Published in Vadose Zone Journal 3:668-675 (2004)
© 2004 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
ORIGINAL RESEARCH
Anion Exchange in Saprolite
K. A. McVay*,a,
D. E. Radcliffeb,
L. T. Westb and
M. L. Cabrerab
a Dep. of Agronomy, Kansas State Univ., Manhattan, KS 66506
b 3111 Miller Plant Sciences Building, Univ. of Georgia, Athens, GA 30602
* Corresponding author (kmcvay{at}ksu.edu).
Received 2 July 2003.
 |
ABSTRACT
|
|---|
Many studies have shown that subsoils in the Piedmont region of the southeastern USA have anion exchange capacity (AEC) and can adsorb anions such as nitrate 
. Groundwater discharging to wells and streams often passes through thick layers of saprolite beneath Piedmont soils, but little is known about the AEC of saprolite. Our objective was to determine the AEC on saprolite samples collected with a split-spoon sampler during the installation of seven groundwater monitoring wells in north Georgia. Forty four samples were collected from depths ranging from approximately one to 15 m below the surface and analyzed for AEC, extractable iron, and pH. The saprolite thickness was at least 15 m at all seven sites as verified by drilling an adjacent 7.6-cm hole to bedrock. The average AEC for 37 saprolite samples was 0.80 cmolc kg–1. This was slightly less than half of the AEC observed in typical subsoil horizons of Piedmont soils. There was a significant positive correlation (r = 0.64) between saprolite AEC and iron oxide content. Taking into account the differences in thickness of the soil and saprolite, saprolite had about six times the AEC of subsoil. The expected adsorption coefficient (Kd) for NO–3 in saprolite was approximately 0.81 cm3 g–1, based on the relationship between AEC and Kd. Using an estimated bulk density and saturated water content based on texture, this corresponds to a retardation coefficient (R) of 4.4. Models and remediation efforts involving NO–3 in shallow Piedmont groundwater need to account for anion adsorption in saprolite.
Abbreviations: AEC, anion exchange capacity • BTCs, breakthrough curves • CV, coefficient of variation • DCB, dithionite–citrate–bicarbonate
|
INTRODUCTION
|
|---|
MANY STUDIES HAVE SHOWN that acid subsoils of the Ultisol, Oxisol, acid Alfisol, Andisol, and Inceptisol orders exhibit anion adsorption (Berg and Thomas, 1959; Sumner and Davidtz, 1965; Kinjo and Pratt, 1971; McMahon and Thomas, 1974; Black and Waring, 1976a, 1976b, 1976c; Chan et al., 1980; Nkedi-Kizza et al., 1984; Wong et al., 1987; Boggs and Adams, 1992; Ishiguro et al., 1992; Katou et al., 1996). Ultisols are common in the Piedmont region of the southeastern USA, and Grove et al. (1982), Gillman and Sumner (1987), Bellini et al. (1996), and Qafoku et al. (1999) have shown that the anion exchange capacity (AEC) of the unlimed B horizons of soils from this region are typically in the range of 1 to 2 cmolc kg–1.
Retardation coefficients in these soils for monovalent anions such as NO–3 and Cl– are typically 1.5 to 2.5 for solution concentrations less than 30 mmolc L–1 (Bellini et al., 1996; Gupte et al., 1996). Qafoku et al. (1999) studied 16 subsoils from the southeastern USA, South Africa, Australia, Indonesia, Japan, and Hawaii with variable charge mineralogy. They measured AEC and breakthrough curves (BTCs) on packed columns of subsoil using Cl– as a tracer in a 5 mM solution of CaCl2. The AEC ranged from 0.01 to 1.86 cmolc kg–1 with an average value of 0.68 cmolc kg–1. Retardation coefficients were determined from the BTC data and Kd values were calculated using the bulk densities and water contents of the columns. Adsorption coefficients ranged from 0.01 to 2.43 cm3 g–1 with an average value of 0.69 cm3 g–1. Retardation coefficients ranged from 1.1 to 6.2 with an average value of 2.5. For two Piedmont subsoils, the values of R were 3.1 and 4.8. They found a positive, linear relationship between Kd and AEC (r2 = 0.92) and between R and AEC (r2 = 0.93).
Piedmont region soils are commonly underlain by thick layers of saprolite, defined as isovolumetrically weathered bedrock (Pavich et al., 1989). Saprolite forms as the more soluble minerals weather from bedrock leaving behind a porous, brittle material that looks like the parent rock, but can be dug using hand tools. About 50% of the mapped soils in the Piedmont are underlain by light-colored (felsic) igneous and metamorphic rocks. Saprolite that forms from felsic rock is typically 15 to 30 m in thickness (Pavich et al., 1989). Common soil series formed from felsic parent material are Appling, Cecil, Helena, Louisa, Louisburg, Madison, Pacolet, Tallapoosa, Vance, and Wedowee (http://soils.usda.gov/technical/classification/osd/index.html; verified 20 Apr. 2004).
The shallow groundwater system in the Piedmont is an unconfined, two-layer aquifer composed of a zone of saprolite underlain by fractured bedrock (Fig. 1)
. Studies have shown that the saprolite layer is an integral part of the groundwater system and acts as an important water storage zone for the deeper fractures (Brackett et al., 1991, p. 1–160; Cressler et al., 1983; Radtke et al., 1986; Rose, 1992). The depth to the water table may be as shallow as 2 to 6 m in upland areas during the winter and the water table elevation generally follows the surface topography (LeGrand, 1988). Two types of wells are common in this region: shallow bored wells that are installed to bedrock and draw directly from the saprolite storage layer and deep drilled wells that penetrate the bedrock and draw indirectly from the saprolite layer via bedrock fractures. Recharge occurs throughout the uplands in this shallow groundwater system and travel times to streams and wells may be quite long. Rose (1992) measured tritium concentrations in water collected from Piedmont wells screened in the saprolite layer, streams sampled during baseflow (which would represent discharging groundwater), and springs. He estimated that at least some of this water was as old as 25 yr.
Evidence suggests that significant amounts of water pass through the saprolite, so it is important to quantify the effects of ion exchange on solute transport. The only related publication that we are aware of is that by Amoozegar et al. (1991). These authors measured Cl– and NO–3 movement through columns of saprolite from a single, shallow depth at two locations in North Carolina with felsic parent material. Retardation coefficients for NO–3 and Cl– were nearly the same with an average value of 2.2. Our objective was to determine the AEC of a more extensive collection of saprolite samples from the Piedmont of Georgia and to estimate anion adsorption and retardation coefficients for saprolite.
|
MATERIALS AND METHODS
|
|---|
Saprolite samples were collected at three dairies (Fig. 2)
in North Georgia during the installation of seven groundwater monitoring wells that were part of an earlier study (Drommerhausen et al., 1995). Saprolite samples were collected in November 1993 using a split-spoon hammer-driven sampler from the center of a hollow stem auger used for installing monitoring wells. A 30 cm by 8 cm diameter sample was taken every 1.5 m, at depths ranging from 1 to 15 m below the soil surface. Five of the wells were installed on "Dairy-6" described in Drommerhausen et al. (1995) (W-1 through W-5), one well was installed on "Dairy-1" in Morgan County (MO-2), and one well was installed on "Dairy-8" in Putnam County (PU-2). At each location, a separate 7.6-cm auger was used to probe for the depth to bedrock. The bedrock at Dairy-1 and Dairy-6 was mapped as biotite gneiss and the bedrock at Dairy-8 was mapped as granite gneiss (Anonymous, 1976).
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 2. Location of dairies in Georgia where saprolite samples were obtained for physical and chemical analysis.
|
|
Anion exchange was determined on air-dried and crushed (<2 mm) soil samples by the procedure of Gillman (1979) as modified by Sumner et al. (1994). Since temporate soils can have pH dependent charge, both positive and negative, it is important to measure exchange capacity under conditions that typically exist in the field. Gillman's procedure does this by buffering the soil solution with barium chloride and measuring exchange at very low ionic strength. Chloride was determined by coulometric titration (Cotlove et al., 1958). Extractable iron by dithionite–citrate–bicarbonate (DCB) extraction at 85°C was measured using the method of Mehra and Jackson (1960). Soil pH was determined using 1 M KCl and using 0.01 M CaCl2, in a 1:1 solution. Iron concentration was determined by atomic absorption spectroscopy using a PerkinElmer 5000 (Norwalk, CT).
To qualitatively evaluate potential differences in Fe mineralogy in the saprolite materials, Munsell color was determined from ground, air-dry samples (Schwertmann, 1993). Particle-size distribution was determined by pipette and sieving (Kilmer and Alexander, 1949). Mineralogy of the clay separates from PU-2 and W-4 was evaluated by X-ray diffraction using oriented mounts (Drever, 1973) and CuK
radiation. Treatments for mineral identification included Mg-saturation, air-dry; Mg-saturation, glycol solvated; K saturation, air-dry; and K saturation heated to 300 and 550°C.
Statistics were computed using SAS (SAS Inst., Cary NC). The General Linear Model procedure was used to test for significant site and depth effects in AEC. A natural log transform of the AEC was used due to the high coefficient of variation (145%). Means were compared using the Least Squares Mean procedure for pair-wise comparisons. Correlations were computed using the correlation procedure.
|
RESULTS
|
|---|
The depth to which saprolite samples were collected varied depending on the depth of the water table at the time the wells were installed. The intent of the original project (Drommerhausen et al., 1995) was to install the monitoring wells with the center of the 3-m screened section at the elevation of the water table. At all locations except W-1, holes were bored to a depth approximately 1.5 m below the apparent depth of the water table. These depths ranged from 5.6 to 8.4 m and probably represented the annual minimum water table elevation, since the wells were installed in late fall (the driest time of the year). At W-1, we took split-spoon samples as deep as our auger would allow (15 m). The depth to bedrock exceeded 15 m (the length of our longest auger) at all sites. Saprolite (BC or CB horizons) began at depths between 1 and 2 m below the surface at all sites. Hence, the thickness of the saprolite layer was at least 15 m at all sites.
The shallowest split-spoon sample was taken at a depth of 1.1 m, which was above the depth where saprolite started. We will refer to these samples as soil samples and samples from all other depths as saprolite samples. Saprolite samples varied considerably in appearance and color. Some samples were uniformly light colored or dark colored; other samples were striated (Fig. 3) . Color of the ground and sieved saprolite samples ranged from yellow (2.5 YR) to yellowish red (2.5 YR). There was no apparent relationship between color and saprolite AEC.
The average AEC of the seven soil samples was 1.98 cmolc kg–1 (Table 1), which was close to the values reported for two Piedmont B horizon soil samples (1.15 and 1.65 cmolc kg–1) by Qafoku et al. (1999). The overall average AEC for the 37 saprolite samples was 0.80 cmolc kg–1, about 40% of the soil sample AEC. The mean AEC for the shallowest sampling depth (1.1 m) was significantly higher statistically (p = 0.05) than samples from the 5.6-, 7.2-, and 8.7-m depth. Samples below a depth of 8.7 m were not considered in the statistical comparison because there were too few samples. There was a tendency for AEC to decrease with depth from 1.1 to 5.6 m, but below this depth AEC rose slightly, although there were fewer samples below 7.2 m. Anion exchange capacity was more variable in saprolite than in soil. The coefficient of variation (CV) for soil samples was 71% while the CV for saprolite samples was 143%. The average saprolite AEC for each site was similar (0.25–0.76 cmolc kg–1) except for PU-2, which had a statistically significant higher AEC (2.65 cmolc kg–1).
View this table:
[in this window]
[in a new window]
|
Table 1. Anion exchange capacity of saprolite samples from various depths collected during installation of seven monitoring wells.
|
|
Clay content of the saprolite samples varied between 65 and 501 g kg–1 with an average value of 103.1 g kg–1 (Table 2). Clay content was the variable most highly correlated with saprolite AEC (r = 0.67) (Table 3 and Fig 4
, bottom). Dithionite–citrate–bicarbonate Fe of the saprolite samples varied between 0.23 and 71.40 g kg–1 with an average value of 14.59 g kg–1. There was also positive correlation (r = 0.64) with saprolite AEC and DCB Fe (Table 3 and Fig. 4, top). There was a significant inverse correlation between sand content and saprolite AEC. Silt content varied between 6 and 364 g kg–1 with an average value of 141.6 g kg–1.
View this table:
[in this window]
[in a new window]
|
Table 2. Selected physical and chemical properties of saprolite samples from various depths collected during installation of seven monitoring wells.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 4. Anion exchange capacity (AEC) as a function of Fe content (top figure) and clay content (bottom figure).
|
|
Saprolite CaCl2 pH ranged between 3.70 and 5.19 with an average value of 4.18. The KCl pH ranged between 3.39 and 5.91 with an average value of 4.12. On average, KCl pH was 0.16 units below CaCl2 pH, indicating that the samples had a net positive charge. There were no significant correlations between saprolite AEC and either pH or the difference between pH values (Table 3).
X-ray diffraction patterns of samples from W-4 and PU-2 (Fig. 5 and 6)
indicated kaolinite was the dominant mineral in the clay separate of all samples. Kaolinite is confirmed by the presence of peaks at 0.72 and 0.357 nm that both disappear on heating to 550°C. Site W-4 contained minor amounts of mica at all depths (peaks at 1.0 and 0.33 nm) and small quantities of hydroxy interlayered vermiculite (peak at 1.4 nm) in samples from the upper two depths (Fig. 5). The sample from 8.7 m at W-4 appeared to also have a small quantity of halloysite as indicated by the small peak at 0.445 nm. Kaolinite was the only mineral identified in clay separates of samples from PU-2 (Fig. 6). At both sites, kaolinite peaks broadened and became less intense with depth suggesting crystallites were smaller and/or less crystalline at deeper depths (Moore and Reynolds, 1989).
View larger version (43K):
[in this window]
[in a new window]
|
Fig. 5. X-ray diffraction patterns for W-4 saprolite samples (depth for each pattern is shown along the left margin).
|
|
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 6. X-ray diffraction patterns for PU-2 saprolite samples (depth for each pattern is shown along the left margin).
|
|
|
DISCUSSION
|
|---|
The source of the AEC in saprolite appeared to be in the clay-size fraction, based on the high correlation between clay content and AEC (Table 3). Since the clay and iron oxide content both had similar correlation coefficients, this could be due primarily to iron oxides. However, the correlation between DCB Fe and clay content was not especially high (r = 0.59). This suggests that a portion of the iron oxides were coarser than clay size or that a component of the clay in addition to the iron oxides, probably kaolinite, was a source of AEC in these samples. Although visually there appeared to be a lot of mica particles in the silt and sand fractions, the low specific surface of these particles probably limited AEC from this source. Minor amounts of mica were present in the clay separate of W-4, but the contribution of mica edge-charge to AEC would be minor. Poor crystallinity or small crystallite size at deeper depths would tend to increase the contribution of kaolinite relative to iron oxides to the total AEC.
Saprolite had an AEC that was a little less than half of the value (1–2 cmolc kg–1) observed for B horizons of soils in the Piedmont (Grove et al., 1982; Gillman and Sumner, 1987; Bellini et al., 1996; Qafoku et al., 1999). In assessing the impact of AEC on anion transport in these two materials, it is critical to account for the differences in thickness. Subsoils at our sites were about 1 m in thickness, whereas saprolite was at least 15 m in thickness. Even though saprolite has a lower AEC, due to its greater thickness in this region of the Piedmont, it provides much greater anion retention than that of the soil above. Using NO–3 as an example, a 1-m thick subsoil with an AEC of 1.98 cmolc kg–1 (average for soil samples, Table 1) and bulk density of 1.5 g cm–3 could adsorb 9.2 Mg ha–1 of NO–3, if half of the anion adsorption sites were occupied by NO–3. By comparison, a 15-m thick saprolite layer with an AEC of 0.80 cmolc kg–1 (average for saprolite samples, Table 1) and the same bulk density of 1.5 g cm–3 could adsorb 55.8 Mg ha–1 of NO–3 (six times that of the subsoil).
Another way to look at this is in terms of adsorption (Kd) and retardation (R) coefficients and the resulting travel times for anions such as NO–3 and Cl. Using the relationship developed by Qafoku et al. (1999) between Kd and AEC for subtropical and tropical subsoils, including Piedmont subsoils, the Kd predicted for saprolite with an AEC of 0.80 cmolc kg–1 (the average) would be 0.81 cm3 g–1. Since saprolite and subsoil mineralogies are similar (Fig. 5 and 6), the relationship should apply to our saprolite samples. The predicted average saprolite Kd is larger than the average value (0.69 cm3 g–1) for subtropical and tropical subsoils found by Qafoku et al. (1999). However, it is less than the value found by Qafoku et al. (1999) for two Piedmont B horizon soil samples (0.99 and 1.74 cm3 g–1).
To convert Kd to R, we need estimates of saprolite bulk density and saturated water content. We did not measure bulk density of our samples but we can estimate bulk density based on texture. All but four of our 37 saprolite samples had a texture of loamy sand or sandy loam. The saturated water content estimated by Rosetta, a pedo-transfer function database developed by the U.S. Salinity Lab (Schaap, 1999), for these two texture classes is the same, 0.38 cm3 cm–3. Assuming a particle density of 2.65 g cm–3, this water content corresponds to a bulk density of 1.62 g cm–3. This value is similar to the range of bulk densities (1.51–1.83 g cm–3) found for three different saprolite sites in North Carolina (Vepraskas et al., 1991). Using these estimates of bulk density and saturated water content, the estimated R for the average saprolite sample is 4.4.
This is about twice the values reported by Amoozegar et al. (1991), who measured R on saprolite samples from two locations with felsic parent material in North Carolina. They found there was little difference in R for intact vs. packed columns and that NO–3 and Cl– had similar values: 2.1 and 2.3, respectively. The higher estimated R in our study is probably due to differences in column input concentrations used by Qafoku et al. (1999) and Amoozegar et al. (1991). Qafoku et al. (1999) (whose relationship we used to convert AEC to Kd) used a 5 mM concentration and Amoozegar et al. (1991) used a 30 mM concentration. Bellini et al. (1996) showed that as concentrations increased in the range 5 to 30 mM, adsorption of Cl– in a Piedmont subsoil decreased.
Our results indicate that dilute concentrations of anions such as NO–3 should have travel times from the subsoil through saprolite to a well or stream that are approximately four times that of water. Since saprolite layers are much thicker than subsoil layers, the effect of anion exchange in saprolite has more of an effect on travel times than in subsoil, which has been the focus of most research. Rose (1992) showed that travel times of water through saprolite to shallow wells and streams were on the order of years. Since anion exchange could cause a significant delay of a NO–3 pulse that passes through saprolite, anion exchange in saprolite needs to be taken into account in models that simulate NO–3 transport in Piedmont shallow groundwater. Our results also indicate that there may be a considerable delay in detecting NO–3 in a well or stream caused by land use in the uplands that produce excess N. Similarly, there may be a considerable lag between the cessation of practices in the uplands that produce excess N and a reduction in well or stream NO–3 concentrations.
We estimate delay in anion arrival to groundwater assuming flow through the entire thickness of saprolite. There is considerable evidence that bypass, or preferential flow occurs in saprolite and therefore the amount of delay is uncertain. In addition, bypass flow can occur in the B horizons of soil as well, so the relative impact on anion exchange would need to be adjusted based on evidence and measurements of bypass flow of both mediums. It would also stand to reason that the greatest concentration of iron oxides within the saprolite would be located in areas of preferential flow. These flow paths would be more highly weathered, which would cause greater amounts of Fe release and creation of more positive charge.
|
CONCLUSIONS
|
|---|
Our results show that anion exchange is an important characteristic of saprolite and needs to be considered in studies of NO–3 movement from soils through saprolite to wells and streams in the Piedmont region. The average saprolite AEC was 0.80 cmolc kg–1, which was a little less than half of the AEC typically observed in B horizons of Piedmont soils. The sources of the AEC appear to be iron oxides and edges of kaolinite minerals that are positively charged at the natural pH of saprolite (average of 4.18). In places where saprolite is much thicker (e.g., 15 times as thick) than the soil solum above it, there may be a great capacity to adsorb anions such as NO–3, retarding their movement from surface sources into groundwater.
|
REFERENCES
|
|---|
- Amoozegar, A., P.J. Schoeneberger, and M.J. Vepraskas. 1991. Characterization of soils and saprolites from the Piedmont region for waste disposal purposes. Rep. 255. Water Resources Research Inst. of the Univ. of North Carolina, Raleigh, NC.
- Anonymous. 1976. Geological map of Georgia. Georgia Dep. of Natural Resources, Atlanta, GA.
- 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 Sci. Soc. Am. J. 60:132–137.[Abstract/Free Full Text]
- Berg, W.A., and G.W. Thomas. 1959. Anion elution patterns from soils and soil clays. Soil Sci. Soc. Am. Proc. 23:348–350.
- Black, A.S., and S.A. Waring. 1976a. Nitrate leaching and adsorption in a Krasnozem from Redland Bay, Qld.: I. Leaching of banded ammonium nitrate in a horticultural rotation. Aust. J. Soil Res. 14:171–180.
- Black, A.S., and S.A. Waring. 1976b. Nitrate leaching and adsorption in a Krasnozem from Redland Bay, Qld.: II. Soil factors influencing adsorption. Aust. J. Soil Res. 14:181–188.
- Black, A.S., and S.A. Waring. 1976c. Nitrate leaching and adsorption in a Krasnozem from Redland Bay, Qld.: III. Effect of nitrate concentration on adsorption and movement in soil columns. Aust. J. Soil Res. 14:189–195.
- Boggs, M.J., and E.E. Adams. 1992. Field study of dispersion in a heterogeneous aquifer: IV. Investigation of adsorption and sampling bias. Water Resour. Res. 28:3325–3336.
- Brackett, D.A., W.M. Steele, T.J. Schmitt, R.L. Atkins, M.F. Kellam, and J.A. Lineback. 1991. Hydrogeologic data from selected sites in the Piedmont and Blue Ridge Provinces. Georgia Dep. of Natural Resources, Atlanta, GA.
- Chan, K.Y., H.R. Geering, and B.G. Davey. 1980. Movement of chloride in a soil with variable charge properties: I. Chloride systems. J. Environ. Qual. 9:579–582.[Abstract/Free Full Text]
- 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. J. Lab. Clin. Med. 51:461–468.[Medline]
- Cressler, C.W., C.J. Thurmond, and W.G. Hester. 1983. Ground water in the greater Atlanta region. Georgia Dep. of Natural Resources, Atlanta, GA.
- Drever, J.E. 1973. The preparation of oriented clay mineral specimens for X-ray diffraction analysis by a filter–membrane technique. Am. Mineral. 58:553–554.[ISI]
- Drommerhausen, D.J., D.E. Radcliffe, D.E. Brune, and H.D. Gunter. 1995. Electromagnetic conductivity surveys of dairies for groundwater nitrate. J. Environ. Qual. 24:1083–1091.[Abstract/Free Full Text]
- Gillman, G.P. 1979. A proposed method for the measurement of exchange properties of highly weathered soils. Aust. J. Soil Res. 17:129–139.
- Gillman, G.P., and M.E. Sumner. 1987. Surface charge characterization and soil solution composition of four soils from the Southern Piedmont in Georgia. Soil Sci. Soc. Am. J. 51:589–594.[Abstract/Free Full Text]
- Grove, J.H., C.S. Fowler, and M.E. Sumner. 1982. Determination of the charge character of selected acid soils. Soil Sci. Soc. Am. J. 46:32–38.[Abstract/Free Full Text]
- Gupte, S.M., D.E. Radcliffe, D.H. Franklin, L.T. West, E.W. Tollner, and P.F. Hendrix. 1996. Anion transport in a Piedmont Ultisol: II. Local-scale parameters. Soil Sci. Soc. Am. J. 60:762–770.[Abstract/Free Full Text]
- Ishiguro, M., K.-C. Song, and K. Yuita. 1992. Ion transport in an allophanic Andisol under the influence of variable charge. Soil Sci. Soc. Am. J. 56:1789–1793.[Abstract/Free Full Text]
- Katou, H., B.E. Clothier, and S.R. Green. 1996. Anion transport involving competitive adsorption during transient water flow in an Andisol. Soil Sci. Soc. Am. J. 60:1368–1375.[Abstract/Free Full Text]
- Kilmer, V.J., and L.T. Alexander. 1949. Methods of making mechanical analysis of soils. Soil Sci. 68:15–24.
- Kinjo, T., and P.F. Pratt. 1971. Nitrate adsorption: I. In some acid soils of Mexico and South America. Soil. Sci. Soc. Am. Proc. 35:722–725.
- LeGrand, H.E. 1988. Piedmont and Blue Ridge. p. 201–208. In W. Back et al. (ed.) Hydrogeology. Geol. Soc. of Am., Boulder, CO.
- McMahon, M.A., and G.W. Thomas. 1974. Chloride and titrated water flow in disturbed and undisturbed soil cores. Soil Sci. Soc. Am. Proc. 38:727–732.
- Mehra, O.P., and M.L. Jackson. 1960. Iron oxide removal from soils and clays by a dithionite–citrate system buffered with sodium bicarbonate. p. 317–342. In A. Swinefored (ed.) Clays Clay Miner., Proc. Conf. 7th, Washington, DC. 1958. Pergamon Press, Elmsford, NY.
- Moore, D.M., and R.C. Reynolds, Jr. 1989. X-ray diffraction and the identification and analysis of clay. Oxford Univ. Press, New York.
- Nkedi-Kizza, P., J.W. Biggar, H.W. Selim, M.Th. van Genuchten, P.J. Wierenga, J.M. Davidson, and D.R. Nielsen. 1984. On the equivalence of two conceptual models for describing ion exchange during transport through an aggregated Oxisol. Water Resour. Res. 20:1123–1130.
- Pavich, M.J., G.W. Leo, S.F. Obermeir, and J.R. Eastabrook. 1989. Investigations of the characteristics, origins, and residence time of the upland residual mantle of the Piedmont of Fairfax County, Virginia. USGS Prof. Pap. 1352. U.S. Gov. Print. Office, Washington, DC.
- Qafoku, N.P., M.E. Sumner, and D.E. Radcliffe. 1999. Anion transport in columns of variable charge subsoils: Nitrate and chloride. J. Environ. Qual. 29:484–493.
- Radtke, D.B., C.W. Cressler, H.A. Perlman, H.W. Blanchard, Jr., K.W. McFadden, and R. Brooks. 1986. Occurrence and availability of ground water in the Athens region, northeastern Georgia. U.S. Geol. Surv., Doraville, GA.
- Rose, S. 1992. Tritium in groundwater of the Georgia Piedmont: Implications for recharge and flow paths. Hydrol. Proc. 6:67–78.
- Schaap, M.G. 1999. Rossetta. Version 1.0 [Online]. Available at www.ussl.ars.usda.gov/models/rosetta/rosetta.htm (accessed 17 June 2003; verified 11 Mar. 2004). USDA-ARS George E. Brown, Jr. Salinity Lab, Riverside, CA.
- Schwertmann, U. 1993. Relations between iron oxides, soil color, and soil formation. p. 51–69. In J.M. Bigham and E.J. Ciolkosz (ed.) Soil color. SSSA Spec. Publ. 31. SSSA, Madison, WI.
- Sumner, M.E., and J.C. Davidtz. 1965. Positive and negative charges in some Natal soils. S. Afr. J. Agric. Sci. 8:1045–1050.
- Sumner, M.E., L. de Ramos, and U. Kukier. 1994. Compulsive exchange method for determining soil exchange capacities: Proposed time and labor savings modifications. Commun. Soil Sci. Plant Anal. 25:567–572.
- Vepraskas, M.J., M.T. Hoover, and J. Bouma. 1991. Sampling strategies for assessing hydraulic conductivity and water-conducting voids in saprolite. Soil Sci. Soc. Am. J. 55:165–170.[Abstract/Free Full Text]
- Wong, M.T.F., A. Wild, and A.S.R. Juo. 1987. Retarded leaching of nitrate measured in monolith lysimeters in south-east Nigeria. J. Soil Sci. 518:511–518.
TOP
ABSTRACT
INTRODUCTION
