Distribution and mobility of selenium and other trace elements in

Jun 1, 1988 - Trace Element Geochemistry in Water from Selected Springs in Death Valley National Park, California. David K. Kreamer , Vernon F. Hodge ...
1 downloads 0 Views 2MB Size
Download PDF
Environ. Scl. Techno/. 1000, 22, 697-702

Distribution and Mobility of Selenium and Other Trace Elements in Shallow Groundwater of the Western San Joaquin Valley, California Steven J. Deverel" and Steven P. Mlllard

Water Resources Division, U.S. Geological Survey, 2800 Cottage Way, Federal Buildlng, Room W-2234, Sacramento, Callfornia 95825

rn Samples of shallow groundwater that underlies much of the irrigated area in the western San Joaquin Valley, CA, were analyzed for various major ions and trace elements, including selenium. Concentrations of the major ions generally were similar for groundwater collected in the two primary geologic zones-the alluvial fan and basin trough. Selenium concentrations are significantly (a = 0.05) higher in the groundwater of the alluvial-fan zone than in that of the basin-trough zone. The concentrations of oxyanion trace elements were significantly correlated (a= 0.05) with groundwater salinity, but the correlations between selenium and salinity and between molybdenum and salinity were significantly different (a= 0.05) in the alluvial-fan geologic zone compared with those in the basin-trough geologic zone. The evidence suggests that the main factors affecting selenium concentrations in the shallow groundwater are the degree of groundwater salinity and the geologic source of the alluvial soil material.

Introduction Selenium in agricultural drain water has caused a high incidence of deformity and mortality of waterfowl at Kesterson National Wildlife Refuge in the western San Joaquin Valley, CA (1, 2). Subsurface drain water was transported to Kesterson from artificially drained agricultural fields in the westerfi San Joaquin Valley by the San Luis Drain, a concrete-lined conveyance canal. Agricultural drainage systems that are installed within lowlying areas of the western valley maintain water tables below depths (about 1.5 m) where evaporative concentration of solutes can occur in the shallow groundwater and unsaturated zone. Because of shallow groundwater, about 100000 ha of land in the western San Joaquin Valley requires artificial drainage to maintain long-term agricultural productivity (3). High groundwater selenium concentrations generally are in areas where saline soils and groundwater are common (4). A better understanding of the origin, distribution, and mobility of selenium and its relation to the chemical composition of the shallow groundwater in this agricultural area is essential for effective water management. This paper describes the resulta of a study to assess the relation of selenium concentrations and overall chemical composition of shallow groundwater to the two primary geologic zones of the western San Joaquin Valley. As part of this assessment, the interrelations among trace elements also were evaluated. This study is part of a comprehensive investigation of the hydtology and geochemistry of the San Joaquin Valley by the U.S.Geological Survey. The studies are being done in cooperation with the San Joaquin Valley Drainage Program and as part of the Regional Aquifer System Analysis Program of the U.S.Geological Survey. Background and Approach Deverel and others (4) evaluated differences in the chemistry of shallow groundwater among three physiographic zones in the western valley, which are defined

primarily by soil and topographic characteristics: the alluvial-fan zone adjoining the Coast Ranges, the basintrough zone adjoining the San Joaquin River, and the basin-rim zone in between the two (5). The alluvial-fan physiographic zone, with moderately sloping terrain and mostly low-salinity soils, solely consists of material eroded from the Coast Ranges. The basin-trough physiographic zone, which lies closest to the San Joaquin River, is separated from the alluvial-fan zone by the basin-rim zone. The basin-trough and basin-rim zones overlie a mixture of Coast Range and Sierra Nevada alluvium. Soils of the basin-rim zone are moderately to highly saline, whereas the soils of the basin trough are of low salinity (5, 6). The investigation by Deverel and others ( 4 ) indicated that shallow groundwater in the basin-trough physiographic zone had significantly lower selenium concentrations than in the alluvial-fan and basin-rim zones. Though selenium concentrations did not differ significantly between the alluvial-fan and basin-rim zones, the correlation between selenium concentrations and salinity was different. Selenium concentrations and salinity as measured by specific conductance were significantly correlated (a = 0.05) in the alluvial-fan zone. In the basin-rim physiographic zone, saline soils are more common than in the alluvial-fan zone (6),yet the correlation is not statistically significant. The difference in the correlation of salinity and selenium concentrations may be due to the variable geologic origin of deposits in the basin-rim zone. The approach taken in this paper was to evaluate the effect of geologic source material on groundwater chemistry by analyzing the interrelations among specific conductance, pH, major ions, and trace elements in relation to the geologic origin of soil materials. To facilitate this analysis, the study area was divided into the two primary geologic zones: the alluvial-fan and basin-trough zones (7,8). This division was selected rather than the physiographic divisions because it results in a more even distribution of samples between the zones and links the samples more closely to the geologic origin of the soil materials.

Geohydrologic Setting California's San Joaquin Valley is a trough between the Sierra Nevada on the east and the folded and faulted Coast Ranges on the west (Figure 1). The Sierra Nevada, a fault block that dips southwestward, is composed primarily of igneous and metamorphic rock of pre-Tertiary age. The Diablo Range of the Coast Ranges borders the study area to the west and consists of an exposed Cretaceous and Late Jurassic core assemblage of marine origin overlain and juxtapositioned with Cretaceous marine and Tertiary marine and continental deposits. The deposits of the alluvial-fan geologic zone on the westernmost edge of the San Joaquin Valley (Figure 1) were derived exclusively from Diablo Range rocks. The alluvial fans were laid down by intermittent streams that drain the Coast Ranges. In the lower lying basin-trough geologic zone, continental stream-laid and still-water deposita are Pleistocene and Holocene. These deposits are

Not subject to U.S. Copyright. Published 1988 by the American Chemical Society

Envlron. Sci. Technol., Vol. 22, No. 6, 1988 697

Typically, about 1m of water is applied annually by flood or furrow irrigation. This is the primary source of groundwater recharge.

Flgure 1. Location of study area, boundaries of geologic zones, and distribution of sampling sites.

of mixed Coast Range and Sierra Nevada origin (9). Saline soil and groundwater are common in the alluvial-fan and basin-trough geologic zones (6). The presence of saline soil and shallow groundwater within 3 m of land surface is due to a combination of hydrologic and climatic factors. Downward movement of irrigation water is impeded by soils of low permeability, and low-lying park of the western San Joaquin Valley are natural groundwater discharge areas (IO). These factors result in extensive areas with shallow groundwater. Evapotranspiration losses of 780-1040 mm/annum (11) and low precipitation of less than 250 mm/annum in most of the valley historically caused saline conditions due to evaporation of the shallow groundwater. Groundwater movement in the shallow aquifer is affected by irrigation and drainage practices. 698

Environ. Sci. Technol.. Vol. 22, No. 6,1988

Sampling and Analytical Methods Field. Samples of shallow groundwater were collected at 118 sites in the western San Joaquin Valley during an 18-dayperiod in the spring of 1984. Thirty-five sites were farm drain sumps. A drain sump collects subsurface drain water from a single farm drainage system that typically drains less than 250 ha and consists of a grid of buried permeable tile or perforated plastic pipe. Eighty-three sites were observation wells with poly(viny1 chloride) casings ranging in depth from 3 to 10 m. Sixty-eight sites in the alluvial-fan geologic zone and 50 sites in the hasin-trough geologic zone were sampled. An additional 65 groundwater samples were collected in the alluvial-fan geologic zone during a 4-week period in the spring of 1985. These samples were collected from Observationwells installed in three agricultural fields that are within 15 km of each other. This sampling is part of an ongoing study of the processes that affect mobility of selenium and other inorganic constituents in shallow groundwater associated with drained agricultural lands. Water samples were collected from sumps and collector drains with a peristaltic pump to remove water from the collecting structure, usually a concrete cistern, through Teflon tubing into a water-sediment churn splitter, and then into sample containers. Water samples were collected from the observation wells by using tubing and positivedisplacement bladder pumps constructed of Teflon. The submersible, 1.91 cm diameter bladder pump operates by allowing water to enter through a one-way valve in the bottom of the pump. Pressurized nitrogen gas is cycled in and out of a Teflon bladder inside the pump casing, displacing the water up the tubing. Prior to sampling, each well was pumped with peristaltic and bladder pumps simultaneously until the specific conductance of the water did not vary by more than 10% for three consecutive well-casing volumes of water. Before each sample was taken, another well-casing volume was pumped with the bladder pump only. Prior to sampling at each site, all sampling apparatus and containers were rinsed 3 times with the water to be sampled. Well water was used to rinse filters, filter holden, and sample containers. Prior to collection of samples for laboratory analysis of inorganic constituents, unfiltered 50-mL samples were titrated incrementally with dilute sulfuric acid for bicarbonate and carbonate concentrations. The pH was measured with portable meters that were calibrated at each site with standardized solutions. Temperature was measured with mercury thermometers that were checked against a standardized laboratory thermometer. Specific conductance was measured with a meter calibrated with a KC1 standard within 2000 pS/cm of the sample. Samples for analysis of dissolved constituents were pumped through 0.45-pm membrane filters with a peristaltic pump. Samples for determination of major ions and all trace elements were collected and stored in polyethylene bottles with Teflon-lined caps. All samples were acidified with nitric acid to a pH of less than 2 except those collected for determination of specific conductance, pH, bicarbonate, carbonate, sulfate, and chloride. Laboratory. Calcium, magnesium, sodium, potassium, lithium, iron, and zinc were determined by atomic absorption spectrometric methods (12). Cadmium, chromium, copper, lead, manganese, and molybdenum were determined by atomic absorption spectrometric methods

I EXPLANATION + MEDUX GEOLOGIC ZONE B i s bssin frough A is s l l ~ v i a laen l

B

A

-0

B

A

G E O L O G I C

+

Z O N E

EXPLANATION MEDLAN

GEOLOGIC ZONE B is basin trough A i s a l h ~ i a fan l

B

Figure 3. Ranges of v a b s for selected bace element mncanfratkas.

A G E O L O G I C

Z O N E

specnlc conductance. pH, and maw

between zero and the detection limit. For principal component analysis, nondetections were assigned a random value between zero and the detection limit.

with chelation extractions (12). Chloride and vanadium were determined by automated colorimetric methods, and sulfate was determined by the turbidimetric procedure. Arsenic and selenium were determined by hydride generation and atomic absorption spectrometry (13). The method used for selenium is designed to determine the totalconcentration of all forms of selenium present in the water sample. A sample is first subjected to an oxidative digestion to release any selenium from the organic fraction. The selenium released by this digestion, together with the inorganic selenium originally present, then is reduced to the selenite form hy a stannous chloridepotassium iodide mixture. The selenium hydride is generated by reducing the selenite form with sodium borohydride. The hydride gas is stripped from the solution by a stream of nitrogen gas, and ita concentration is determined by atomic ahsorption spectrometry. For regression analysis and nonparametric statistical testa, analytical nondetections were set to a value midway

Chemical Composition of Shallow Groundwater in Alluvial-Fan and Basin-Trough Geologic Zones The major ion chemistry of shallow groundwater is generally similar in the alluvial-fan and basin-trough geologic zones (Figure 2). Sulfate and sodium were the dominant anion and cation in most samples. Salinity, as measured by specific conductance, and concentrations of sulfate and sodium tend to be greater in the basin-trough zone, hut the Wilcoxon rank sum test (14)did not indicate any significant differences (a= 0.05) between zones. Concentrations of most trace elements generally are similar in the two geologic zones (Figure 3). However, the Wilcoxon rank sum test indicated that chromium and selenium are significantly (a = 0.05) greater in the alluvial-fan zone, whereas manganese is greater in the basintrough zone. The reason for the chromium and manganese differences are not apparent from the existing data and are not discussed further. The focus of this paper is on

Figure 2. Ranges of values for Ion concentrations.

Environ.

Sci. Technol., VoI.

22. NO. 6, 1988 6S8

Table I. Correlations of Specific Conductance, pH, and Major Ions with the First Three Principal Components principal component number variable physical properties specific conductance PH major ions calcium magnesium sodium potassium bicarbonate plus carbonate sulfate chloride total variance. %

I

I1

I11

-0.94 -0.16

0.11 0.58

-0.05 -0.79

-0.29 -0.95 -0.98 -0.70 0.06 -0.98 -0.95

0.67 -0.14 -0.01 -0.31 -0.75 0.04 -0.14

0.27 0.06 -0.03 -0.45 -0.39 0.00 0.08

58

17

9

Table 11. Correlations of Trace Elements with the First Three Principal Components principal component number variable

I

I1

I11

arsenic boron chromium iron lithium manganese molybdenum selenium vanadium zinc total variance. %

0.09 -0.91 -0.51 0.10 -0.60 0.29 -0.82 -0.79 -0.72 -0.05

-0.47 -0.09 0.22 -0.71 0.11 -0.65 -0.14 -0.08 -0.14 -0.62

-0.71 0.04 -0.15 -0.47 -0.20 0.22 0.07 0.05 0.01 -0.40

35

15

10

the reasons for the differences between selenium concentrations in the two geologic zones.

Geochemical Interrelations and Trace Element Mobility Principal component analysis (15)was used to examine the interrelations among major ions and physical properties and separately among trace elements. Specific conductance, pH, and major ions were evaluated as one group of variables, and trace elements were separately evaluated as a second group. For both groups of variables, the principal component analysis was based on the correlation matrix (variance-covariance matrix of the standardized variables). For specific conductance, pH, and major ions, 58% of the total variance was explained by the first principal component (Table I), which is associated with specific conductance and the dominant major ions-magnesium, sodium, sulfate, and chloride. This indicates that most variability in this group of variables is associated with variation in groundwater salinity. For trace elements, 35% of the total variance is accounted for by the first principal component (Table 11), which is primarily associated with boron, molybdenum, selenium, and vanadium. The second principal component accounts for only 15% of the total variance and is dominated by iron, manganese, and zinc, which are usually associated with mineral phases. Boron, molybdenum, selenium, and vanadium are probably present as dissolved oxyanions in the oxidized, alkaline water of the western San Joaquin Valley and, thus, are likely to be associated with salinity. Evidence of this is provided by an examination of the literature, speciation data, and correlations 700

Envlron. Scl. Technol., Vol. 22, No. 6, 1988

of these elements with salinity. The most oxidized form of selenium is selenate (+6 valence), which is relatively mobile in aqueous environments and does not associate with most solid-phase materials (16-18). Samples collected at 18 wells and drain sumps included in this study were analyzed for selenate and selenite (+4 valence) by Dr. Gregory Cutter at Old Dominion University, Norfolk, VA. Selenate represented an average of 98% of the dissolved selenium concentration in the samples. In addition, Fujii and Deverel (19) reported that (1) the selenium in soil solutions and shallow groundwater is in the selenate form and (2) a very small percentage of soil selenium is in the absorbed phase. Selenium has been reported to Qccurextensively in association with reduced metal sulfide minerals in marine sedimentary rocks (20, 21), which are similar to Coast Range rocks. Low Concentrations of selenium generally were reported for igneous rocks such as those of the Sierra Nevada (22). Boron, molybdenum, and vanadium are geochemically mobile and present as oxyanions in oxidized, alkaline environments such as the western San Joaquin Valley shallow groundwater (23-25,28). These elements are present in approximately equal amounts in both sedimentary and igneous rocks (23-25,28,30). Boron has been reported in concentrations as high as 15 000 wg/L in surface water that drain the Coast Ranges (27) and in concentrations of several hundred micrograms per liter in surface water that drain the Sierra Nevada. Molybdenum and vanadium have been reported at concentrations of several micrograms per liter in surface water that drain both mountain ranges (26,27, 29). Although boron, molybdenum, selenium, and vanadium are of relatively similar geochemical mobility in oxidized and alkaline aqueous environments, selenium seems to be unique in its mineralogical origin. Selenium originates primarily from sulfide minerals present in sedimentary rocks, which are predominant in the Coast Ranges (20-22). The results of principal component analysis for specific conductance, pH, and major ions are similar to the results for the trace elements. In both cases, a large part of the variance is accounted for by mobile and conservative species. Tidball and others (31)reported analogous results for factor analysis of 721 analyses of soil samples collected in the western valley. Selenium and sulfur (the major anionic component of soil salinity) were highly correlated with the same factor. Scores for this factor were highest in areas where groundwater salinity and selenium are high. Results of Tidball and others (31) and of this study demonstrate the association of mobile oxyanions and salinity in soils and shallow groundwater. This association is discussed further in the next section.

Correlation of Salinity and Trace Elements' Regression analysis of mobile trace element concentrations and groundwater salinity (as determined by specific conductance) in the two geologic zones was used to further elucidate the processes controlling the concentrations and distribution of inorganic trace elements for the data presented in Deverel and others (4). Because the trace element concentrations and specific conductance were log-normal distributed, the natural logs (log,) of the concentrations and specific conductance were used for regression analysis (Table 111). The correlation coefficients and regression equations are presented in Table I11 for the two geologic zones. All correlation coefficients were significant at a = 0.05. The regressions of salinity with boron, molybdenum, selenium, and vanadium concentrations were compared by

Table 111. Correlation Coefficients and Regression Equations for Natural Logarithms of Specific Conductance versus Natural Logarithms of Concentrations of Boron, Molybdenum, Selenium, and Vanadium"

correlation log, [boron] vs log, specific conductance log, [molybdenum] vs log, specific conductanceC log, selenium vs log, specific conductanceC log, vanadium vs log, specific conductance

correlation coefficient basintrough alluvial-fan geologic geologic zone zone 0.90 0.58 0.79

0.93 0.85 0.44

0.62

0.75

regression equationb alluvial-fan geologic zone basin-trough geologic zone slope intercept slope intercept 1.37 f 0.11 1.03 & 0.18 2.10 f 0.44 0.92 f 0.13

-3.2 -5.9 -14.4 -5.3

f 2.1 f 8.6 f 30.2 f 6.3

1.34 1.37 0.93 0.76

f 0.11 f 0.17 f 0.27 f 0.07

-3.2 -7.8 -6.5 -3.8

f 2.1 f 8.4 f 15.3 f 3.2

aData collected in 1984 and reported in Deverel and others ( 4 ) . Number of samples: 68 samples, alluvial-fan geologic zone; 50 samples, basin-trough geologic zone. Form of equation: log, y = a + b (log, x ) , where y is the constituent, a is the intercept, b is the slope, x is specific conductance. Confidence intervals are based on a = 0.05. Equations are significantly (a = 0.05) different between the two geologic zones as determined by analysis of covariance.

analysis of covariance (32). Only the selenium and salinity and molybdenum and salinity regressions are significantly different at the 95% confidence level between the two zones. The difference in the correlation of these two mobile trace element concentrations and salinity is related primarily to the geologic origin of selenium and molybdenum in the San Joaquin Valley. As discussed previously, selenium is often present in sedimentary marine deposits in association with, or oxidized from, metal sulfides from which the deposits of the alluvial-fan zone were derived (22). The plot of selenium concentrations and specific conductance (Figure 4) shows a scatter of observations for the basin-trough zone but a linear correlation for the alluvial-fan zone. The difference in relations between selenium and salinity in the two geologic zones is even more evident when data from sites near the geologic boundary are examined more closely. Three samples collected in the mapped area of the alluvial-fan geologic zone, but near the geologic boundary, had selenium concentrations that were less than 1pg/L, yet had relatively high specific-conductancevalues of 2300, 2830, and 3860 HS/cm. Two samples were collected in the central part (near Los Banos) and one in the northern part (near Tracy) of the study area. Removal of these outlying observations from the regression analysis resulted in a larger part of the variance (r = 0.83) in selenium concentrations being explained by variations of specific conductance. Results were analyzed separately for all sites within 1.6 km of the geologic boundary in the basin-trough zone. Shallow groundwater at these sites could have originated in the alluvial-fan zone, which is upgradient. Of the 16 sampling sites, two sites had large specific-conductance values (9200 and 30 400 pS/cm) but concentrations of selenium less than 1 pg/L. Regression analysis of the remaining 14 sites resulted in a correlation coefficient of 0.95 and a regression relation having a slope not significantly different from that for the data from the alluvial-fan zone. Regression analysis of the 34 observations from sites in the basin-trough zone and more than 1.6 km from the alluvial-fan zone shows even less evidence of any correlation between selenium and salinity than the analysis of all basin-trough observations (r = 0.17). The lack of correlation of selenium and specific conductance in the basin-trough zone is probably because of the mixture of Sierra Nevada and Coast Range deposits in the basin-trough geologic zone. In fact, Sierra Nevada deposits dominate. The enrichment in deuterium and oxygen-18 of saline groundwater samples collected regionally and in agricultural fields (33) demonstrates that the highest groundwater salinity is the result of evapotranspiration from a shallow water table (within 1.5 m of

ALLUVIAL-FAN GEOLOGIC ZONE

10

r=0.79

I

1

loge Se = 2.10 x loge SC-14.4 A

5

2

0

rl

z vj

z

P

t E E

-5

x

I I 8 10 12 LOGE SPECIFIC CONDUCTANCE, IN 1 S I C M

0

BASIN-TROUGH GEOLOGIC ZONE

v)

Y A

aa

n

-

0

8 10 12 LOGE SPECIFIC CONDUCTANCE, IN p S I C M

6

Figure 4. Natural logarithms of selenium concentrations versus natural logarithms of specific conductance for two geologic zones.

land surface). Though partially evaporated saline groundwater is present throughout the low-lying areas of the western valley, high selenium concentrations seem to coincide only with the combination of high salinity and Coast Range alluvial deposits. The correlation of selenium and salinity for groundwater of the Coast Range alluvium was tested further with selenium and specific-conductance data for the shallow groundwater of three agricultural fields that are in the alluvial-fan geologic zone. In 1985, groundwater samples were collected and analyzed from 65 observation wells in the three fields according to methods described under Sampling and Analytical Methods. The relation between selenium and specific conductance for the 65 samples is Envlron. Scl. Technol., Vol. 22, No. 6, 1988

701

I

2

I

r:0.94 lOgeSe=2.43 x IogeSC-16.4

I .A

2

0

d

A

v)

A

u

u

9

I 8

O,

I

9

I 10

11

LOGE SPECIFIC CONDUCTANCE, IN fiS/CM

Figure 5. Natural logarithms of selenium concentratlons versus natural logarithms of specific conductance for data collected In 1985.

shown in Figure 5. The slope of the regression relation is not significantly different from the slope of the relation for the alluvial-fan data in Figure 4.

Summary and Conclusions The results of this study indicate that hydrologic processes that have resulted in high groundwater salinity in the low-lying areas of the western San Joaquin Valley also have resulted in high concentrations of mobile trace elements. The effects of geologic origin of the alluvial deposits and the distribution of groundwater salinity on the mobility and distribution of selenium and other mobile oxyanions have been described in this paper. This information provides a basis for evaluating and assessing sources of these elements in agricultural drain water and receiving water. The specific conclusions of this study are as follows: Much of the total variance in concentrations of 14 trace elements was associated with boron, molybdenum, selenium, and vanadium, which are mobile oxyanions in the oxidized and alkaline water of the western San Joaquin Va11ey. Concentrations of boron, molybdenum, and vanadium were significantly correlated (a= 0.05) with groundwater salinity in the alluvial-fan and basin-trough geologic zones. Boron and vanadium have similar relations to salinity in both geologic zones probably because they are present in the geologic source materials for both geologic zones in relatively equal amounts. Selenium is much more enriched in the Coast Range source materials of the alluvial-fan zone than those of the Sierra Nevada with the other oxyanions. Selenium is more strongly correlated with salinity and occurs in higher concentrations in the groundwater of the alluvial-fan geologic zone. Registry No. Se, 7782-49-2; Ca, 7440-70-2;Mg, 7439-95-4; Na, 7440-23-5; K, 7440-09-7;As, 7440-38-2;B, 7440-42-8;Cr, 7440-47-3; Fe, 7439-89-6; Li, 7439-93-2; Mn, 7439-96-5; Mo, 7439-98-7; V, 7440-62-2; Zn, 7440-66-6.

Literature Cited (1) U S . Bureau of Reclamation “Kesterson Reservoir and Waterfowl”; Information Bulletin 2; U.S. Government Printing Office: Washington, DC, 1984. (2) Presser, T. S.; Barnes, 1. Water-Resour.Invest. (U.S. Geol. Surv.) 1984, No. 84-4122. (3) U.S.Bureau of Reclamation “Drainage and Salt Disposal”; Information Bulletin 1; US. Government Printing Office: Washington, DC, 1984. 702

Environ. Scl. Technol., Vol. 22, No. 6, 1988

(4) Deverel, S. J.; Gilliom, R. J.; Fujii, R.; Izbicki, J. A.; Fields, J. C. Water-Resour. Invest. (U.S. Geol. Suru.) 1984, No. 84-4319. (5) California Department of Water Resources, “Land and Water-Use h p e & of the San Joaquin Valley Investigation”; memorandum report; State of California: Sacramento, CA, 1970; p 157. (6) Harradine, F. Soil Survey of Western Fresno County; University of California Press: Berkeley, CA, 1950. (7) Jenninge, C. W.; Strand, R. G. Santa Cruz Sheet, Geologic Map of California;California Division of Mines and Geology: Sacramento, CA, 1958. (8) Mathews, R. A.; Burnett, J. L. Fresno Sheet, Geologic Map of California;California Division of Mines and Geology: Sacramento, CA, 1965. (9) Miller, R. E.; Green, J. H.; Davis, G. H. Geol. Surv. Prof. Pap. (U.S.) 1971, No. 497-E. (10) Davis, G. H.; Poland, J. F. US.Geol. Surv. Water-Supply Pap. 1957, NO. 1360-6. (11) Williamson,A. K. Water-Resour.Invest. (U.S. Geol. Surv.) 1981, NO.81-45. (12) Techniques of Water-Resources Investigations of the U.S. Geological Survey; Skougstad, M. W., Fishman, M. J., Friedman, L. C., Erdmann, D. E., Duncan, S. S., Eds.; U.S. Geological Survey: Alexandria, VA, 1979; Book 5, Chapter

Al. (13) Fishman, M. J.; Bradford, W. L. Open-File Rep.-U.S. Geol. Suru. 1982, No. 82-272. (14) Hollander, M.; Wolfe, D. A. Nonparametric Statistical Methods; Wiley: New Yoik, 1973. (15) Johnson, R. A,; Wichern, D. W. Applied Multivariate Statistical Analysis; Prentice-Hall: Englewood Cliffs, NJ, 1982. (16) Leckie, J. 0.;Benjamin, M. M.; Hayes, K.; Kanjman, G.; Altman, S. Co-precipitation of Trace Elements from Water with Iron Oxyhydroxide;Electric Power Research Institute report CS-1513; Electric Power Research Institute: Palo Alto, CA, 1980; pp 4-1-4-18. (17) Frost, R. R.; Griffin, R. A. Soil Sci. SOC.A m . J. 1977, 41, 53-57. (18) Hingston, F. J.; Posner, A. M.; Quirk, J. P. J. Soil Sci. 1974, 25, 16-26. (19) Fujii, R.; Deverel, S. J. A S A Spec. Publ., in press. (20) Luttrell, G. W. U.S. Geol. Surv. Bull. 1959, No. 1019-M. (21) Gent, C. A. U.S. Geol. Surv. Bull. 1974, No. 1958-74. (22) Sindeeva, N. D. Mineralogy and Types of Deposits of Selenium and Tellurium;Wiley-Interscience: New York, 1964; 363 p. (23) Waskowiak, R. In Geochemistry ofBoron; Walker, C. T., Ed.; Dowder, Hutchinson and Ross: Stroudsburg, PA, 1975; pp 64-77. (24) Ellis, B. G.; Knezek, B. D.; Jacobs, L. W. In Chemical Mobility and Reactivity i n Soil Systems; Nelson, D. W., Elvick, D. E., Tanji, K. K., Eds.; Soil Society of America: Madison, WI, 1983; pp 109-122. (25) Harder, H. In Geochemistry of Boron; Walker, C. T., Ed.; Dowder, Hutchinson and Ross, Stroudsburg, PA, 1975; pp 83-104. (26) Mitten, H. T.; LeBlanc, R. A.; Bertoldi, G. D. Open-File Rep.-US. Geol. Surv. 1970. (27) Berkstresser, C. F. Open-File Rep.--US. Geol. Surv. 1968. (28) Hem, J. D. U.S. Geol. Sum. Water-Supply Pap. 1985, No. 2254. (29) Silvey, W. D. Open-File Rep.-U.S. Geol. Surv. 1971. (30) Sen, N.; Nockolds, S. R.; Allen, R. Geochim. Cosmochim. Acta 1959, 16, 58-78. (31) Tidball, R. R.; Severson, R. C.; Gent, C. A,; Riddile, G. 0. Open-File Rep.-U.S. Geol. Surv. 1986, No. 86-583. (32) Steel, R. G. D.; Torrie, J. H. Principles and Procedures of Statistics; McGraw-Hill: New York, 1960; 481 p. (33) Deverel, S. J.; Fujii, R. Water Resour. Res. 1988, 24(4), 516-524.

Received for review December 29, 1986. Revised manuscript received December 11, 1987. Accepted January 14,1988.