Kinetics of Phosphorus Release from Spodosols: Effects of Oxalate and FormateT. R. Fox,* N. B. Comerford, and W. W. McFee

ABSTRACTOrganic anions have been implicated in the release of P to soil

solution and may thus increase the availability of P to plants. Ox-alate and formate have been identified as the most prevalent low-molecular-weight organic anions in the soil solutions in Spodosolsunder forest vegetation in northern Florida. The purpose of this studywas to investigate the effect that these two organic anions have onthe kinetics of inorganic- and organic-P release from the A, Bh, andBt horizons of a representative forested Spodosol. Soil samples wereequilibrated for various periods of time with solutions of eitheroxalate, formate, or water. The extracts were analyzed for pH, in-organic P, organic P, Al, and organic acids. Results showed thatoxalate had little effect on P release in the A horizon. This was dueto the lack of Al oxides to hold P. Conversely, oxalate greatly in-creased solution P in the Bh and Bt horizons. In the subsoil horizons,the release of P was rapid, followed the disappearance of oxalatefrom solution, and was coincident with an increase in pH. Thesefacts suggested that P was released via a ligand-exchange reaction.A longer term reaction was consistent with the reprecipitation of theP as an amorphous Al phosphate when oxalate was either totallysorbed or degraded. The presence of formate did not increase Prelease.

THE INFLUENCE of low-molecular-weight (LMW)organic anions on P availability has drawn con-

siderable attention in recent years (Sollins et al., 1981;Gardner et al., 1983; Marschner et al., 1986; Jurinaket al., 1986). Because LMW organic anions functionas organic ligands, they can increase P in solution by(i) replacing P sorbed at metal-hydroxide surfacesthrough ligand-exchange reactions (Stumm, 1986), (ii)dissolving metal-oxide surfaces that sorb P (Stummand Morgan, 1981; Martell et al., 1988), and (iii) com-plexing metals in solution and thus preventing pre-cipitation of metal phosphates (Ng Kee Kwong andHuang, 1977).

In Spodosols of the southeastern USA, P deficienciescommonly limit growth and development of foreststands. Since reactions with Al tend to control P avail-ability in these soils (Ballard and Fiskell, 1974), theaction of LMW organic anions on release of P dependson their ability to complex Al. Organic acids that formstable complexes with Al should, theoretically, in-crease P desorption.

Low-molecular-weight organic anions have beenshown to affect the kinetics of P release. Kuo and Lotse(1974) reported that the rate of P desorption from gibb-site was faster in the presence of ethylenediamine te-traacetate (EDTA) than in the presence of formate.Likewise, Traina et al. (1987) observed a more rapidinitial release of P from an acid, montmorillonitic soilin the presence of citrate compared with formate. InT.R Fox, ITT Rayonier, Inc., Forest Research Center, P.O. Box 819,Yulee, FL 32097; N.B. Comerford, Soil Science Dep., Univ. of Flor-ida, Gainesville, FL 32611; W.W. McFee, Agronomy Dep., PurdueUniv., West Lafayette, IN 47907. A contribution of the FloridaAgric. Exp. Stn., Journal Series no. R-00289. Received 20 Nov. 1989.Corresponding author.

Published in Soil Sci. Soc. Am. J. 54:1441-1447 (1990).

both cases, the rate of P release was faster with theligand that formed more stable complexes with Al.

Fox and Comerford (1990) found oxalate and for-mate to be the dominant LMW organic anions in agroup of representative Spodosols in north Florida.Concentrations ranged from 0.1 to 1 mM. They foundthat the oxalate levels in soil solutions were an orderof magnitude higher in Bh horizons than in A hori-zons. Therefore, oxalate may affect the release of P,and may possibly have a significant effect on P nutri-tion of forest stands growing in these Spodosols.

One approach toward understanding the effect ofLMW organic anions on P release and subsequentavailability to plants growing in these soils is to in-vestigate the kinetics of P release. The objectives ofthis research were to (i) examine the effects of oxalateand formate on the kinetics of P release from A, Bh,and Bt horizon soil material from a representative,poorly drained Spodosol; and (ii) through the use ofreaction rate, pH change, and the fate of the organicanions, illucidate reaction mechanisms.

MATERIALS AND METHODSSoil Material

Samples of soil material from A, Bh, and Bt horizons froma Pomona series (a sandy, siliceous, hyperthermic UlticHaplaquod) soil were collected from a single soil pit locatedin Alachua County, Florida. The soil material was air dried,passed through a 2-mm sieve, and stored for up to a year inplastic barrels prior to use. Selected chemical and physicalproperties, determined by standard methods (Page et al.,1982; Klute, 1986), are presented in Table 1.

Desorption ExperimentsThe kinetics of P desorption in oxalate and formate so-

lutions, as well as water, were investigated using a batchprocedure. Selected properties of oxalic and formic acid arelisted in Table 2. Oxalate solutions of 0.1 and 1.0 mM wereprepared from reagent-grade oxalic acid, monohydrate(Fisher Scientific, Pittsburgh, PA). Formate solutions of 1mM were prepared from reagent-grade calcium formate(Fisher Scientific). Deionized water was also used. The or-ganic-acid solutions and the deionized water were adjustedto pH 4.3 with HC1 or NaOH. A 4.3 pH was used to ap-proximate the soil-solution pH. The solutions had the fol-lowing ionic strengths: 1.0 mM oxalate = 0.0031 Af;0.1 mMoxalate = 0.0008 M; 1.0 mM formate = 0.0029 M; pH 4.3acidified water = 0.0001 M.

Duplicate 10-g samples of air-dried soil material wereplaced in 250-mL polyethylene bottles and 100 mL of ex-tracting solution were added to the bottle along with twodrops of toluene. The bottles were capped and shaken on areciprocating shaker at 100 cycles per minute. Samples wereremoved after 0.5, 1, 3, 6, 12, 24, 48, and 72 h and filteredthrough 0.45-jum nylon membrane filters. The filtrates werestored frozen at 0 °C until analyzed.

Chemical AnalysisThe conductivity of the filtered extracts was measured

with a Pt conductivity electrode, and solution pH was meas-ured with a combination glass electrode. Aluminum was de-

1441

1442 SOIL SCI. SOC. AM. J., VOL. 54, SEPTEMBER-OCTOBER 1990

Table 1. Selected physical and chemical characteristics of the Po-mona series soil.

Table 2. Physical and chemical properties of oxalic and formic acids.

PropertySand (%)Clay (%)Organic C (g/kg)pH (H20)PhosphorusTotal P (mg/kg)Total organic P (mg/kg)Mehlich I (mg/kg)0.01 MCaCl, (mg/kg)Distilled water (mg/kg)PotassiumMehlich I (mg/kg)CalciumMehlich I (mg/kg)MagnesiumMehlich I (mg/kg)AluminumAcid oxalate (mg/kg)Pyrophosphate (mg/kg)Mehlich I (mg/kg)1MKC1 (mg/kg)0.01 MCaCl2 (mg/kg)Distilled water (mg/kg)IronAcid oxalate (mg/kg)Pyrophosphate (mg/kg)Mehlich I (mg/kg)0.01 MCaCl2 (mg/kg)Distilled water (mg/kg)

A931

17.74.03

21.89.65.17.09.1

13

111

30

226215

3715102

1235320.7

Horizon

Bh894

21.94.23

65.549.27.62.00.4

3

13

4

135713794321589713

619220.6

Bt

78183.14.74

138.627.19.71.00.1

2

14

8

1075769230118423

9892250.8

Oxalic acid Formic acid

termined using flame emission spectrophotometry with aN2O-C2H2 flame. In selected samples, Fe, K, Ca, Na, Mg,Zn, Mn, and Cu were also determined by inductively cou-pled plasma emission spectrophotometry.

Inorganic P in the nitrate was determined by a molyb-denum blue colorimetric procedure using ascorbic acid as areductant (Murphy and Riley, 1962). This is an operationaldefinition of inorganic P, since molybdenum hydrolysessome organic P compounds (Stainton, 1980; Tarapchak etal., 1982).

Total P was also measured in each filtrate. A 20-mL ali-quot of the filtrate was evaporated to dryness at 100 °C ina 50-m Pyrex beaker. The breaker was then placed in a mufflefurnace overnight at 500 °C. The beaker was allowed to cooland was then placed on a hot plate. A 10-mL aliquot of 4.8M HC1 was added to each sample and evaporated to dryness.This was followed by 5 mL of 12 M HC1, which was alsoevaporated to dryness. The beakers were allowed to cool andthe sample was redissolved in 20 mL of 0.1 M HC1. Phos-phorus in solution was again measured by the Murphy-Rileyprocedure. National Bureau of Standards (NBS) pine tissuesamples were used to check the accuracy of the total-P pro-cedure.

Soluble organic P was operationally defined as the differ-ence between Murphy-Riley P in the dry-ashed and acid-digested sample (total P) and Murphy-Riley P in the initialleachate (inorganic P).

Organic-Anion AnalysisLow-molecular-weight aliphatic organic anions in the fil-

tered solutions were analyzed by isocratic high-performanceliquid chromatography (HPLC) (Lee and Lord, 1986). TheHPLC system consisted of a Gilson single-piston high-pres-

FormulaFormula wt.Dissociation constant (p/0Ligand (£)Stability constant (log A^J

HO2CCO2H90.041.23; 4.19H2L6.10

HCO2H46.03

3.75HZ,

1.36

sure pump and pressure module (Gilson Medical Electron-ics, Middleton, WI), a Rheodyne Model 7125 injection valvefitted with a 20 ^L injection loop (Rheodyne, Cotati, CA),a Hamilton PRP-X300 150 by 4.1 mm organic-acid column(Hamilton Co., Reno, NV), a Gilson Holochrom variable-wavelength ultraviolet (UV) detector and a Gilson comput-erized integrator. Organic anions were separated at ambienttemperature with H2SO4 as the eluent at a flow rate of 2 mL/min and detected at 210 nm. Organic anions were quantifiedby comparing peak areas with an external standard curve.

Statistical AnalysisStatistical differences in the pattern of P-desorption ki-

netics were tested using a regression procedure. The observedrelationship between desorption and time (t) suggested a cu-bic equation of the form

P desorbed = ft, + Pit + 02t2 + [1]where /30, /?„ /32, and /33 are regression coefficients and t istime. Selected pairs of lines were compared, utilizing theprinciple of conditional error (Swindel, 1970). The differencein error sums of squares between one full and two reducedmodels were compared using an F test. The full model in-cluded the data common to the two groups being compared,while the two reduced models included only the data fromeach individual group. The test statistic for this procedureis given by

F = (SSE* - SSE)/(f* - f)SSE// [2]

where SSE* is the error sums of squares for the full or com-mon regression; /* is the error degrees of freedom for thefull or common regression; SSE is the pooled error sums ofsquares for the reduced models; and / is the pooled errordegrees of freedom for the reduced models. This test statisticis compared with an F value with (f* — f) degrees of freedomin the numerator and/degrees of freedom in the denomi-nator. This is essentially a multivariate procedure that testsfor differences in the overall regression rather than differ-ences between individual pairs of data points.

Phosphorus-desorption data were fitted to the cubic equa-tion using the REG procedure of the Statistical Analysis Sys-tem (SAS Institute, 1982). All differences stated as significantand discussed in the text are significant at the P < 0.05 level.

RESULTSInorganic Phosphorus

In the A horizon, there were no significant differ-ences in inprganic-P release among the oxalate solu-tions and distilled water. Desorption of inorganic P informate was lower than in the other extracting solu-tions. In general, inorganic-P desorption in all the ex-tracting solutions increased rapidly for the first 6 to12 h. After 24 h, solution inorganic P declined slightly(Fig. la).

In the Bh horizon, all extracting solutions produceda significantly different effect (Fig. Ib). Inorganic-P re-

FOX ET AL.: KINETICS OF PHOSPHORUS RELEASE FROM SPODOSOLS 1443

20 40 60 80

Time (h)Fig. 1. Release of inorganic P into 0.1 mM and 1.0 mM oxalate,

1.0 mM formate, and distilled-water solutions adjusted to pH 4.3for the (a) A horizon, (b) Bh horizon, and (c) Bt horizon materialfrom a Pomona series soil. (Note differences in ordinal scale.)

lease was greater in the oxalate solutions than in water.Between two and three times more inorganic P wasdesorbed with the 1.0 mM oxalate solution than withthe 0.1 mM oxalate. In both oxalate solutions, inor-ganic-P desorption increased rapidly for 3 to 6 h andthen immediately decreased again. The point where Pdesorption began to decline coincides with the com-plete sorption of the added oxalate (Fig. 2b). Phos-phorus desorption in water declined monotonicallyduring the first 6 h, while inorganic-P desorption inthe formate solution was significantly less than inwater.

The general pattern of inorganic-P desorption in theBt horizon was similar to that in the Bh horizon. Inthe Bt horizon material, water and formate releasedthe same levels of P, while release by the two levelsof oxalate statistically differed from these and fromeach other (Fig. Ic). There was a short period whereinorganic-P release increased with the 1.0 mM oxalatesolution, which continued until all the added oxalatewas sorbed (Fig. 2c). In the Bt horizon, the 1.0 mMoxalate solution caused a 9- to 10-fold increase in Pdesorption. Phosphorus desorption in the 0.1 mM ox-alate was slightly greater than in water or formate,where P desorption was negligible.

Organic PhosphorusAppreciable amounts of organic P were released in

the A and the Bh horizons, but not in the Bt horizon(Fig. 3a, b, c). Organic P composed some 10 to 30%of the total P released in the A horizon (Fig. 3a) and30 to 70% in the Bh horizon (Fig. 3b). The overall

COo> 1'0.8-0.6-0.4-0.2'

n i

f f m

(c)'•••...

m • ,• ————— r—— • ———— , ———— •——20 40

Time (h)60 80

Fig. 2. Disappearance from solution of 0.1 mM and 1.0 mM oxalateand 1.0 mM formate added to the (a) A horizon, (b) Bh horizon,and (c) Bt horizon material from a Pomona series soil.

20 40 60 80

Time (h)Fig. 3. Release of organic P into 0.1 mM and 1.0 mM oxalate, 1.0

mM formate, and distilled-water solutions adjusted to pH 4.3 forthe (a) A horizon, (b) Bh horizon, and (c) Bt horizon materialfrom a Pomona series soil.

pattern of organic-P desorption in the various solu-tions was similar in the A and Bh horizons. Desorption

1444 SOIL SCI. SOC. AM. J., VOL. 54, SEPTEMBER-OCTOBER 1990

of organic P increased gradually and then leveled offafter 6 to 12 h. Initially, organic-P desorption in bothoxalate solutions was greater than in water and, in the1.0 mM solution, it was greater than in the 0.1 mMsolution. Organic-P desorption in formate was lessthan in water for both horizons. In the A horizon,release of organic P by the two oxalate solutions de-creased sharply after 24 h, whereas the decrease inorganic P was small and gradual in the rest of thesolutions.

Oxalate and Formate SorptionApproximately 1 mM/kg of oxalate was sorbed in

the first 12 h from both oxalate solutions in the Ahorizon (Fig. 2a). This represents all the added oxalatein the 0.1 mM solution and one-tenth that in the 1.0mM solution. Formate was not sorbed at all duringthis same period. After 12 h, the remaining oxalateand formate were rapidly degraded. The term degra-dation is used to describe the abrupt drop in solutionconcentrations of organic acids after 24 h, since it isunlikely that this is an adsorption reaction.

In the two subsurface horizons (Fig. 2b, c), all theadded oxalate disappeared within 6 h, with the greatbulk of it gone within 3 h. None of the added formatedisappeared during this same time. After 12 to 24 h,however, the formate disappeared rapidly, presumablyby degradation of the compound.

AluminumAmong the three horizons, Al release increased in

the order Bh horizon > Bt horizon > A horizon (Fig.

0.8-06-

0.4-

0,?-

£»

1.0mMcOxalate0.1 mM,O)ialateLOmMformate

Water

-e"-~°

(a)

^^

*-^""~^t80

jg> 1.0'o 0.8

^0.6-

•g 0.4

I °'2

iS3 0.0 0 20 40 60 80

1.0

0.8-

0.6-

0.4-

0.2-

0.0

(c)

20 40 60 80

Time (h)Fig. 4. Release of Al into 0.1 mM and 1.0 mM oxalate, 1.0 mM

formate and distilled water adjusted to pH 4.3 for the (a) A ho-rizon, (b) Bh horizon, and (c) Bt horizon material from a Pomonaseries soil.

4). The overall pattern of release was, however, thesame in each horizon. Release of Al in water was con-sistently small. The release of Al in 1.0 mM oxalatesolution was greater than in 0.1 mM oxalate or for-mate. In each of the organic-acid solutions, there wasa short period of rapid Al release lasting approximately3 h. In the 1.0 mM oxalate solution, this was followedby an extended period of gradual Al release. The Allevels in the formate solution began to decline after12 to 24 h. The Al levels in the oxalate solutions alsodeclined but, for the most part, not until after 48 h.

Solution pHIn the A horizon, the pH of the oxalate solutions

and water were initially similar, while the pH of theformate solution was lower (Fig. 5a). All were stablethrough 12 to 24 h, after which the pH of the organic-acid solutions rose sharply, corresponding with thedisappearance of the organic anions (Fig. 2).

In the subsoil (Fig. 5b, c), the pH of the oxalatesolutions was greater than water and tended to increaseslowly, while there was a gradual decrease in pH withtime in the water extract. As in the A horizon, the pHof the formate solution was lower than water until 24h. At the point where the formate began to degrade,the pH of the solution again rose sharply.

DISCUSSIONThe results suggest that the mechanism controlling

inorganic-P release in the A horizon differed from that

20 40 60 80

Time (h)Fig. 5. Temporal changes in solution pH of 0.1 mMand 1.0 mM

oxalate, 1.0 mM formate, and distilled water extracts initiallyadjusted to pH 4.3 from the (a) A horizon, (b) Bh horizon, and(c) Bt horizon material of a Pomona series soil. The first pointplotted is for 0.5 h.

FOX ET AL.: KINETICS OF PHOSPHORUS RELEASE FROM SPODOSOLS 1445

in the Bh and Bt horizons. In light of this, the surfaceand subsurface soils are discussed separately. The dis-cussion focuses on the effects of and mechanisms ofP release by oxalate, and the implications of these datato forest nutrition in the lower Coastal Plain.

A Horizon — Inorganic PhosphorusPrevious work with surface soil material from a wide

variety of Haplaquods has shown that they have vir-tually no ability to sorb P and, in consequence, P fer-tilizers are rapidly leached from the surface (Ballardand Fiskell, 1974). These data show that inorganic andorganic P are dominantly present in a water-solubleform. This conclusion is supported by the facts thatoxalate had no effect on P release, that oxalate wasminimally consumed during the time when most Pwas released to solution, and that the pH remainedstable during the time of P release. Ligand exchangeor dissolution of metal-P compounds would havecaused changes in at least one of these parameters.

We attribute the rapid drop in oxalate and formateconcentrations after 12 h to microbial degradation ofthe organic anions, rather than to sorption. Simpleorganic anions serve as the substrates for growth ofmicroorganisms (Harder, 1973; Hodgkinson, 1977).McColl et al. (1990) have observed rapid degradationof organic acids in soil solution, which they attributedto both biotic and abiotic pathways. Although toluenewas added to retard microbial growth in the samplesof this study, microbial activity was only inhibited fora short while. O'Keefe et al. (1987) observed significantactivity of both bacteria and fungi after 24 h in similarsoils treated with toluene. Hodgkinson (1977) sum-marized a variety of chemical degradation pathwaysfor both oxalate and formate. The oxidation of organicanions tends to consume H ions, which would causethe solution pH to rise. As seen in these data, a pHincrease accompanied the disappearance of oxalateand formate from solution.

The gradual drop in the concentration of inorganicP after 24 h was interpreted as the slow formation ofcolloidal Al phosphates. Chemical speciation with thecomputer program GEOCHEM (Sposito and Matti-god, 1980) supported this statement. While these datado not match the equilibrium assumptions inherentin GEOCHEM, it is still useful to investigate the plau-sible directions of reactions in this system (Traina etal., 1987). Calculated ion-activity products (IAP)showed that there was a tendency for the solutions tobecome saturated with respect to amorphousA1(OH)2H2PO4. The log values of the IAP forA1(OH)2H2PO4 in the oxalate solutions were stable at28 to 29 between 6 and 24 h, suggesting undersatur-ation with this solid phase (log^Tsp = 27-29; Colemanet al., 1960; Veith and Sposito, 1977). At 48 h, thecalculated log IAP for A1(OH)2H2PO4 dropped to ap-proximately 26, suggesting that the solution was thensaturated with respect to this solid phase. It was at thistime that the levels of P in solution began to drop,probably in association with the formation of a col-loidal A1(OH)2H2PO4 solid.

In the A horizon, the overall lower P release in theformate solutions can also be attributed to the for-

mation of Al phosphate. Since formate does not formvery stable complexes with Al (Table 2), colloidal Al-P may form and reduce the apparent release of P. Thismechanism was supported by GEOCHEM calcula-tions, which indicated saturation with respect to amor-phous A1(OH)2H2PO4, even at 3 h.

Bh and Bt Horizons — Inorganic PhosphorusWhile the exact nature of the reactive surfaces in

the Bh and Bt horizons is unknown, both contain sur-faces that are dominated by Al and strongly sorb largeamounts of P (Ballard and Fiskell, 1974). Since P re-lease was greater in the more concentrated oxalate so-lution, and low in both formate and water, the abilityto form stable complexes with Al was important forthe release of P from the subsoil. Numerous studieshave reported similar results (Swenson et al., 1949;Deb and Datta, 1967; Lopez-Hernandez et al., 1979;Comerford and Skinner, 1989).

These kinetic data suggest that ligand exchange isthe mechanism for the rapid P release observed inthese Bh and Bt horizons. Ligand exchange of oxalatefor P should be accompanied by a rapid sorption ofoxalate and an increase in solution pH (Goldberg andSposito, 1985). Furthermore, one would expect theconclusion of P release to closely coincide with totaldisappearance of oxalate from solution. All these ex-pected results are present in the data from these soilmaterials. The rapid disappearance of oxalate must beattributable to sorption phenomena and not degra-dation. Formate can be viewed as a control for deg-radation of the organic anions in these experimentsbecause, in the A horizon, oxalate and formate de-graded in the same time frame and neither was sorbed.In these horizons, formate was not sorbed yet did notdegrade until after the P was released and oxalate wasgone.

As before, GEOCHEM was used to investigate theeventual drop in solution inorganic-P levels in the Bhhorizon. The calculated log IAP of A1(OH)2H2PO4 wasapproximately 26 in the Bh horizon at 12 h, showingsaturation with respect to this previously mentionedsolid phase.

Organic PhosphorusSoluble organic P is not often included in studies of

soil P, even though it has been recognized as an im-portant component of soil solution P (Anderson,1980). These data show that the same arguments ap-plied to inorganic P above apply to organic P. Of mostinterest is that oxalate was able to access pools of or-ganic P in the Bh horizon that the water was not. Thespeciation of the organic P is not known for these soils,but a number of organic-P compounds are known toform metal-organic P complexes, much the same asinorganic P (Martin, 1970; Anderson et al., 1974).Since the forms of organic P in these soils is unknown,the mechanism by which oxalate influences organic Pis uncertain. Additional work is needed to identify theforms of organic P in these soils, their availability toplants, and the influence of organic anions on theirsolubility.

1446 SOIL SCI. SOC. AM. J., VOL. 54, SEPTEMBER-OCTOBER 1990

The presence of large amounts of soluble organic P,between 10 and 30% of the total pool of P released inthe A horizon and up to 70% of that released from theBh horizon, strongly suggests that organic P cycling inthe Bh horizon is an important pathway in the P nu-trient cycle within the soil.

Implications for Tree NutritionIt is well established that subsoil horizons can be

important to the P nutrition of plants (Comerford etal., 1984). It is not well established, however, that itis important in Sppdosols of the southeastern lowerCoastal Plain. Previous work in this region has shownthat P availability often limits growth of southern pinegrowing on Spodosols of the flatwoods (Pritchett andComerford, 1983). In addition, Van Rees and Com-erford (1986) have shown that fine roots of slash pinecan proliferate in the argillic horizon of Spodosols,while Neary et al. (1990), using a mechanistic nutrient-uptake model, suggested that the subsurface horizonsmay contribute a substantial portion of the P takenup by a young forest stand. This latter report did not,however, incorporate the increased P availability thatmight result from the presence of oxalate.

The results of this study add evidence to furthersupport the potential of P supply from subsoil hori-zons. In summary, we have demonstrated that oxalatecan greatly influence the release of inorganic and or-ganic P from the spodic and argillic horizons, but hasno effect on P in the surface soil. We have presentedempirical evidence from the kinetic data to suggestmechanisms of release. All these show the potential ofoxalate to supply P to a forest stand from an otherwisesparingly soluble source of P in subsoils. At the sametime, these data indicate that soil solution P replen-ishment in the A horizon will have to come frommineralization of organic P or P leaching from theoverlying forest floor.

ACKNOWLEDGMENTSThe authors thank Union Camp Corp. for primary fund-

ing, as well as the Intensive Management Practices Assess-ment Center (USDA Forest Service) and Mclntyre-Stennisfor subsidiary funding for this study.

JAYAWEERA & MIKKELSON: AMMONIA VOLATILIZATION FROM FLOODED SOIL SYSTEMS: I. 1447