Water balance of a dairy loafing lot using geotextile and its impact on water quality.
Adriano et al. (1971) documented significant ground water contamination from nitrates and high soil nitrate (N[O.sub.3]) concentration attributed to management of dairies and animal manure in California. Similar land-use management is occurring today in North Central Georgia. Nearly 150 of Georgia's 422 dairies are found in the five Piedmont counties directly south of Atlanta and Athens. The University of Georgia Cooperative Extension Service sampled 138 rural wells from 1991 through 1992 and found 13 percent exceeded the 10 mg [L.sup.-1] N[O.sub.3]-N maximum contamination limit set by the Environmental Protection Agency (EPA) (Gould, 1993). When samples were grouped into farm versus non-farm, 17 percent of the on-farm wells exceeded the maximum contamination limit standard compared to only five percent for non-farm wells. Since more than 90 of these farms were operating dairies, the source of pollutants was attributed to the concentration of animals on these dairies.
Many dairies in North Central Georgia consist of small (< 300 animal) operations that do not confine the animals to barns. These operations utilize adjacent pastures to graze animals between milkings. To keep milk production at a high level, many producers limit the amount of time the cows are allowed to graze, and feed the animals a grain ration in the barn. When the animals are not on pasture or being milked, they are confined near the barn in an unpaved paddock. One result of this practice is the paddock becomes mostly barren and accumulates large amounts of manure. This accumulated manure is infrequently removed. During periods of abundant rainfall, these areas become extremely muddy and during drier times sustain only weedy species such as spiny pigweed (Amaranthus spinosus L.) and musk thistle (Carduus nutans L.). The paddocks are called "loafing areas" or "loafing lots" by local producers and are of various dimensions and shapes. Management varies widely, as these lots become the overflow pen, or holding area between milkings.
In previous work at the University of Georgia, Drommerhausen et al. (1995) documented high levels of salts in loafing lots at several dairies. Nitrate levels extracted from soil samples taken in the loafing lot at a dairy near Rutledge, Georgia averaged 30 to 40 mg N[O.sub.3] k[g.sup.-1] to depths greater than two meters. Monitoring wells were installed and groundwater nitrate levels greater than 100 mg N[O.sup.3] [L.sup.-1] were found directly below the loafing area. Chloride levels in the ground water were of similar magnitude. Two other dairies in this study had ground water N[O.sub.3] concentrations of 60 and 120 mg [L.sup.-1] directly below their loafing areas.
Overcash and Phillips (1978) estimated runoff potential from loafing lots for North Carolina dairies to be around 30 to 40 percent of precipitation, with greater percentages through the wetter winters. Greater runoff losses would be expected with increased amounts of impervious surfaces (concrete). They did not address percolation, or attempt a complete water balance for loafing lots.
During extended wet periods, soils within loafing lots can become damaged under high stocking density resulting in conditions where cows sink to their bellies in mud. To avoid this management problem, the Natural Resources Conservation Service (NRCS, 2004) in Georgia, has begun to install a manufactured geotextile in area loafing lots. For use in these loafing areas, the land area is smoothed and graded, the geotextile material is laid down on the surface and a 15 cm (6 in) depth of crushed gravel is placed on top. The geotextile material serves to keep the gravel separate from soil below and thereby distributes the load of cattle and keeps them from sinking into the mud (Ruhl et al., 2003). The gravel surface also improves infiltration and eliminates areas of standing water. Another alternative for improving loafing lots is to pave the areas with concrete, but this produces more runoff and requires frequent wash-downs and manure handling. Also, utilizing gravel and geotextile material for loafing lots is less expensive, costing about 25 to 33 percent of the cost of concrete (Ruhl et al., 2003).
Currently there are a number of these new loafing areas in use in North Central Georgia. Local herd management has improved but questions concerning runoff and percolation and the impact to water quality remain. We hypothesize that a significant portion of annual precipitation will move to ground water at this loafing lot site, and that the quality of percolating water will be impaired. The objective of this study was to determine the concentration of nutrients and the amount of runoff and percolation occurring at an NRCS designed loafing lot system.
Methods and Materials
Loafing lot construction. A loafing lot, designed by the engineering staff of the Natural Resources Conservation Service, Watkinsville, Georgia, was constructed at a 60-cow dairy in Oglethorpe County Georgia (Figure 1). Approximate dimensions were 20 m (65 ft) by 40 m (130 ft). The soil in the loafing lot was mapped as a Pacolet sandy loam (clayey, kaolinitic, thermic, Typic Kanhapludult) on an approximate 5 percent slope (Table 1). The water table at this upland site was estimated to be approximately 10 m (33 ft) below the surface (LeGrand, 1988). Tile drains with 0.10 m (4 in) diameter perforated plastic pipe were placed on intervals of 7.5 m (25 ft) under the entire area at approximately 0.65 m (26 in) depth below the original soil surface in a 0.3 m (12 in) wide bed of No. 60 gravel. This placed the bottom of the drain lines near the middle of the Bt2 horizon. Soil above the drain lines was mechanically repacked to inhibit preferential flow in the disturbed area above the drains. Construction of earthen berms on the downslope sides (west and south) resulted in the removal of most of the surface A horizon in the loafing lot. The remaining surface of the lot was smoothed, covered with a geotextile material (Series 180-N, manufactured by T.C. Mirafi, Pendergrass, Georgia), then covered with 7.6 cm (3 in) of crusher run [mixed sized of 5 cm (2 in) diameter and less] gravel, and an additional 7.6 cm (3 in) of granite dust. Finally, the area was fenced for animal confinement.
Sample collection and analysis. All drain lines were connected to a common header that discharged into a sump for flow measurement and sampling. Water flow from tile drains was quantified using a custom made 2.5 L (0.66 gal) tipping bucket where total number of tips was recorded on five-minute intervals. Surface water was routed to a 7.6 cm (3 in) parshall flume for flow measurement and automated sampling. For comparison, runoff was also estimated using the NRCS curve number (CN) method utilizing a CN of 82 as suggested by NRCS (U.S. Department of Agriculture, 1969).
Water samples were collected during storm events by a refrigerated sampler (model 6700FR, ISCO, Lincoln, Nebraska) and held at 4[degrees]C (39[degrees]F) until analysis. Samples were collected and processed usually within one week of an event. Surface samples and tile drainage water samples were each collected starting at three-minute intervals during storm and drainage events. Samples were collected more frequently in the beginning of storms, and less frequently as storms continued. Results from events were combined to provide monthly estimates of nutrient loss. On average, six surface water samples were collected per event with an average of three events per month. Background levels of nutrients were determined on runoff and drainage samples collected and analyzed over a six-month period following construction but prior to placing cattle in the loafing lot.
Construction was completed in the spring of 1997. Approximately 60 cows were confined to the loafing lot beginning in September of 1997 and were there for 6 weeks. After that period, they were in the loafing lot only on weekends. This provided a stocking density of approximately one animal per 13 [m.sup.2] (140 f[t.sup.2]) for a six-week period. The NRCS technical standard specifies a stocking density of no more than one dairy cow per 18.6 [m.sup.2] (200 f[t.sup.2]) (NRCS, 2003). According to van Vuuren and Meijs (1987), a 600 kg (1320 lb) lactating cow excretes about 40 g N [d.sup.-1] (0.09 lb N [d.sup.-1]), as waste. This rate would result in N accumulation in the first 6 weeks of approximately 100 kg-N (222 lbs-N) in the loafing lot. P deposition was approximately 17 kg (38 lb) based on this estimate for N and an N/P ratio for dairy manure of 5.9 (Forster, 1998).
Daily weather data were obtained for Watkinsville, Georgia from the Georgia Automated Environmental Monitoring Network (Hoogenboom, 2001), for determination of evaporation potential. This location is within 40 km (25 mi) of the research site. Rainfall and runoff were recorded locally on five-minute intervals during storm events using a CR10 datalogger (Campbell Scientific, Inc., Logan, Utah). Rainfall was measured with a tipping bucket rain gauge that had a minimum sensitivity of 0.1 mm (.004 in). Runoff was determined using a pressure transducer (Microswitch, Minneapolis, Minnesota) located in a stilling well connected to the side of the flume and calibrated in place. The quantity of water discharged was a function of height of water in the flume and was calculated using the method of Brakensiek et al. (1979).
Evaporation from the loafing lot was calculated using a modified version of the FAO-56 reference crop evapotranspiration equation (Allen et al., 1998; Allen, 2000). The original FAO-56 model is based on the Penman-Monteith equation and is parameterized for a 12 cm (4.7 in) tall grass with a surface resistance of 70 s [m.sup.-1] (21 s f[t.sup.-1]) and an albedo of 0.23. To adjust the equation for the surface conditions at the loafing lot, aerodynamic roughness length for momentum and heat were changed to 0.1 and 0.02 cm (.04 and 008 in), respectively; and surface resistance was adjusted to 525 s [m.sup.-1] (160 s f[t.sup.-1]). This value of surface resistance is representative of a soil with a dry surface but moist conditions immediately below the surface (Ham and Heilman, 1991). Albedo was adjusted to 0.20, which is typical of a drying soil and appropriate for a gravel surface. With the exception of these adjustments, evaporation was computed exactly as described by Allen et al. (1998). Because the surface moisture conditions at the loafing lot were not quantified, there is considerable uncertainty in the estimate of the surface resistance used to model evapotranspiration. In reality, resistance probably ranged from 100 to 2000 s [m.sup.-1] (30 to 610 s f[t.sup.-1]) during wetting and drying cycles. A value of 525 s [m.sup.-1] (160 s f[t.sup.-1]) was selected as the best estimate of the median value.
Soil samples were obtained from grid sampling at ten equidistant locations across the lot in increments to 1.6 m (5 ft) prior to lot construction and after 1.5 years of operation. Soil samples were extracted (1:5) with 0.5 M [K.sub.2][SO.sub.4] and analyzed for nitrate, ammonium, and chloride for both sampling dates. Nitrate was determined colorimetrically using the Griess-Ilosvay method (Keeney and Nelson, 1982) after reduction of nitrate to nitrite with a cadmium column, as the sum of nitrate plus nitrite. Ammonium was determined colorimetrically using the salicylate-hypochlorite method (Crook and Simpson, 1971). Chloride was determined by coulometric titration (Cotlove et al., 1958). Phosphorus was determined on soil samples extracted (1:5) with Mehlich 1 and the concentration determined colorimetrically by reduction with molybdate (Murphy and Riley, 1962). Storm water samples were filtered through 0.45 (m filters and then analyzed for soluble nitrate, ammonium, phosphorus and chloride by the same methods. Physical parameters of soil were measured using intact cores (6-cm X 8.6 cm diameter or 2.3 in X 3.5 in diameter) obtained from each horizon. Saturated hydraulic conductivity was determined using the constant head method (Reynolds et al., 2002). Bulk density was calculated by determining the mass of soil weighed after oven-drying cores at 105[degrees]C (221[degrees]F). Soil water characteristic curve data were collected using pressure cells and a pressure plate extractor (Dane and Hopmans, 2002a; Dane and Hopmans, 2002b). Van Genuchten parameters were fitted to the soil water characteristic curve data (van Genuchten, 1980). Field capacity and permanent wilting point water contents were calculated using these fitted parameters with water potentials of -0.03 MPa (-1/3 bar) and -1.5 MPa (-15 bar), respectively.
Results and Discussion
Soil profile characteristics. The measured physical properties of the soil horizons for the loafing lot are shown in Table 1. This profile is typical for a Pacolet soil and for other similar soils such as the Cecil soil. The Cecil and Pacolet series occupy approximately 20 percent of the mapped land area in the southern Piedmont region (Radcliffe and West, 2000). The minimum [K.sub.s] typically occurs in the lower Bt or upper BC horizons in these soils (Bruce et al., 1983), as it did at this site (in the Bt2 horizon). Following significant rainfall, this characteristic periodically perches water at this interface, whereas the water table at an upland site such as this normally occurs at depths of 6 to 15 m (20 to 50 ft) below the surface (LeGrand, 1988).
Water balance. For the water balance we measured precipitation, runoff, and tile drainage at this site for approximately two years. There was no vegetation on the loafing lot so transpiration did not occur. The remaining unmeasured pools of water are evaporation, soil water storage, and deep percolation. For most storms, runoff occurred soon after initiation of rainfall, and followed the intensity pattern of the rainfall. Monthly averages, each containing several storms, indicated that approximately 18 percent of total rainfall became surface runoff (Table 2). This is low compared to percentages cited by Overcash and Phillips (1978), which range from 31 to 100 percent for Wisconsin dairy loafing lots and 20 to 39 percent for western beef cattle feed lots. Measured runoff compared well with that predicted using NRCS curve number (CN), especially for larger events (Figure 2). This independent estimate of surface water yield increased confidence in measured values reported in Table 2. The slope of the regression (Figure 2) was slightly higher (1.06) than unity indicating that less runoff occurred than was expected for this surface condition. Reduced runoff in this system as compared to other lots can be attributed to the protection and improved load distribution the gravel surface provides that reduces compaction of the soil beneath the gravel.
Tile drains were installed well above the ground water table, and collected water only under saturated conditions. They recovered a very small percentage of total precipitation, totaling less than one percent. This low recovery is being attributed to the mechanical packing above the drains during installation, which disconnected the flow path between the remaining A horizon and rock layer at the surface. Since the A horizon was essentially replaced by a layer of rock and gravel, this layer provided ready storage for intense storms, but drained and remained dry between storm events. Some lateral drainage within the rock mulch could have occurred which would have surfaced near the lower berm. Water flowing through the rock mulch and then surfacing would have increased our estimate of surface runoff. Our flume measurements don't show evidence of significant delayed runoff, which would account for such an error. The saturated hydraulic conductivity of the Bt1 horizon is six times greater than that of the Bt2. For significant lateral flow to occur within those horizons the conductivity of a restrictive layer should be at least 10 times lower than those layers immediately above. Therefore it is conceivable that little lateral flow occurred, which helps explain the reduced volume of water intercepted by tile drains as compared to the designed system.
Although evaporation was not measured directly at the loafing lot, data collected at a complete weather station near the area provides a reasonable prediction of water loss through this pathway. Little ponding was evident following storms, and the gravel surface remained fairly dry between storms, which created a barrier to vapor flow between the soil and atmosphere. Estimates of evaporative losses accounted for approximately 41 percent of the rainfall received over the two years of monitoring.
Two unmeasured pools for water remain. Soil water storage was estimated using the saturated and permanent wilting point water contents in Table 1. This difference, when used as an estimate for water storage in each horizon, multiplied by the horizon thickness and summed provided an estimated total water storage of 167 mm (6.5 in) of water in the first 1.5 m (60 in) of soil profile (ignoring water storage in the gravel layer). Thus soil water storage could account for a maximum of 9 percent of the total water balance. By difference, the remaining pool, deep percolation, was estimated to be 32 percent of precipitation. Thus a very significant portion of the water balance appears to have gone to deep percolation at this site. In 1998, when a more complete data set for drainage was collected, 13 percent of the days monitored had subsurface flow in the tile drains. Using a conservative estimate of K, for the gravel/sand layer of 60 mm [d.sup.-1](2.3 in [d.sup.-1]), the effective saturated hydraulic conductivity to 4 m (157 in) was 26 mm [d.sup.-1] (1 in [d.sup.-1]). Assuming gravity flow, potentially more than 1000 mm (39 in) of water could have moved to ground water during times of saturation in 1998. Unsaturated flow between times of profile saturation would have contributed even greater volumes to ground water.
Water chemistry. Changes in chemistry of both surface water and percolating water occurred rather quickly after cattle were confined in the loafing lot starting in September 1997. The concentrations of most analytes in runoff water samples collected after the cattle were confined in the lot were significantly higher than concentrations prior to this date (Table 3). Since sodium is an important constituent of the cattle diet, and sodium chloride is a low cost source of sodium, a significant amount of chloride (up to one percent on a dry weight basis) can be found in dairy cattle rations (Granzin and Gaughan, 2002). Chloride excreted in urine and in manure then provides a convenient tracer for marking manure deposition. The increase in concentration of chloride over time in runoff water follows the trend in stocking rate of cattle and thus the loading rate of manure (Table 3). This spike in chloride concentration of runoff water dropped as the stocking density of the lot was reduced over time. Orthophosphate concentrations of runoff water increased to approximately 10 mg P [L.sup.-1] as soon as the cattle were confined, and those levels were maintained throughout the rest of the monitoring period. Nitrate levels increased above background as soon as cattle were confined, but remained low until March 1998, when warmer temperatures allowed mineralization to occur. At that point levels quickly exceeded the 10 mg N [L.sup.-1] EPA drinking water standard in surface runoff. Ammonium levels followed the trend of chloride where maximum values occurred around the time of maximum deposition with some increase due to mineralization in March, 1998.
Percolation water captured by tile drains had a different chemical signature than surface runoff (Table 4). Little percolation occurred, and sample numbers were limited. But by the end of one month of confinement, nitrate levels of percolation water had increased twelve-fold over those measured as background, and remained high throughout the monitoring period. The maximum monthly N[O.sub.3] concentration was 60.4 mg [L.sup.-1]. Chloride concentrations peaked during November, 1997, and declined thereafter as herd management dictated less frequent use of the loafing lot. Ortho-phosphate concentrations remained low, but were significantly greater than background concentrations measured prior to confinement of cattle.
Soil chemistry. Increased concentrations of nitrate and chloride in the soil beneath the loafing lot were confirmed through analysis of soil samples after 1.5 years of operation (Figure 3). Using a paired student t-test, levels of both chloride and nitrate were significantly increased over the period of the study in the upper 0.5 m (20 in) of soil. At greater depths, very little change occurred over the 1.5-year period. A significant pulse of chloride evident in the initial samples and centered around 1.2 m (47 in) below the surface remained at that depth but showed a reducing trend over time. Nitrate also showed a decreasing trend with increased depth. Increased solute concentration of leachate would reduce anion exchange capacity of the subsoil as shown by Bellini et al. (1996) which could account for this reduction of anions held on exchange sites (Figure 2). A wetter environment under the rock mulch would lead to greater leaching losses of N[O.sub.3] and Cl over time, which could also cause a reduction in anion concentration.
Management implications. The NRCS designed loafing lot performed well in terms of preventing muddy conditions. Very little sediment was collected in runoff samples and virtually no soil erosion was observed from the lot over the course of this study (data not shown). However, our results show the environmental impact of this practice could be quite negative. The water quality of runoff from NRCS designed loafing lots was poor due to high concentrations of N[O.sub.3], N[H.sub.4], and P[O.sub.4], as one might expect. Also, as one might expect, runoff was a relatively small part of the water balance (18 percent of precipitation) due to the improved infiltration/percolation characteristics of the gravel and geotextile surface material. In this design, to lower the risk of surface water pollution, runoff water could be routed to a lagoon or to a large vegetated filter strip. The more difficult problem is the large percentage of the water balance constituted by deep percolation and the poor quality of this water due to high N[O.sub.3]. Just as was found with conventional loafing lots (Drommerhausen et al., 1995), N[O.sub.3] leaching to ground water can be expected with the NRCS designed loafing lots. Since the loafing lots are usually located near milking barns where wells are common, the water in these wells is subject to contamination with N[O.sub.3]. Unless the NRCS designed loafing lots can be located in an area where ground water contamination is not a concern, concrete loafing areas or some other low permeability surface may be a preferred alternative, despite the higher costs.
Summary and Conclusion
We measured rainfall, runoff, and tile drainage from a newly installed NRCS designed loafing area over a two-year period. The water balance for the loafing lot indicated that approximately 18 percent of annual rainfall can be expected as runoff. The quality of this runoff water was impaired due to high concentrations of N[O.sub.3], N[H.sub.4], and P[O.sub.4] and should be routed to a lagoon or managed through a large vegetated filter strip to utilize nutrients. More than 30 percent of annual rainfall was estimated to move through the rock mulch and soil of the loafing lot toward ground water. Percolation water intercepted by the tile drains was high in nitrate (>37 mg [L.sup.-1]) and chloride (> 50 mg [L.sup.-1]). Water that became deep percolation likely had similar concentrations and will negatively impact ground water confirming our hypothesis that ground water can be significantly impaired with these loafing lot systems. In areas with a ground water resource, loafing lot designs should be modified by use of concrete, or by developing some low permeable surface to prevent adverse affects to ground water. Due to its poor quality, both surface and subsurface water that flows through the loafing lot should be captured to prevent direct movement into the environment.
Table 1. Soil profile description and physical properties for Pacolet soil series at loafing lot, Oglethorpe County, Georgia. Bulk Horizon Depth Sand Silt Clay density (cm) (%) (Mg [m.sup.-3]) A 0-18 65 23 12 1.45 Bt1 18-38 32 18 50 1.52 Bt2 38-76 23 9 68 1.32 BC 76-117 52 25 23 1.42 C 117+ 49 21 30 1.43 Field Wilting Horizon capacity point [K.sub.s] ([m.sup.3] [m.sup.-3]) ([m.sup.3] [m.sup.-3]) (cm [d.sup.-1]) A 0.25 0.08 10.2 Bt1 0.37 0.31 1.9 Bt2 0.38 0.27 0.3 BC 0.33 0.19 6.3 C 0.32 0.15 38.5 Table 2. Monthly totals of water flux for various pools of a water balance determined to impact the loafing lot, Oglethorpe County, Georgia. Year Month Precipitation ([double dagger]) Runoff ([double dagger]) mm 1997 Jan 118.0 [dagger] Feb 198.1 40.0 Mar 68.8 18.3 Apr 139.2 36.1 May 88.2 19.0 Jun 64.5 12.9 Jul 88.6 3.6 Aug 42.1 3.0 Sep 144.3 19.8 Oct 127.8 27.4 Nov 76.8 4.2 Dec 137.3 [dagger] 1998 Jan 129.6 [dagger] Feb 29.6 [dagger] Mar 87.1 26.3 Apr 112.6 3.7 May 174.5 37.2 Jun 78.9 23.3 Jul 61.4 0 Aug 36.6 4.6 Sep 43.9 0.1 Oct 80.8 0.9 Nov 39.6 0.3 Dec 45.9 [dagger] Totals 1753.8 310.7 Year Month Evaporation ([section]) Tile ([double dagger]) mm 1997 Jan 10.0 [dagger] Feb 11.9 [dagger] Mar 27.7 [dagger] Apr 29.0 [dagger] May 36.4 0.00 Jun 42.8 2.50 Jul 66.4 [dagger] Aug 65.1 [dagger] Sep 45.2 [dagger] Oct 31.1 [dagger] Nov 13.4 [dagger] Dec 10.0 [dagger] 1998 Jan 11.6 [dagger] Feb 12.7 0.27 Mar 18.8 0.46 Apr 24.4 0.28 May 48.0 0.85 Jun 54.5 0.50 Jul 52.3 0.09 Aug 55.0 0.09 Sep 44.8 0.03 Oct 35.1 0.00 Nov 16.6 0.00 Dec 10.3 [dagger] Totals 718.3 4.8 ([dagger]) Monitoring equipment was not available for some or all of this period. ([section]) Evaporation estimated using a modified version of the FAO-56 reference crop evapotranspiration equation from daily weather data obtained for Watkinsville, Georgia from the Georgia Automated Environmental Monitoring Network. ([double dagger]) Values estimated by local measurements and summed to monthly totals. Table 3. Concentration of soluble ions in surface water runoff from loafing lot, averaged on a monthly basis, Oglethorpe County, Georgia. Number of Date N[O.sub.3]-N N[H.sub.4]-N P[O.sub.4]-P CI samples mg [L.sup.1] Background 0.5 0.2 0.01 13.3 24 Sep-97 8.7* 2.3* 10.75* 24.9* 17 Oct-97 4.9* 1.4* 7.95* 24.6* 21 Nov-97 0.1 32.0* 6.09* 111.8* 12 Dec-97 2.9* 22.0* 10.45* 141.0* 22 Jan-98 8.8* 3.1* 12.83* 73.1* 9 Feb-98 1.0 12.1* 8.23* 61.9* 35 Mar-98 22.2* 10.2* 13.90* [dagger] 8 Apr-98 45.6* 2.3* 10.07* [dagger] 21 * Values significantly different from background using student's t test at 0.05 probability level. ([dagger]) Analysis not completed. Table 4. Concentration of soluble ions in tile drained water from loafing lot, Oglethorpe County, Georgia. Date N[O.sub.3]-N N[H.sub.4]-N P[O.sub.4]-P CI mg [L.sup.-1] Background 5.9 0.1 0.01 9.3 25-Sep-97 5.3 0.0 1.98* 50.2* 26-Sep-97 60.4* 0.0 0.25* 109.0* 18-Nov-97 46.0* 0.7 0.53* 95.3* 26-Nov-97 40.8* 3.8 1.34* 98.1* 6-Feb-98 40.5* 1.4 0.22* 52.4* 11-Feb-98 37.0* 0.9 0.67* 56.2* * Concentrations were significantly greater than the average of background samples using a student's t-test at the 0.05 probability level.
This research was supported in part by U.S. Environmental Protection Agency 319 funds in cooperation with the Georgia Soil and Water Conservation Commission. The authors thank Dr. Gerrit Hoogenboom and the Georgia Automated Environmental Monitoring Network for providing appropriate weather data, and thanks to Dr. Jay Ham for his assistance in estimating evaporative losses from the loafing lot.
Adriano, D.C., P.F. Pratt, and S.E. Bishop. 1971. Nitrate and salt in soils and ground waters from land disposal of dairy manure. Soil Science Society of America Proceedings 35:759-762.
Allen, R.G. 2000. Using the FAO-56 dual crop coefficient method over an irrigated region as part of an evapotranspiration intercomparison study. Journal of Hydrology 229:27-41.
Allen, R.G., L.S. Pereira, D. Raes, and M. Smith. 1998. Crop evapotranspiration: Guidelines for computing crop requirements. Irrigation and Drainage Paper No. 56. FAO, Rome, Italy, 300 pp.
Bellini, G., M. E. Sumner, D. E. Radcliffe, and N.P. Qafoku. 1996. Anion transport through columns of highly weathered acid soil: Adsorption and retardation. Soil Science Society of America Journal 60:132-137.
Brakensiek, D.L., H.B. Osborn, and W.J. Rawls. 1979. Field manual for research in agricultural hydrology. U.S. Department of Agriculture, Agricultural Handbook No. 224.
Bruce, R.R., J.H. Dane, V.L. Quisenberry, N.L. Powell, and A.W. Thomas. 1983. Physical characteristics of soils in the Southern Region: Cecil Southern Cooperative Series Bulletin 267. Georgia Agricultural Experiment Station Athens, Georgia.
Burkholder, J.M., E.J. Noga, C.H. Hobbs, and H.B. Glasgow, Jr. 1992. New "phantom" dinoflagellate is the causative agent of major estuarine fish kills. Nature. 35:407-410.
Cotlove, E., H. V. Trantham, and R. L. Bowman. 1958. An instrument and method for automatic, rapid, accurate and sensitive titration of chloride in biologic samples. Journal of Laboratory and Clinical Medicine 51:461-468.
Crook, W.M. and W.E. Simpson. 1971. Determination of ammonium on kjeldahl digests of crops by an automated procedure. Journal of the Science of Food and Agriculture 22:9-10.
Dane, J.H. and J.W. Hopmans. 2002a. Pressure cell. Pp. 684-687. In: W.A. Dick (ed.) Methods of soil analysis. Part 4 Physical methods. Soil Science Society of America Madison, Wisconsin.
Dane, J.H. and J.W. Hopmans. 2002b. Pressure plate extractor. Pp. 688-689. In: W.A. Dick (ed.) Methods of soil analysis. Part 4 Physical methods. Soil Science Society of America. Madison, Wisconsin.
Drommerhausen, D.J., D.E. Radcliffe, D.E. Brune, and H.D. Gunter. 1995. Electromagnetic conductivity surveys of dairies for groundwater nitrate. Journal of Environmental Quality 24:1083-1091.
Forster, D.L. 1998. Economic Issues in Animal Waste Management. Pp. 37-48. In: Animal Waste Utilization: Effective Use of Manure as a Soil Resource. J.L. Hatfield and B.A. Stewart (eds.). Ann Arbor Press, 121 South Main Street, Chelsea, Michigan.
Gould, M.C. 1993. A summary of nitrate levels in well water samples. Water Quality Courier: A Newsletter on Water Quality Issues 2:3-4. University of Georgia Cooperative Extension Service. Athens, Georgia.
Granzin, B.C. and J.B. Gaughan. 2002. The effect of sodium chloride supplementation on the milk production of grazing Holstein Friesian cows during summer and autumn in a humid sub-tropical environment. Animal Feed Science and Technology 96:147-160.
Ham, J.M. and J.L. Heilman. 1991. Aerodynamic and surface resistances affecting energy transport in a sparse crop. Agricultural and Forest Meteorology 53:267-284.
Hoogenboom, G. 2001. Weather monitoring for management of water resources. Pp. 778-781. In: K. J. Hatcher (ed.) Proceedings of the 2001 Georgia Water Resources Conference. Institute of Ecology. The University of Georgia, Athens, Georgia (ISBN 0-935835-07-5).
Keeney, D.R. and D.W. Nelson. 1982. Nitrogen--inorganic forms. Pp. 643-689. In: A.L. Page, R.H. Miller, and D.R. Keeney (eds.), Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. 2nd edition American Society of Agronomy, Madison, Wisconsin.
LeGrand, H.E. 1988. Piedmont and Blue Ridge. Pp. 201-208. In: W. Back et al. (ed.) Hydrogeology. Geological Society of America. Boulder, Colorado.
Murphy, J. and J.P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Analytica Chimica Acta 27:31-36.
Overcash, M. R. and R. L. Phillips. 1978. Dairy feedlot hydrology. Trans. American Society of Agricultural Engineers 21:1193-1198.
Rabalais, N.N., W.J. Wiseman. Jr., R.E. Turner, D. Justic, B.K.S. Gupta, and Q. Dortch. 1996. Nutrient changes in the Mississippi River and system responses on the adjacent continental shelf. Estuaries 19:386-407.
Radcliffe, D.E. and L. T. West. 2000. Major land resource area 136: Southern Piedmont. In: D.D. Scott (ed.). Water and chemical transport in soils of the Southeastern U.S.A. Southern Cooperative States Bulletin #395 [Online]. Available at http://soilphysics.okstate.edu/S257/cover.html.
Reynolds, W.D., D.E. Elrick, E.G. Youngs, H.W.G. Booltink, and J. Bouma. 2002. Laboratory methods. Pp. 802-815. In: W.A. Dick (ed.) Methods of soil analysis. Part 4 Physical methods. Soil Science Society of America. Madison, Wisconsin.
Ruhl, S., J. Overmoyer, D. Barker, and L.C. Brown. 2003. Using geotextile fabric in livestock operations. Ohio State University Fact Sheet [Online]. Available at http://ohioline.osu.edu/aex-fact/0304.html.
Sharpley, A.N., S.C. Chapra, R. Wedepohl, J.T. Sims, T.C. Daniel, and K.R. Reddy. 1994. Managing agricultural phosphorus for protection of surface waters: issues and options. Journal of Environmental Quality 23:437-451.
Spalding, R.F. and M.E. Exner. 1993. Occurrence of nitrate in groundwater: A review. Journal of Environmental Quality 22:392-402.
U.S. Department of Agriculture--Natural Resources Conservation Service (USDA-NRCS). 1969. Hydrologic Soil-Cover Complexes. Chapter 9. In: National Engineering Handbook. USDA-NRCS. Washington, D.C.
U.S. Department of Agriculture--Natural Resources Conservation Service (USDA-NRCS). 2003. Heavy use area protection. Code 561. Electronic field office technical guide [Online]. Available at http://www.nrcs.usda.gov/technical/efotg.
van Genuchten, M.T. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Science Society of America Journal 44:892-898.
van Vuuren, A.M. and J.A.C. Meijs. 1987. Effects of herbage composition and supplement feeding on the excretion of nitrogen in dung and urine by grazing dairy cows. Pp. 17-25. In: H.G. van Der Meer, R.J. Unwin. R.A. van Dijk, and G.C. Ennik (eds). Animal Manure on Grassland and Fodder Crops. Fertilizer or Waste? Martinus Nijhoff Publishers. The Netherlands.
Kent A. McVay is an assistant professor in the Department of Agronomy at Kansas State University in Manhattan, Kansas. David E. Radcliffe and Miguel L. Cabrera are both professors in the Department of Crop and Soil Science at the University of Georgia in Athens. Georgia. Gerrit Hoogenboom is a professor in the Department of Biological and Agricultural Engineering at the University of Georgia in Athens, Georgia.