- Email: [email protected]
Adsorption and movement of selected pesticides at high concentrations in soils
Wuler Research Vol. 13~ pp, 375 to 380
Pergamon Press l.td 1979. Printed in Great Britain
ADSORPTION AND MOVEMENT OF SELECTED PESTICIDES AT HIGH CONCENTRATIONS tN SOILS P. S. C. RAO and J. M. DAVIDSON Soil Science Department, University of Florida, Gainesville, FL 3261l, U.S.A.
(Received 29 June 1978) Abstract--Adsorption of 2,4-D amine, atrazine, terbacil and methyl parathion pesticides on Webster, Cecil and Eustis soils was measured at pesticide solution concentrations ranging from zero to the aqueous solubility limit of each pesticide. Measured equilibrium adsorption isotherms for nearly all soil-pesticide combinations were of nonlinear Freundlich type. The Freundlich adsorption constant (K) based on soil organic carbon was much less variable for a given pesticide among the four soils than was the K based on total soil mass. The influence of the shape of the adsorption isotherm on the movement of 2,4-D amine and atrazine through water-saturated soil columns Was also examined. Pesticide effluent concentrations from soil columns were measured at two input solution concentrations (50 and 5000/~gmi- t for Z4-D amine; 5 and 50/~g ml - ~ for atrazine). In all cases, pesticide mobility was significantly greater for the higher concentrations. Thus, serious errors may he introduced by assuming a linear adsorption isotherm (i.e. pesticide mobility is invariant with input concentration) when predicting pesticide transport from waste disposal sites where high concentrations exist.
Laboratory studies were initiated to investigate the physical, chemical and microbiological behavior of pesticides in soils when applied at high concentrations. The microbiological aspects of this study are reported elsewhere by Ou et aL (1978a, b). The influence of adsorption characteristics on pesticide mobility are considered in this paper.
INTRODUCTION
Because of a continued increase in the number and quantity of pesticide compounds being placed on the market, the safe disposal of surplus and/or waste pesticide materials has become an acute problem (yon Rumker et al, 1974). Incineration, encapsulation, isolation in underground caves and mines, chemical stabilization, land spreading and landfilling are some of the procedures being considered for the disposal of pesticides and other hazardous wastes (van Everdingen & Freeze, 1971; Schomaker, 1976). Of these methods, disposal by landfills and land spreading appear to be more common and economical (Fields & Lindsey,'1975; Lindsey et al, 1976). Placing hazardous waste in the land has come under severe attack recently (Rouston & Wiidung, 1969; Atkins, 1972) because it does not guarantee that the hazardous chemicals disposed in this manner will not migrate from the disposal site. A thorough understanding of the various processes that influence the persistence, retention, and leaching of pesticides in soils is required to develop technology for the selection and management of pesticide disposal sites involving soils. The fate of pesticides in soils when applied at concentrations similar to those associated with agricultural practices has been welldocumented in several reviews (Bailey & White, 1970; Sanborn et al, 1977). However, the direct extrapolation of these data to systems containing high pesticide concentrations, such as those occurring at or below disposal sites, may not always be feasible (Davidson et al., 1976).
MATERIALS AND METHODS
SoiLs
SoiLs used in this study were: Webster silty clay loam (Mollisol) from Iowa, Cecil sandy loam (Ultisol) from Georgia, and Eustis fine sand (Entisol) from Florida. These soils were selected on the basis of their iaxonomic and textural representation of major U.S. soils. Surface samples (0-30cm depth) of each soil were air-dried and passed through a 2-ram sieve prior to storage and use. Selected physical and chemical properties of these soils, pertinent to this study, are listed in Table 1.
Correspondence to: P. S. C. Rao, 2169 McCarty Hall. Soil Science Department, University of Florida, Gainesville, FL 32611, U.S.A., Phone: 904-392-1951.
PestiCides Four pesticides included in the study were: 2,4-D [2.4-dichiorophenoxyacetic acid], atrazine [2-chloro..4ethylamino-(>.isopropylamino-s-triazine], terbacil [3-tertbntyl-5-cldoro-6-methyluracil], and methyl parathion [OO-dimethyI-O-pnitrophenyl phospborothionte]. A stock solution of 10,000/~gml- 1 of 2,4-v in 0.01 N CaCI2 was prep~'ed using the commercial formulation Deal-Weed (Tbomson-Hayward Chem. Co, Kansas City, MS). DealWeed formulation consists of the dimethylamine salt of 2,4-D (41~ acid equivalent). A stock solution of 50/~gml- : of atrazine in 0.01 N CaCI: was prepared by dissolving appropriate amounts of AATREX B0W (80°/0 wettable powder; Ciba Geigy Corp., Greensboro, NC). Technical grade terbacil (97°/0 pure; E. I. DuPont Co., Wilmington, DE) was used to prepare a stock solution of 500/~gmlin 0.01 N CaCI,. A concentrated xylene solution of methyl parathion (80% solution; Monsanto Co, Agricultural Division, St. Louis, MO) was diluted in 0.01 N CaCi, to give a stock solution of 46/~gml. -~ Pesticide solutions at all
375
P. S. C. RAO and J. M. DAVIDSON
376
Table 1. Physical and chemical properties of the soils used in this study
Soil
Particle size fraction i",,) sand Silt Clay
Webster Cecil Eustis
18.4 65.8 93.8
45.3 19.5 3.0
38.3 14.7 3.2
pH ( 1: 1 pastel Water 1 N KCI 7.3 5.6 5.6
6.5 4.8 4.1
lower desired concentrations were prepared by successive dilutions of the stock solutions in 0.01 N CaClz. A mixture of antibiotics consisting of penicillin G and potymixin B sulfate (Sigma Chemical Co.. St. Louis, MO) was added to all pesticide solutions to prevent microbial degradation during storage and use. All pesticide dilutions were spiked with the appropriate l'tC-lahelled compound (uniformly ring labelled) to give specific activities in the range of 2-5 nCi ml.- 1
Adsorption experiments Equilibrium adsorption isotherms for all soil-pesticide combinations were measured using the batch procedure. Equilibrium was achieved by shaking duplicate samples of 5 or 10g soil with 10ml of pesticide solution in Pyrex screw-cap glass test tubes for 48 h. Preliminary experiments had indicated that there was no measurable increase in pesticide adsorption beyond this time. Following equilibration, the test tubes were centrifuged at 800 0 for 10rain and the t'C-activity in 1-ml aliquots of the clear supernatant solution was assayed by the liquid scintillation method. Dczrea.~ in pesticide solution concentration was attributed to adsorption by the soil. All adsorption experiments were performed at a constant temperature (23° + I°C).
Column displacement experiments Pesticide movement through saturated columns of Webster, Cecil, and Eastis soils was studied using the miscible displacement technique described by Davidson et al. (1968). Air-dry soil was packed in small increments into glass cylinders {15 cm long; 45 cmz cross-sectional area). Medium porosity fritted glass endplates served to retain the soil in the column. The soil was initially saturated with 0.01 N CaCI2 solution. A known volume of pesticide solution at a desired concentration was introduced into the soil at a constant flux using a constant-volume peristaltic pump. The applied pesticide solution was subsequently displaced through the soil column with 0.01 N CaCI_, at the same flux. Effluent solutions were collected in 5 or 10 ml aliquots using an automatic fraction collector. A pulse of 3H20 (specific activity 5nCiml -t) was also displaced through each soil column to characterize the transport of non-adsorbed solutes. The activity of I'~C and ~H in effluent fractions was assayed by liquid scintillation. The efficiencies of counting exceeded 90% for l'tC and 60~o for 3H in all cases. The column experiments consisted of displacing 2,4.-0 amine solutions at two concentrations (50 and 5000 #g ml- i) through columns of Cecil, Eustis and Webster soil, and 5 and 50#g ml- 1 solutions of atrazine through a Enstis soil column. All displacements were performed at a Darcy flux of approximately 0.22 cm h - t to ensure near.equilibrium conditions for pesticide adsorption during flow. The total volume of water held in the soil column (Vo) was gravimetrically determined at the end of each displacement by extruding the soil from the glass cylinders and oven-drying. The number of pore volumes (V/Vo) of solution displaced through the column was calculated by dividing the cumulative outflow volume (V) by total water volume (Vo) in the soil column. Effluent pesticide concentrations are expressed as relative concentrations (C/Co),
CEC (me 100g-t~
Organic C (o)
Base saturation (%)
Extractable acidity Ime 100g -~)
54.7 6.8 5.2
3.87 0.90 0.56
9l 31 10
5.15 4.68 4.68
where C and Co are. respectively, effluent and input concentrations. Plots of C/Co vs V.'Vo are referred to as breakthrough curves tor BTC). RESULTS AND DISCUSSION
Adsorption experiments Equilibrium adsorption isotherms were determined for each soil-pesticide combination by measuring pesticide adsorption at five or more concentrations ranging from zero to the pesticide's aqueous solubility limit. All adsorption isotherms were described by the Freundlich equation, S = KCN; where K and N are constants, and $ and C are adsorbed (#g g - t soil) and solution ( # g m l - t ) concentrations, respectively. The values of the Freundlich adsorption constants K and N for a given soil-pesticide combination were obtained using a least-squares fit to the adsorption isotherm. These values are summarized in Table 2. Because soil organic carbon content generally correlates well with pesticide adsorption, the use of an adsorption partition coefficient based upon organic carbon content rather than total soil mass has been proposed by Hamaker & Thompson (1972). Using this procedure, the amount of pesticide adsorbed was expressed as/zg g - I organic carbon and the Freundlich constants (Koc) for each adsorption isotherm were recomputed. These values are also presented in Table 2. As indicated by the CV values, the Koc values for a given pesticide are much less variable among the three soils studied than are the K values uncorrected for organic carbon. These results are in general agreement with the observation of Hamaker (1975) where the Koc values for a given pesticide were nearly independent of soil type. It should be recognized, however, that soil pH, clay content, cation exchange capacity and chemical nature of the pesticide may also play a significant role in determining pesticide adsorption by soils (Bailey & White, 1970). On the basis of the Koc values listed in Table 2, the extent of pesticide adsorption on soils was in the order of terbacii < 2,4-0 amine < atrazine < methyl parathion. Two important conclusions can be made based on the data presented in Table 2. First, the fact that the Freundlich equation describes all pesticide adsorption isotherms over a wide concentration range suggests that adsorption sites were not saturated at any concentration considered in this study. The amount of pesticide adsorbed by the soil continued to increase, at a decreasing rate, with each increase in solution
Adsorption and movement of selected pesticides
377
Table 2. Freundlich constants calculated from equilibrium adsorption isotherms for various soil-pesticide combinations Pesticide
Soil
2.4-D amine
Atrazine
Terbacil
Methyl parathion
K
N
Webster Cecil Eustis Average + ?oCV* Webster Cecil Eustis Average + ?oCV Webster Cecil Eustis Average + ° oCV
4.62 0.65 0.76 2.01 + 112 6.03 0.89 0.62 2.51 + 121 2.46 0.38 0.99 + 130
0.70 0.83 0.73 0.75 + 9 0.73 1.04 0.79 0.85 -+ 19 0.88 0.99 0.88 0.92 + 7
119.4 72.2 135.7 109.1 +_. 30 155.8 98.9 110.7 121.8 _+ 25 63.6 42.2 21.4 42.4 + 50
Webster Cecil • Eustis Average +_.0,oCV
13.39 3.95 2.72 6.69 + 87
0.75 0.85 0.86 0.82 -t- 7
346.0 438.6 486.4 423.7 +_ 17
0.12
Koc
* CV is the coefficient of variation. 0r CV = (standard deviation/average) x 100. concentration. This behavior may not hold for other pesticide adsorbents (Weber & Usinowicz, 1973). Second, contrary to a frequent assumption, pesticide adsorption isotherms measured were nonlinear, i.e. N less than 1 (Table 2). Linear adsorption isotherms have been generally accepted for low pesticide concentrations because it simplifies computer simulation modeling (Kay & Elrick, 1967~ Davidson et ai., 1968; Hugenberger et al, 1972; Davidson & Chang, 1972). The significance of nonlinear adsorption isotherms with regard to pesticide mobility in softs is discusssed in the following section.
Column displacement experiments The partial differential equation generally assumed to describe the movement of pesticides and other adsorbed solutes through soils under steady-state water flow conditions is (van Genuchten et al., 1974): ~C
~2C
~C
p ~S a-'~"
a-TffiDp'~x2-v~x--O
(1)
When the adsorption isotherm obeys the Freundlich equation, the convective-dispersive solute transport model, equation (1), reduces to:
R(C)
~C
= D~
OC
- v ~xx"
(2)
+pKNCN~11.
(3)
where. R(C)=I1
The retardation term R(C) is a quantitative index of the pesticide's mobility in that its value is equal to the ratio of the positions of the adsorbed and nonadsorbed solute fronts in soft. The value of the adsorption coefficient K in equation (3) for nonadsorbed solutes is equal to zero; hence, R(C) = I. For. adsorbed solutes, R(C) is greater than unity since the
value of K is larger than zero. Thus, larger values of R(C) indicate reduced pesticide mobility in soils. It may be noted from equation (3) that for the case of nonlinear adsorption isotherms (N < 1), the retardation term varies with solution concentration C, while for a linear isotherm (N = 1), R(C) is independent of pesticide solution concentration. Thus, the mobility of pesticides and other adsorbed solutes through soils is directly influenced by the shape of the equilibrium adsorption isotherms. Effluem breakthrough curves (BTC) were measured for 2,4-D amine at two input concentrations (Co ffi 50 and 5000flgm1-1) and tritiated water (aH20) using Webster, Cecil and Eustis soft columns. These BTC are shown in Figs. 1, 2 and 3. Tritiated water represents a non-adsorbed solute and serves as a reference for the adsorbed solutes (2,4-D amine in this case). A shift of the BTC for adsorbed solutes to the right of 3H20 BTC is due to an adsorption-induced retardation. A greater shift of the BTC to the right indicates increased adsorption; thus, a decreased mobility. It is apparent from the data presented in Figs. 1, 2 and 3 that the mobility of 2,4-D amine significantly increased as the input concentration (Co) increased from 50 to 5000/~g ml.- 1 Note that for the 5000/~gm]-1 input concentration. 2,4-D amine was nearly as mobile as was aH20. The effect of increased mobility at high concentration was more pronounced in the Webster soil (Fig. 1) than in the Cecil Soil (Fig. 2) or Eustis soil (Fig. 3). These column resfilts are consistent with the expected trend based upon equation (3) and the measured nonlinear adsorption isotherms (Table 2) for 2,4-D amine. Breakthrough curves for displacement of atrazine through Eustis soil column at two input concentrations (Co = 5 and 50Fgml - l ) and 3H20 are shown in Fig. 4. The trend of increased mobility at higher, atrazine solution concentration is evident. However, the differences in pesticide mobility between
378
P.S.C. RAO and J. M. DAVIDSON w ~
o.e
%
0.6
o
0.4
o
o e
0.2 0.00
~I
2 , 4 - oarnir~
o,.o
.50 ~ rnl -~ !
~
+
--
+
I 2
-'~
-
4
, ' m ~ S-
(~
Pore volumes,
l I0
I 12
14"
V/Vo
Fig. 1. Effluent breakthrough curves for 2,4-0 amine (Co : 50 and 5000pg/ml -~) and for tritiated water displacement through Webster soil column.
Cecil sod 2 , 4 - 0 omlne
1.0 -
°
o
0.8
÷
~
0
~
o
el'
0
o.~ °° 0.4
0 .f.O
o
0.~
O
•
"I-
o
"
~,
;,. 0 o
•
+Co-,~OO /~grrd- o
I
2
S
4
Pore vo4urms,
~Ib41b 5
6
7'
V/Vo
Fig. 2. Effluent breakthrough curves for 2,4-D amine (Co : 50 and 5000pg ml -~) and for triti~ed water displacement through Cecil soil column.
the two concentrations were not as large for atrazine (Fig. 4) as they were for 2,4-D amine (Figs. 1, 2 and 3). Deviation of adsorption isotherms from linearity, i.e. constant retardation term, i n ~ c a ~ exponentially as the concentration diffcrenca become larger and/or as N approaches zero (Hamaker & Thompson, 1972, Davidson et at, 1976). In this study, atrazin¢ concent r a t i o n varied only by 10-fold while the 2,4-D amine concentration varied by 100-fold. Furthermore,, the isotherm for 2,4-D amine adsorption in Webster soil Eusfis m l 2,4-oarnme
~'¢ IC o
q~QMilleoOoo o o o o o o o o o
O.E
"
• o6
•
•*°
~
o
:
•
•
~'04
o
ie •*co° e~*• •oe eel
•
.-~M z O
"--Co•50/.~m~"
°
o
*
•* 02
0
~,~00o ~.q~----o .
•
2
o
4 Pore volumes,
6 V/VQ
* *°•
8
•
iO
Fig. 3. Effluent breakthrough curves for 2,4-0 amine (Co = 50 and 5000 pg ml-t) and for tritiated water displacement through Eustis soil column.
was more nonlinear than that for the Eustis soilatrazin¢ system (Table 2). The position of the BTC for an adsorbed solute is governed by the nature of the equilibrium adsorption isotherm, equation (3), wtg,reas the shape of the BTC (i.e. symmetry or lack of it) is defined by the kinetics of adsorption-~csorption processes. Symmetrical BTC are obtained when adsorption is an instantaneous equilibrium process. However, under nonequilibrium conditions during flow, asymmetrical BTC are generally obtained (van C~nuchten er ai., 1974). All of the pesticide breakthrough curves measured in this study (Figs. 1-4) were asymmetrical in shape with extensive 'tailing' when C/Co approar,hed 1.0 or again at C/Co -- 0. However, su(:h tailing was absent in 3HzO breakthrough curves. Hence, the asymmetry of the measured BTC for the pesticides can be attributed to nonequilihrium conditions existing in the soil column during flow and not to dispersion. As the BTC simulated using equation (2) are symmetrical and sigmoidai in shape, this model fails to describe the reported data. Rao et aL (1979) present an evaluation of two conceptual models where non.equilibrium during flow was due to either kinetics-controlled or diffusion-controlled adsorpt ion-desorption processes. The illustrated increase in pesticide mobility at high
Adsorption and movement of selected pesticides
379
Eustis soil A1tazine
~ 0.11
"
oo:.~ 0,4
÷
o
~ o÷ 0 $
o 0
•
•
0.2 o ° 0
2
+
o Ooo l ~ q ~ o . ~ ~ 4
6
Pore volurn~,
8
I0
IZ
V / V0
Fig. 4. Effluent breakthrough curves for atrazine (Co = 5 and 50 ~tg ml-1) and for tritiated water displacement through Eustis soil column. concentrations limits the usefulness of the present low concentration data base for developing safe management practices for pesticide disposal in the soil. However, underestimation of pesticide movement by assuming linear adsorption isotherms may be less severe for pesticides with low aqueous solubilities. Ou et al. (1978a, b) showed that for high loading rates,
experiments. In Environmental Dynamics of Pesticides. (Edited by R. Haque and V. H. Freed), pp. 115-131. Plenum Press, N.Y. Hamaker J. W. & Thompson, J. M. (1972) Adsorption. In Organic Chemicals in the Soil Environment. (Edited by C. A. I. Goring and J. W. Hamaker), Vol. 1, pp. 49-143. Marcel Dekker, N.Y. Hugenberger F~ Lctey J. & Farmer, W. J. (1972) Observed and calculated distribution of Lindane in soil columns up to 20,000 pg 2,4-D g - 1 soil, there was significant as influenced by water movement. Soil Sci. Soc. Am. decrease in the pesticide degradation rate with a conProc. 36, 544-548. comitant depression of total microbial activity in the Kay B. D. & Elrick, D. E. (1967) Adsorption and movesoil. Thus, due to rapid leaching and minimal microment of Lindane in softs. Soil Sci. 104, 314--322. biological decomposition of pesticides at high concenLindsey A. W, Farb D. & Sanjour W. (1976) OSWMP chemical waste landfill and related projects. In Residue trations, the potential for ground water pollution inManooement by Land Disposal: Proceedinos of the creases. The results presented in this paper may also Hazardous Waste Research Syrup. (Edited by W. H. be applicable for describing the behavior of other Fuller), pp. 14-31. EPA.600/9-76-O15, July. hazardous wastes in soils. Ou L. T., Rothwell, D. F., Wheeler W. B. & Davidson J. M. (1978o) The effect of high 2,4-D concentrations on degradation .and carbon dioxide evolution in soils. J. Acknowledgemems--Tlfis research was supported in part Envir. Qual. 7, 241-246. by the U.S. Environmental Protection Agency (Grant No. 0 u L. T., Davidson, J. M. & Rothwell, D. F. (1978/7) ReR-803849) and in part by special funds from the Center spouse of soil microflora to high 2,4-D applications. Soil for Environmental Programs at the Institute of Food and Biol. Biochem. 10, 443-445. Agricultural Sciences, University of Florida. Published as Ran P. S. C, Davidson J. M., Jessup R. E. & Sclim H. M. Florida Agricultural Experiment Station Journal Series No. 1110. (1979) Evaluation of conceptual models for describing non equilibrium adsorption-desorption of pesticides REFERENCES during steady-flow in soils. Soil Sci. Soc. Am. J. 43 (in press). Atkins P. R. (1972) The pesticide manufacturing indnstry-Ronston R. C. & Wftdung R. E. (1969) Ultimate disposal current waste treatment and disposal practices. Wat, Polof wastes to soft. In Wmer. (Edited by L. K. Cecil). Chem. iut. Control Res. Series. 12020 FYE 61/72, pp. 185. F,ngng. Prog. Syrap. Set. 64.97. pp. 19-25. Bailey G. W. & White J. L. (1970) Factors intiuencing Sanborn J. R, Fracis B. F. & Metcalf R. L. (19777) The adsorption, desorption and movement of pesticides in degradation of selected-pesticides in soft: a review of soft. Residue Rev. 32, 29-92. published literature. EPA-600/9-77-022. Davidson J. M., Rieck C. E. & Santeimann P. W. (1968) Schomaker N. B. (1976) Current research on land disposal Influence of water flux and porous materials on the of hazardous wastes. In Residue Management by Land movement of selected herbicides. Soil Sci. Soc. Am. Proc. Disposal: Proc. of the Hazardous Waste Res. Syrup. 32, 629--633. (Edited by W. H. Fuller), EPA-600/9-76-01.5, pp. 1-13. Davidson J. M. & Chang, R. K. (1972) Transport of picJuly. loram in relation to soil physical conditions and porevan Genuchten M. Th, Davidson J. M. & Wierenga, P. J. water velocity. Soil Sci. Soc. Am. Proc. 36, 257-261. (1974) An evaluation of kinetic and equilibrium equaDavidson J. M, Ou L. T. & Rag P. S. C. (1976) Behavior tions for predicting pesticide movement through porous of high pesticide concentrations in soft-water systems, media. Soil Sci. Soc. Am. Proc. 38, 29-35. In Residual Manooement by Land Disposal: Proceedinos yon Everdingen R. O. & Freeze R. A. (1971) Subsurfaceof the Hazardous Wastes Research Syrup. (Edited by W. disposal of waste in Canada. Tech. Bull No. 49. p. 19. H. Fuller), pp. 206-212. EPA-600/9-76-O15, July, 1 9 7 6 . Inland Waters Branch, Dept. of Environment, Ottawa, Fields T~ Jr. & Lindsey A. W. (1975) Landfill Disposal Canada. of Hazardous Wastes: a Review of Literature and Known yon Rumker R, Lawless E. W~ Meiners A. F~ Lawrence Approaches. U . S . E . P . A . Publication S. W. 15: WashK.A., Keiso G. L. & Horay, F. (1974) Productiog Distrjington, D.C. butio'n, Use and Environmental Impact Potential of Hamaker J. W. (1975) The interpretation of soil leaching Selected Pesticides. EPA-540/1-74-O01.
380
P.S.C.
RAO a n d J. M. DAVIDSON
W e b e r W. J. & U s i n o w i c z P. J. (1973) A d s o r p t i o n f r o m a q u e o u s s o l u t i o n . Tech, Publication, Research Project 17020 E P F , U.S. E n v i r o n . P r o t . A g e n c y , C i n c i n n a t i , Ohio. NOMENCLATURE C -- s o l u t i o n c o n c e n t r a t i o n . / ~ g m l - 1 D -- d i s p e r s i o n coefficient, c m z h K = Freundlich adsorption constant
Koc = F r e u n d l i c h a d s o r p t i o n c o n s t a n t b a s e d o n o r g a n i c N = R(C) = S = t = t' = x = p = 0 =
c a r b o n c o n t e n t o f soil Freundlich exponent r e t a r d a t i o n t e r m . a n i n d e x of pesticide m o b i l i t y a d s o r b e d c o n c e n t r a t i o n . / a g g - 1 soil time. h a v e r a g e p o r e - w a t e r velocity, c m h - t distance, cm soil b u l k density, g soil c m - 3 v o l u m e t r i c s o i l - w a t e r c o n t e n t , c m 3. c m - 3
Pergamon Press l.td 1979. Printed in Great Britain
ADSORPTION AND MOVEMENT OF SELECTED PESTICIDES AT HIGH CONCENTRATIONS tN SOILS P. S. C. RAO and J. M. DAVIDSON Soil Science Department, University of Florida, Gainesville, FL 3261l, U.S.A.
(Received 29 June 1978) Abstract--Adsorption of 2,4-D amine, atrazine, terbacil and methyl parathion pesticides on Webster, Cecil and Eustis soils was measured at pesticide solution concentrations ranging from zero to the aqueous solubility limit of each pesticide. Measured equilibrium adsorption isotherms for nearly all soil-pesticide combinations were of nonlinear Freundlich type. The Freundlich adsorption constant (K) based on soil organic carbon was much less variable for a given pesticide among the four soils than was the K based on total soil mass. The influence of the shape of the adsorption isotherm on the movement of 2,4-D amine and atrazine through water-saturated soil columns Was also examined. Pesticide effluent concentrations from soil columns were measured at two input solution concentrations (50 and 5000/~gmi- t for Z4-D amine; 5 and 50/~g ml - ~ for atrazine). In all cases, pesticide mobility was significantly greater for the higher concentrations. Thus, serious errors may he introduced by assuming a linear adsorption isotherm (i.e. pesticide mobility is invariant with input concentration) when predicting pesticide transport from waste disposal sites where high concentrations exist.
Laboratory studies were initiated to investigate the physical, chemical and microbiological behavior of pesticides in soils when applied at high concentrations. The microbiological aspects of this study are reported elsewhere by Ou et aL (1978a, b). The influence of adsorption characteristics on pesticide mobility are considered in this paper.
INTRODUCTION
Because of a continued increase in the number and quantity of pesticide compounds being placed on the market, the safe disposal of surplus and/or waste pesticide materials has become an acute problem (yon Rumker et al, 1974). Incineration, encapsulation, isolation in underground caves and mines, chemical stabilization, land spreading and landfilling are some of the procedures being considered for the disposal of pesticides and other hazardous wastes (van Everdingen & Freeze, 1971; Schomaker, 1976). Of these methods, disposal by landfills and land spreading appear to be more common and economical (Fields & Lindsey,'1975; Lindsey et al, 1976). Placing hazardous waste in the land has come under severe attack recently (Rouston & Wiidung, 1969; Atkins, 1972) because it does not guarantee that the hazardous chemicals disposed in this manner will not migrate from the disposal site. A thorough understanding of the various processes that influence the persistence, retention, and leaching of pesticides in soils is required to develop technology for the selection and management of pesticide disposal sites involving soils. The fate of pesticides in soils when applied at concentrations similar to those associated with agricultural practices has been welldocumented in several reviews (Bailey & White, 1970; Sanborn et al, 1977). However, the direct extrapolation of these data to systems containing high pesticide concentrations, such as those occurring at or below disposal sites, may not always be feasible (Davidson et al., 1976).
MATERIALS AND METHODS
SoiLs
SoiLs used in this study were: Webster silty clay loam (Mollisol) from Iowa, Cecil sandy loam (Ultisol) from Georgia, and Eustis fine sand (Entisol) from Florida. These soils were selected on the basis of their iaxonomic and textural representation of major U.S. soils. Surface samples (0-30cm depth) of each soil were air-dried and passed through a 2-ram sieve prior to storage and use. Selected physical and chemical properties of these soils, pertinent to this study, are listed in Table 1.
Correspondence to: P. S. C. Rao, 2169 McCarty Hall. Soil Science Department, University of Florida, Gainesville, FL 32611, U.S.A., Phone: 904-392-1951.
PestiCides Four pesticides included in the study were: 2,4-D [2.4-dichiorophenoxyacetic acid], atrazine [2-chloro..4ethylamino-(>.isopropylamino-s-triazine], terbacil [3-tertbntyl-5-cldoro-6-methyluracil], and methyl parathion [OO-dimethyI-O-pnitrophenyl phospborothionte]. A stock solution of 10,000/~gml- 1 of 2,4-v in 0.01 N CaCI2 was prep~'ed using the commercial formulation Deal-Weed (Tbomson-Hayward Chem. Co, Kansas City, MS). DealWeed formulation consists of the dimethylamine salt of 2,4-D (41~ acid equivalent). A stock solution of 50/~gml- : of atrazine in 0.01 N CaCI: was prepared by dissolving appropriate amounts of AATREX B0W (80°/0 wettable powder; Ciba Geigy Corp., Greensboro, NC). Technical grade terbacil (97°/0 pure; E. I. DuPont Co., Wilmington, DE) was used to prepare a stock solution of 500/~gmlin 0.01 N CaCI,. A concentrated xylene solution of methyl parathion (80% solution; Monsanto Co, Agricultural Division, St. Louis, MO) was diluted in 0.01 N CaCi, to give a stock solution of 46/~gml. -~ Pesticide solutions at all
375
P. S. C. RAO and J. M. DAVIDSON
376
Table 1. Physical and chemical properties of the soils used in this study
Soil
Particle size fraction i",,) sand Silt Clay
Webster Cecil Eustis
18.4 65.8 93.8
45.3 19.5 3.0
38.3 14.7 3.2
pH ( 1: 1 pastel Water 1 N KCI 7.3 5.6 5.6
6.5 4.8 4.1
lower desired concentrations were prepared by successive dilutions of the stock solutions in 0.01 N CaClz. A mixture of antibiotics consisting of penicillin G and potymixin B sulfate (Sigma Chemical Co.. St. Louis, MO) was added to all pesticide solutions to prevent microbial degradation during storage and use. All pesticide dilutions were spiked with the appropriate l'tC-lahelled compound (uniformly ring labelled) to give specific activities in the range of 2-5 nCi ml.- 1
Adsorption experiments Equilibrium adsorption isotherms for all soil-pesticide combinations were measured using the batch procedure. Equilibrium was achieved by shaking duplicate samples of 5 or 10g soil with 10ml of pesticide solution in Pyrex screw-cap glass test tubes for 48 h. Preliminary experiments had indicated that there was no measurable increase in pesticide adsorption beyond this time. Following equilibration, the test tubes were centrifuged at 800 0 for 10rain and the t'C-activity in 1-ml aliquots of the clear supernatant solution was assayed by the liquid scintillation method. Dczrea.~ in pesticide solution concentration was attributed to adsorption by the soil. All adsorption experiments were performed at a constant temperature (23° + I°C).
Column displacement experiments Pesticide movement through saturated columns of Webster, Cecil, and Eastis soils was studied using the miscible displacement technique described by Davidson et al. (1968). Air-dry soil was packed in small increments into glass cylinders {15 cm long; 45 cmz cross-sectional area). Medium porosity fritted glass endplates served to retain the soil in the column. The soil was initially saturated with 0.01 N CaCI2 solution. A known volume of pesticide solution at a desired concentration was introduced into the soil at a constant flux using a constant-volume peristaltic pump. The applied pesticide solution was subsequently displaced through the soil column with 0.01 N CaCI_, at the same flux. Effluent solutions were collected in 5 or 10 ml aliquots using an automatic fraction collector. A pulse of 3H20 (specific activity 5nCiml -t) was also displaced through each soil column to characterize the transport of non-adsorbed solutes. The activity of I'~C and ~H in effluent fractions was assayed by liquid scintillation. The efficiencies of counting exceeded 90% for l'tC and 60~o for 3H in all cases. The column experiments consisted of displacing 2,4.-0 amine solutions at two concentrations (50 and 5000 #g ml- i) through columns of Cecil, Eustis and Webster soil, and 5 and 50#g ml- 1 solutions of atrazine through a Enstis soil column. All displacements were performed at a Darcy flux of approximately 0.22 cm h - t to ensure near.equilibrium conditions for pesticide adsorption during flow. The total volume of water held in the soil column (Vo) was gravimetrically determined at the end of each displacement by extruding the soil from the glass cylinders and oven-drying. The number of pore volumes (V/Vo) of solution displaced through the column was calculated by dividing the cumulative outflow volume (V) by total water volume (Vo) in the soil column. Effluent pesticide concentrations are expressed as relative concentrations (C/Co),
CEC (me 100g-t~
Organic C (o)
Base saturation (%)
Extractable acidity Ime 100g -~)
54.7 6.8 5.2
3.87 0.90 0.56
9l 31 10
5.15 4.68 4.68
where C and Co are. respectively, effluent and input concentrations. Plots of C/Co vs V.'Vo are referred to as breakthrough curves tor BTC). RESULTS AND DISCUSSION
Adsorption experiments Equilibrium adsorption isotherms were determined for each soil-pesticide combination by measuring pesticide adsorption at five or more concentrations ranging from zero to the pesticide's aqueous solubility limit. All adsorption isotherms were described by the Freundlich equation, S = KCN; where K and N are constants, and $ and C are adsorbed (#g g - t soil) and solution ( # g m l - t ) concentrations, respectively. The values of the Freundlich adsorption constants K and N for a given soil-pesticide combination were obtained using a least-squares fit to the adsorption isotherm. These values are summarized in Table 2. Because soil organic carbon content generally correlates well with pesticide adsorption, the use of an adsorption partition coefficient based upon organic carbon content rather than total soil mass has been proposed by Hamaker & Thompson (1972). Using this procedure, the amount of pesticide adsorbed was expressed as/zg g - I organic carbon and the Freundlich constants (Koc) for each adsorption isotherm were recomputed. These values are also presented in Table 2. As indicated by the CV values, the Koc values for a given pesticide are much less variable among the three soils studied than are the K values uncorrected for organic carbon. These results are in general agreement with the observation of Hamaker (1975) where the Koc values for a given pesticide were nearly independent of soil type. It should be recognized, however, that soil pH, clay content, cation exchange capacity and chemical nature of the pesticide may also play a significant role in determining pesticide adsorption by soils (Bailey & White, 1970). On the basis of the Koc values listed in Table 2, the extent of pesticide adsorption on soils was in the order of terbacii < 2,4-0 amine < atrazine < methyl parathion. Two important conclusions can be made based on the data presented in Table 2. First, the fact that the Freundlich equation describes all pesticide adsorption isotherms over a wide concentration range suggests that adsorption sites were not saturated at any concentration considered in this study. The amount of pesticide adsorbed by the soil continued to increase, at a decreasing rate, with each increase in solution
Adsorption and movement of selected pesticides
377
Table 2. Freundlich constants calculated from equilibrium adsorption isotherms for various soil-pesticide combinations Pesticide
Soil
2.4-D amine
Atrazine
Terbacil
Methyl parathion
K
N
Webster Cecil Eustis Average + ?oCV* Webster Cecil Eustis Average + ?oCV Webster Cecil Eustis Average + ° oCV
4.62 0.65 0.76 2.01 + 112 6.03 0.89 0.62 2.51 + 121 2.46 0.38 0.99 + 130
0.70 0.83 0.73 0.75 + 9 0.73 1.04 0.79 0.85 -+ 19 0.88 0.99 0.88 0.92 + 7
119.4 72.2 135.7 109.1 +_. 30 155.8 98.9 110.7 121.8 _+ 25 63.6 42.2 21.4 42.4 + 50
Webster Cecil • Eustis Average +_.0,oCV
13.39 3.95 2.72 6.69 + 87
0.75 0.85 0.86 0.82 -t- 7
346.0 438.6 486.4 423.7 +_ 17
0.12
Koc
* CV is the coefficient of variation. 0r CV = (standard deviation/average) x 100. concentration. This behavior may not hold for other pesticide adsorbents (Weber & Usinowicz, 1973). Second, contrary to a frequent assumption, pesticide adsorption isotherms measured were nonlinear, i.e. N less than 1 (Table 2). Linear adsorption isotherms have been generally accepted for low pesticide concentrations because it simplifies computer simulation modeling (Kay & Elrick, 1967~ Davidson et ai., 1968; Hugenberger et al, 1972; Davidson & Chang, 1972). The significance of nonlinear adsorption isotherms with regard to pesticide mobility in softs is discusssed in the following section.
Column displacement experiments The partial differential equation generally assumed to describe the movement of pesticides and other adsorbed solutes through soils under steady-state water flow conditions is (van Genuchten et al., 1974): ~C
~2C
~C
p ~S a-'~"
a-TffiDp'~x2-v~x--O
(1)
When the adsorption isotherm obeys the Freundlich equation, the convective-dispersive solute transport model, equation (1), reduces to:
R(C)
~C
= D~
OC
- v ~xx"
(2)
+pKNCN~11.
(3)
where. R(C)=I1
The retardation term R(C) is a quantitative index of the pesticide's mobility in that its value is equal to the ratio of the positions of the adsorbed and nonadsorbed solute fronts in soft. The value of the adsorption coefficient K in equation (3) for nonadsorbed solutes is equal to zero; hence, R(C) = I. For. adsorbed solutes, R(C) is greater than unity since the
value of K is larger than zero. Thus, larger values of R(C) indicate reduced pesticide mobility in soils. It may be noted from equation (3) that for the case of nonlinear adsorption isotherms (N < 1), the retardation term varies with solution concentration C, while for a linear isotherm (N = 1), R(C) is independent of pesticide solution concentration. Thus, the mobility of pesticides and other adsorbed solutes through soils is directly influenced by the shape of the equilibrium adsorption isotherms. Effluem breakthrough curves (BTC) were measured for 2,4-D amine at two input concentrations (Co ffi 50 and 5000flgm1-1) and tritiated water (aH20) using Webster, Cecil and Eustis soft columns. These BTC are shown in Figs. 1, 2 and 3. Tritiated water represents a non-adsorbed solute and serves as a reference for the adsorbed solutes (2,4-D amine in this case). A shift of the BTC for adsorbed solutes to the right of 3H20 BTC is due to an adsorption-induced retardation. A greater shift of the BTC to the right indicates increased adsorption; thus, a decreased mobility. It is apparent from the data presented in Figs. 1, 2 and 3 that the mobility of 2,4-D amine significantly increased as the input concentration (Co) increased from 50 to 5000/~g ml.- 1 Note that for the 5000/~gm]-1 input concentration. 2,4-D amine was nearly as mobile as was aH20. The effect of increased mobility at high concentration was more pronounced in the Webster soil (Fig. 1) than in the Cecil Soil (Fig. 2) or Eustis soil (Fig. 3). These column resfilts are consistent with the expected trend based upon equation (3) and the measured nonlinear adsorption isotherms (Table 2) for 2,4-D amine. Breakthrough curves for displacement of atrazine through Eustis soil column at two input concentrations (Co = 5 and 50Fgml - l ) and 3H20 are shown in Fig. 4. The trend of increased mobility at higher, atrazine solution concentration is evident. However, the differences in pesticide mobility between
378
P.S.C. RAO and J. M. DAVIDSON w ~
o.e
%
0.6
o
0.4
o
o e
0.2 0.00
~I
2 , 4 - oarnir~
o,.o
.50 ~ rnl -~ !
~
+
--
+
I 2
-'~
-
4
, ' m ~ S-
(~
Pore volumes,
l I0
I 12
14"
V/Vo
Fig. 1. Effluent breakthrough curves for 2,4-0 amine (Co : 50 and 5000pg/ml -~) and for tritiated water displacement through Webster soil column.
Cecil sod 2 , 4 - 0 omlne
1.0 -
°
o
0.8
÷
~
0
~
o
el'
0
o.~ °° 0.4
0 .f.O
o
0.~
O
•
"I-
o
"
~,
;,. 0 o
•
+Co-,~OO /~grrd- o
I
2
S
4
Pore vo4urms,
~Ib41b 5
6
7'
V/Vo
Fig. 2. Effluent breakthrough curves for 2,4-D amine (Co : 50 and 5000pg ml -~) and for triti~ed water displacement through Cecil soil column.
the two concentrations were not as large for atrazine (Fig. 4) as they were for 2,4-D amine (Figs. 1, 2 and 3). Deviation of adsorption isotherms from linearity, i.e. constant retardation term, i n ~ c a ~ exponentially as the concentration diffcrenca become larger and/or as N approaches zero (Hamaker & Thompson, 1972, Davidson et at, 1976). In this study, atrazin¢ concent r a t i o n varied only by 10-fold while the 2,4-D amine concentration varied by 100-fold. Furthermore,, the isotherm for 2,4-D amine adsorption in Webster soil Eusfis m l 2,4-oarnme
~'¢ IC o
q~QMilleoOoo o o o o o o o o o
O.E
"
• o6
•
•*°
~
o
:
•
•
~'04
o
ie •*co° e~*• •oe eel
•
.-~M z O
"--Co•50/.~m~"
°
o
*
•* 02
0
~,~00o ~.q~----o .
•
2
o
4 Pore volumes,
6 V/VQ
* *°•
8
•
iO
Fig. 3. Effluent breakthrough curves for 2,4-0 amine (Co = 50 and 5000 pg ml-t) and for tritiated water displacement through Eustis soil column.
was more nonlinear than that for the Eustis soilatrazin¢ system (Table 2). The position of the BTC for an adsorbed solute is governed by the nature of the equilibrium adsorption isotherm, equation (3), wtg,reas the shape of the BTC (i.e. symmetry or lack of it) is defined by the kinetics of adsorption-~csorption processes. Symmetrical BTC are obtained when adsorption is an instantaneous equilibrium process. However, under nonequilibrium conditions during flow, asymmetrical BTC are generally obtained (van C~nuchten er ai., 1974). All of the pesticide breakthrough curves measured in this study (Figs. 1-4) were asymmetrical in shape with extensive 'tailing' when C/Co approar,hed 1.0 or again at C/Co -- 0. However, su(:h tailing was absent in 3HzO breakthrough curves. Hence, the asymmetry of the measured BTC for the pesticides can be attributed to nonequilihrium conditions existing in the soil column during flow and not to dispersion. As the BTC simulated using equation (2) are symmetrical and sigmoidai in shape, this model fails to describe the reported data. Rao et aL (1979) present an evaluation of two conceptual models where non.equilibrium during flow was due to either kinetics-controlled or diffusion-controlled adsorpt ion-desorption processes. The illustrated increase in pesticide mobility at high
Adsorption and movement of selected pesticides
379
Eustis soil A1tazine
~ 0.11
"
oo:.~ 0,4
÷
o
~ o÷ 0 $
o 0
•
•
0.2 o ° 0
2
+
o Ooo l ~ q ~ o . ~ ~ 4
6
Pore volurn~,
8
I0
IZ
V / V0
Fig. 4. Effluent breakthrough curves for atrazine (Co = 5 and 50 ~tg ml-1) and for tritiated water displacement through Eustis soil column. concentrations limits the usefulness of the present low concentration data base for developing safe management practices for pesticide disposal in the soil. However, underestimation of pesticide movement by assuming linear adsorption isotherms may be less severe for pesticides with low aqueous solubilities. Ou et al. (1978a, b) showed that for high loading rates,
experiments. In Environmental Dynamics of Pesticides. (Edited by R. Haque and V. H. Freed), pp. 115-131. Plenum Press, N.Y. Hamaker J. W. & Thompson, J. M. (1972) Adsorption. In Organic Chemicals in the Soil Environment. (Edited by C. A. I. Goring and J. W. Hamaker), Vol. 1, pp. 49-143. Marcel Dekker, N.Y. Hugenberger F~ Lctey J. & Farmer, W. J. (1972) Observed and calculated distribution of Lindane in soil columns up to 20,000 pg 2,4-D g - 1 soil, there was significant as influenced by water movement. Soil Sci. Soc. Am. decrease in the pesticide degradation rate with a conProc. 36, 544-548. comitant depression of total microbial activity in the Kay B. D. & Elrick, D. E. (1967) Adsorption and movesoil. Thus, due to rapid leaching and minimal microment of Lindane in softs. Soil Sci. 104, 314--322. biological decomposition of pesticides at high concenLindsey A. W, Farb D. & Sanjour W. (1976) OSWMP chemical waste landfill and related projects. In Residue trations, the potential for ground water pollution inManooement by Land Disposal: Proceedinos of the creases. The results presented in this paper may also Hazardous Waste Research Syrup. (Edited by W. H. be applicable for describing the behavior of other Fuller), pp. 14-31. EPA.600/9-76-O15, July. hazardous wastes in soils. Ou L. T., Rothwell, D. F., Wheeler W. B. & Davidson J. M. (1978o) The effect of high 2,4-D concentrations on degradation .and carbon dioxide evolution in soils. J. Acknowledgemems--Tlfis research was supported in part Envir. Qual. 7, 241-246. by the U.S. Environmental Protection Agency (Grant No. 0 u L. T., Davidson, J. M. & Rothwell, D. F. (1978/7) ReR-803849) and in part by special funds from the Center spouse of soil microflora to high 2,4-D applications. Soil for Environmental Programs at the Institute of Food and Biol. Biochem. 10, 443-445. Agricultural Sciences, University of Florida. Published as Ran P. S. C, Davidson J. M., Jessup R. E. & Sclim H. M. Florida Agricultural Experiment Station Journal Series No. 1110. (1979) Evaluation of conceptual models for describing non equilibrium adsorption-desorption of pesticides REFERENCES during steady-flow in soils. Soil Sci. Soc. Am. J. 43 (in press). Atkins P. R. (1972) The pesticide manufacturing indnstry-Ronston R. C. & Wftdung R. E. (1969) Ultimate disposal current waste treatment and disposal practices. Wat, Polof wastes to soft. In Wmer. (Edited by L. K. Cecil). Chem. iut. Control Res. Series. 12020 FYE 61/72, pp. 185. F,ngng. Prog. Syrap. Set. 64.97. pp. 19-25. Bailey G. W. & White J. L. (1970) Factors intiuencing Sanborn J. R, Fracis B. F. & Metcalf R. L. (19777) The adsorption, desorption and movement of pesticides in degradation of selected-pesticides in soft: a review of soft. Residue Rev. 32, 29-92. published literature. EPA-600/9-77-022. Davidson J. M., Rieck C. E. & Santeimann P. W. (1968) Schomaker N. B. (1976) Current research on land disposal Influence of water flux and porous materials on the of hazardous wastes. In Residue Management by Land movement of selected herbicides. Soil Sci. Soc. Am. Proc. Disposal: Proc. of the Hazardous Waste Res. Syrup. 32, 629--633. (Edited by W. H. Fuller), EPA-600/9-76-01.5, pp. 1-13. Davidson J. M. & Chang, R. K. (1972) Transport of picJuly. loram in relation to soil physical conditions and porevan Genuchten M. Th, Davidson J. M. & Wierenga, P. J. water velocity. Soil Sci. Soc. Am. Proc. 36, 257-261. (1974) An evaluation of kinetic and equilibrium equaDavidson J. M, Ou L. T. & Rag P. S. C. (1976) Behavior tions for predicting pesticide movement through porous of high pesticide concentrations in soft-water systems, media. Soil Sci. Soc. Am. Proc. 38, 29-35. In Residual Manooement by Land Disposal: Proceedinos yon Everdingen R. O. & Freeze R. A. (1971) Subsurfaceof the Hazardous Wastes Research Syrup. (Edited by W. disposal of waste in Canada. Tech. Bull No. 49. p. 19. H. Fuller), pp. 206-212. EPA-600/9-76-O15, July, 1 9 7 6 . Inland Waters Branch, Dept. of Environment, Ottawa, Fields T~ Jr. & Lindsey A. W. (1975) Landfill Disposal Canada. of Hazardous Wastes: a Review of Literature and Known yon Rumker R, Lawless E. W~ Meiners A. F~ Lawrence Approaches. U . S . E . P . A . Publication S. W. 15: WashK.A., Keiso G. L. & Horay, F. (1974) Productiog Distrjington, D.C. butio'n, Use and Environmental Impact Potential of Hamaker J. W. (1975) The interpretation of soil leaching Selected Pesticides. EPA-540/1-74-O01.
380
P.S.C.
RAO a n d J. M. DAVIDSON
W e b e r W. J. & U s i n o w i c z P. J. (1973) A d s o r p t i o n f r o m a q u e o u s s o l u t i o n . Tech, Publication, Research Project 17020 E P F , U.S. E n v i r o n . P r o t . A g e n c y , C i n c i n n a t i , Ohio. NOMENCLATURE C -- s o l u t i o n c o n c e n t r a t i o n . / ~ g m l - 1 D -- d i s p e r s i o n coefficient, c m z h K = Freundlich adsorption constant
Koc = F r e u n d l i c h a d s o r p t i o n c o n s t a n t b a s e d o n o r g a n i c N = R(C) = S = t = t' = x = p = 0 =
c a r b o n c o n t e n t o f soil Freundlich exponent r e t a r d a t i o n t e r m . a n i n d e x of pesticide m o b i l i t y a d s o r b e d c o n c e n t r a t i o n . / a g g - 1 soil time. h a v e r a g e p o r e - w a t e r velocity, c m h - t distance, cm soil b u l k density, g soil c m - 3 v o l u m e t r i c s o i l - w a t e r c o n t e n t , c m 3. c m - 3