is 168.2 ha and for LCE is 133.7 ha.
|
Lake Creek East had a cultivated riparian zone (CR) in which the perennial ryegrass field extended to the edge of Lake Creek. The shallow concave creek banks within 10 to 15 m of the center of the creek had a sparse collection of other grass species, including barnyard grass [Echinochloa crus-galli (L.) P. Beauv.], Bermuda grass [Cynodon dactylon (L.) Pers.], and redtop (Agrostis gigantea Roth).
At both the LCE and LCW sites, the perennial ryegrass seed crop was established in the fall of 1994 and followed previous perennial ryegrass seed crops (Horwath et al., 1998). The new ryegrass seed crop was established as a row crop with conventional tillage. The grass seed crop was managed by a local grower with standard production practices for western Oregon (Youngberg, 1980). Nitrogen fertilizer (163–199 kg N/[ha yr]) was applied in the spring of each year in form of urea–ammonium sulfate.
The Lake Creek sites are typical of grass seed growing areas in the southern and mid-Willamette Valley in several ways (Parsons et al., 1970). Topography at the Lake Creek sites is quite flat, with slopes < 3%. Soils at the study sites are poorly drained (Dayton–Holcomb association) and are marginally productive for most agricultural crops except those such as perennial ryegrass, which can tolerate waterlogged conditions. A band of Dayton soil (fine, smectitic, mesic Vertic Albaqualf) encompasses the stream to a width of 25 m at LCW and a width of 18 m at LCE. At LCE, the presence of Dayton soils was used to define the cultivated riparian zone. The Dayton soil series is located on nearly level terrain that was formed by alluvial deposits found adjacent to streams (Knezevich, 1975). Dayton soils are characterized by a very slowly permeable to impermeable, clay IIBt horizon. Because of the presence of this layer, little downward seepage of water can take place and a perched water table occurs during the wet season (Boersma et al., 1972). The Dayton soils also have a silty clay loam texture IIIC horizon. Holcomb (fine, smectitic, mesic Typic Argialboll) soils occur throughout the remainder of the study sites. The Holcomb soil series has a clay texture IIBt horizon, a silt loam IIIC horizon, and occupies a more upland landscape position than Dayton soils. A and E horizons with silt loam or silty clay loam textures overlay the IIBt horizons in both the Dayton and Holcomb soils.
Instrumentation, Sampling, and Analyses
To examine the changes in water quality parameters of ground water as it moved from the grass seed fields through the riparian zones at LCW and LCE, we installed a series of nested piezometers along three transects at each site (Fig. 1). Each piezometer nest consisted of one piezometer in the A and E horizons (designated as A/E-horizon piezometers) and one in the C horizon at a given location. Piezometer nest positions along the transects were determined by the relative position of riparian vegetation, grass seed fields, and soil types.
At LCW, six piezometer nests were located along each of three transects located along the predominant ground water flowpath to Lake Creek (Fig. 1). Transects were 17.5 m apart. Piezometers in the NCR were located at 0.5 m (Row 1), 8 to 9 m (Row 2), 15 to 17 m (Row 3), and 24 to 32 m (Row 4) from Lake Creek. Piezometers in the FLD were 36 to 44 m (Row 5) and 51 to 58 m (Row 6) from Lake Creek. The NCR had Dayton soil near the stream (Rows 1–3) and Holcomb soil near the FLD (Row 4). All piezometers in FLD (Rows 5 and 6) were located in Holcomb soil.
At the LCE site, four piezometer nests were located along each of three transects, which were 15 m part (Fig. 1). Piezometers in the CR were located at 5 m (Row 1) and 16 m (Row 2) from Lake Creek, and piezometers in the FLD were located 22 m (Row 3) and 40 to 47 m (Row 4) from Lake Creek. The CR piezometers were located in Dayton soil (Rows 1 and 2), and the FLD piezometers (Rows 3 and 4) were located in Holcomb soil.
We used 51-mm-i.d. TIMCO (Prairie du Sac, WI) PVC slotted, high-flow piezometers with a slot width of 0.25 mm and vented caps. Deep (C horizon) piezometers in the FLD and CR had a series of watertight screw joints to allow for disassembling at the soil surface and below the plow depth to allow for agricultural harvesting and tillage operations in the grass seed field. Shallow (A/E horizon) piezometers in the FLD and CR had joints at the soil surface to allow disassembling during harvest operations. All joints were self-sealing and did not require the use of glues or O rings. The C-horizon piezometers were screened at a depth of 1.22 to 1.83 m below the soil surface, a zone just below the IIBt horizons of the Dayton and Holcomb soils. The dimensions of the shallow A/E-horizon piezometers varied with the soil series in which they were located. Shallow piezometers in the Dayton soils had screens 15 to 36 cm below the soil surface and in the Holcomb soils had screens 18 to 53 cm below the surface. These screening intervals allowed water samples and hydrologic measurements to be made in the A and E horizons above the IIBt horizon.
We installed the piezometers during late summer and early fall 1995 by driving the piezometers into hand-augured holes that were same size as the outside diameter of the piezometers. A collar of native clay from the Dayton IIBt horizon was packed around the piezometers at the soil surface to prevent movement of ponded water down along the piezometer casing. The soils at the study sites, especially the IIBt horizons, have large amounts of montmorillinitic (high shrink–swell) clays. During the early rains of October and early November, soil around the piezometer casings swelled and a tight seal was formed. The native clay collars were repacked each summer. After installation, we developed each piezometer according to standard protocols to ensure proper operation (Kill, 1990).
Ground water samples were collected from piezometers every two weeks from November 1995 through June 1996 and every three to four weeks from November 1996 through June 1997. Ground water levels dropped below the depths of the piezometers from late June through late October of each year and occasionally during the sampling periods. To ensure that water samples represented actual soil water characteristics, one case volume of water was removed from each piezometer with a peristaltic pump and the piezometer was allowed to recharge prior to a water sample being pumped for collection.
Water samples were transported to the laboratory immediately after collection, filtered with a 0.45-µm filter, and stored at 4°C until analyses were performed. We determined NO3–N and NH4–N colorimetrically with a Lachat (Milwaukee, WI) Quikchem 4200 flow injection (American Public Health Association, 1992). Quality assurance measures included routine use of blanks, duplicate samples, and quality control check samples.
We used two-sided Wilcoxon rank-sum tests (Sprent, 1990) to evaluate differences in solute chemistry of water collected from piezometers installed in the same soil horizons (e.g., C horizon) in FLD and riparian zones at LCW and LCE. At LCW, piezometers in Rows 5 and 6 represented FLD conditions and piezometers in Rows 1 through 4 represented NCR conditions. At LCE, piezometers in Rows 3 and 4 represented FLD conditions and piezometers in Rows 1 and 2 represented CR conditions.
We measured water level elevations in each of the A/E-horizon and C-horizon piezometers at LCW and LCE prior to pumping the piezometers for water quality sampling. We also equipped three additional C-horizon piezometers with vented, Geokon (Lebanon, NH) vibrating wire pressure transducers to allow continuous monitoring of water table elevations (Fig. 1). Precipitation was measured at LCW with a Texas Electronics (Dallas, TX) tipping bucket rain gauge. Stream stage was measured at LCW with a vented vibrating wire pressure transducer installed in a slotted PVC stilling well set into the stream bank (Fig. 1). The stream channel at LCW was much better defined than at LCE and was better suited for stream gauging. Campbell (Logan, UT) CR10X data loggers collected data from sensors every 10 min. A professional surveyor prepared a topographic map with a contour interval of 0.15 m of the study sites and established the locations and elevations of all piezometers and monitoring equipment.
We estimated stream discharge at LCW with a rating curve (r2 = 0.99) based on stream stage and discharge values calculated with Manning's equation (Albertson and Simons, 1964). The rating curve was validated with flow measurements made with a Swoffer (Seattle, WA) flow meter at varying stage heights up to bank full flow. On average, measured streamflow and Manning equation estimates of streamflow agreed within ±49%, with the largest relative errors occurring at low flows. We chose not to use precalibrated flumes or weirs because we did not want to disturb the natural flooding regime of the study sites.
We conducted a series of slug tests (Hvorslev, 1951) in 10 shallow piezometers at LCE and LCW during wet, winter conditions to determine saturated hydraulic conductivity values of A/E horizon soils. We also measured soil redox potentials (platinum vs. standard hydrogen electrode [SHE]) with in situ soil electrodes installed at two soil depths (Patrick et al., 1996). All redox readings are reported relative to the standard hydrogen electrode with laboratory potentials at 25°C for the reference electrodes.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
To evaluate the net effectiveness of the riparian zones at LCE and LCW, we needed to be able to estimate the proportion of the streamflow in Lake Creek that is likely to flow through the biologically active A and E horizons of riparian zones along Lake Creek. Based on the saturated hydraulic conductivity values (Table 1) that we measured in shallow piezometers at LCE and LCW, we used Darcy's law (Harr, 1991):
 | [1] |
0.3 m. Furthermore, we assumed that the average gradient of the shallow ground water table was equivalent to the average land surface slope of 1.5%. For the moderate-flow conditions of 24 April, we assumed that the length of wetted channels was half of the 21 January length and that the average depth of water in the channel was 0.1 m. These assumptions were based on field observations of channel expansion and contraction over the period of the study.
|
|
Based on the hydrometric data collected during this study and field observations at the study sites, we propose three primary mechanisms contributing to streamflow for Lake Creek and similar watersheds: (i) ground water rising to watershed surfaces and exfiltrating into stream channels, swales, and ponded areas; (ii) rainfall on saturated watershed surfaces with rapid movement to swales, channels, and saturated depressions; and (iii) drainage of water from soil A/E horizons to channels and swales (Dunne and Leopold, 1978). Based on our estimates, the third mechanism, flow from soil A horizons, appears to be a small component of streamflow during moderate and high streamflow periods (Table 2). Therefore, we believe that most of the water in Lake Creek has reached the stream channel without significant interaction with riparian zone soils along the intermittent channels near Lake Creek West and Lake Creek East. Temporal and spatial patterns of diuron [3-(3,4-dichlorophenyl)-1,1-dimethyl urea] movement in surface water and ground water support these conclusions (Field et al., 2003). Diuron transport occurred primarily as streamflow with little evidence of movement in the shallow or deep piezometers at the Lake Creek sites. Dunne and Black (1970a)(b) noted the importance of similar mechanisms to a small agricultural watershed with permeable soils in Vermont.
Riparian Zone Influence on Nitrate and Ammonium Nitrogen
In the previous section, we demonstrated that a relatively small percentage of the water draining into Lake Creek moves through riparian zone soils adjacent to Lake Creek. Nevertheless, it is important to understand the influence that riparian zones and associated biogeochemical processes have on the nitrogen concentrations of water that does move through the riparian zone soils. Table 3 presents median NO3–N concentrations, and Table 4 presents NH4–N concentrations for sampling dates with complete samples for all of the piezometers in either the A/E horizons or C horizon at a given site. Complete sample sets (a water sample available for each piezometer at a given depth) were required for the two-sided Wilcoxon rank sum test that were used to evaluate the differences between ground water solute chemistry in the FLD (piezometers in Rows 5 and 6 at LCW and in Rows 3 and 4 at LCE) and riparian zones along Lake Creek (NCR at LCW and the CR at LCE). Statistical comparisons were performed by soil horizon within a given site. Water levels at the study sites were more frequently at or above the screens on the C-horizon piezometers, and consequently, there are many more sampling dates for the C-horizon piezometers with complete samples. Complete A/E-horizon data sets generally were collected during the wet winter season. Lake Creek East had more sampling dates with complete A/E-horizon data sets because there were fewer piezometers at LCE and the piezometer transects at LCE did not extend as far away from the stream as at LCW.
|
|
Ammonium concentrations were much lower than NO3–N concentrations at both LCE and LCW (Table 4). For C-horizon NH4–N concentrations, there were few dates with statistically significant differences between FLD and CR at LCE or between FLD and NCR at LCW. When differences did occur, they were very small. There were no sampling dates with statistically significant differences between LCW FLD and NCR A/E-horizon NH4–N concentrations. The A/E-horizon NH4–N concentrations were greater in the Lake Creek East CR than in the FLD for seven sampling dates.
These results provide important insights regarding the influence of riparian zones along intermittent streams in poorly drained agricultural landscapes of the Willamette Valley. Noncultivated riparian (NCR) zones, such as the one along Lake Creek West, can reduce the NO3–N concentrations of waters draining from grass seed fields. Nitrate concentrations in A/E-horizon and C-horizon ground waters in the noncultivated riparian zone at LCW were distinctly lower than in the grass seed field at LCW. Similar reductions in nitrate concentrations did not occur in the LCE cultivated riparian zone. Plant uptake and microbial processes, including denitrification, are important processes in reducing NO3- concentrations of shallow ground water flowing from grass seed agricultural fields (FLD) through noncultivated riparian zones (NCR) (S.M. Griffith, unpublished data, 2002).
A major indicator of the biogeochemical differences between the NCR at Lake Creek West and the CR at Lake Creek East was soil redox potential (Fig. 5) . In spite of the fact that the Lake Creek East CR was flooded more frequently that the NCR at Lake Creek West, A/E-horizon redox potentials were higher in Lake Creek East CR soils than in the Lake Creek West NCR soils. Lake Creek West NCR A/E-horizon redox potentials also were much lower than in the Lake Creek West FLD or Lake Creek East FLD. These differences in redox potential were probably caused by the higher levels of soil C in the Lake Creek West NCR soil and led to higher denitrification rates at the Lake Creek West NCR (S.M. Griffith, unpublished data, 2002). Consequently, NO3–N concentrations were lower in A/E-horizon and C-horizon piezometers in the Lake Creek West NCR than in the Lake Creek East CR.
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H. Davis, S. M. Griffith, W. R. Horwath, J. J. Steiner, and D. D. Myrold Denitrification and Nitrate Consumption in an Herbaceous Riparian Area and Perennial Ryegrass Seed Cropping System Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1299 - 1310. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Vidon, L. P. Tedesco, J. Wilson, M. A. Campbell, L. R. Casey, and M. Gray Direct and Indirect Hydrological Controls on E. coli Concentration and Loading in Midwestern Streams J. Environ. Qual., August 8, 2008; 37(5): 1761 - 1768. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Davis, S. M. Griffith, W. R. Horwath, J. J. Steiner, and D. D. Myrold Mitigation of Shallow Groundwater Nitrate in a Poorly Drained Riparian Area and Adjacent Cropland J. Environ. Qual., April 5, 2007; 36(3): 628 - 637. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Davis, S. M. Griffith, W. R. Horwath, J. J. Steiner, and D. D. Myrold Fate of Nitrogen-15 in a Perennial Ryegrass Seed Field and Herbaceous Riparian Area Soil Sci. Soc. Am. J., April 19, 2006; 70(3): 909 - 919. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Rupp, R. E. Peachey, K. L. Warren, and J. S. Selker Diuron in Surface Runoff and Tile Drainage from Two Grass-Seed Fields J. Environ. Qual., January 5, 2006; 35(1): 303 - 311. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bedard-Haughn, K. W. Tate, and C. van Kessel Using Nitrogen-15 to Quantify Vegetative Buffer Effectiveness for Sequestering Nitrogen in Runoff J. Environ. Qual., November 1, 2004; 33(6): 2252 - 2262. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Field, R. L. Reed, T. E. Sawyer, S. M. Griffith, and P. J. Wigington Jr. Diuron Occurrence and Distribution in Soil and Surface and Ground Water Associated with Grass Seed Production J. Environ. Qual., January 1, 2003; 32(1): 171 - 179. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
piezometers.