JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hutchison, K. J.
Right arrow Articles by Hesterberg, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hutchison, K. J.
Right arrow Articles by Hesterberg, D.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hutchison, K. J.
Right arrow Articles by Hesterberg, D.
Related Collections
Right arrow Redox Processes
Right arrow Phosphorus
Right arrow Other Environmental Contamination
Right arrow Nutrient Management
Right arrow Water Pollution
Published in J. Environ. Qual. 33:1793-1802 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Surface Water Quality

Dissolution of Phosphate in a Phosphorus-Enriched Ultisol as Affected by Microbial Reduction

Kimberly J. Hutchison* and Dean Hesterberg

Department of Soil Science, Box 7619, North Carolina State University, Raleigh, NC 27695-7619

* Corresponding author (kim_hutchison{at}ncsu.edu).

Received for publication July 20, 2003.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phosphorus dissolution often increases as soils become more reduced, but the mechanisms are not fully understood. The objectives of this research were to determine rates and mechanisms of P dissolution during microbial reduction of a surface soil from the North Carolina Coastal Plain. Duplicate suspensions of silt + clay fractions from a Cape Fear sandy clay loam (fine, mixed, semiactive, thermic Typic Umbraquult) were reduced in a continuously stirred redox reactor for 40 d. We studied the effects of three treatments on P dissolution: (i) 2 g dextrose kg–1 solids added as a microbial carbon source at time 0 d; (ii) 2 g dextrose kg–1 solids split into three additions at 0, 12, and 26 d; and (iii) no added dextrose. After 40 d of reduction, concentrations of dissolved reactive phosphorus (DRP) were similar for all treatments and increased up to sevenfold from 1.5 to 10 mg L–1. The initial rate of reduction and dissolution of DRP was significantly greater for the 0-d treatment. A linear relationship (R2 = 0.79) was found between DRP and dissolved organic carbon (DOC). Dissolved Fe and Al and pH increased, suggesting the formation of aqueous Fe– and Al–organic matter complexes. Separate batch experiments were performed to study the effects of increasing pH and citrate additions on PO4 dissolution under aerobic conditions. Increasing additions of citrate increased concentrations of DRP, Fe, and Al, while increasing pH had no effect. Results indicated that increased dissolved organic matter (DOM) during soil reduction contributed to the increase in DRP, perhaps by competitive adsorption or formation of aqueous ternary DOM–Fe–PO4 or DOM–Al–PO4 complexes.

Abbreviations: CBD, citrate–bicarbonate–dithionite • DOC, dissolved organic carbon • DOM, dissolved organic matter • DRP, dissolved reactive phosphate • FAAS, flame atomic absorption spectrometry • LDPE, low-density polyethylene • XANES, X-ray absorption near-edge structure


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
SOME AGRICULTURAL LANDS located in areas of intensive livestock farming have received P applications in excess of the quantity of P removed by crop harvest (Mikkelsen, 1997), resulting in elevated soil P concentrations. Accumulation of P in surface soils has increased to the extent that the loss of P in subsurface drainage waters and in runoff discharged to surface waters has become a management concern because of P contributing to the eutrophication of surface waters (Beauchemin et al., 1998; Breeuwsma and Schoumans, 1987; Federico et al., 1981; Sallade and Sims, 1997a; Pierzynski et al., 1994). Many of the dissolution and transport mechanisms in P-enriched soils are not understood well enough to predict P loss quantitatively. Soil oxidation–reduction (redox) potential has been negatively correlated with P dissolution, where soils with lower redox potentials (more reduced conditions) may show enhanced P dissolution (Patrick and Khalid, 1974; Holford and Patrick, 1981; Willet, 1989; Vadas and Sims, 1998; Phillips, 1998). Therefore, there is a concern for increased risk of P loss from poorly drained (more reduced) soils that contain elevated levels of P.

Dissolution of P under reducing conditions in flooded soils and sediments has been related to the reductive dissolution of Fe(III)-oxides (Holford and Patrick, 1981; Jensen et al., 1998; Maguire et al., 2000; Phillips, 1998; Sallade and Sims, 1997b; Willet, 1989). Sallade and Sims (1997b) correlated oxalate-extractable Fe with dissolved P after sediments were flooded under anoxic conditions for 21 d. Operationally defined chemical fractionation of solid phase inorganic P in sediments (Sallade and Sims, 1997b) and in soil (Maguire et al., 2000) suggested that P associated with Fe and Al were the predominant forms of P. In general, changes in Fe-oxide mineral solubility or the relative distribution of PO4 between Fe-oxide (redox active) and Al-oxide (non-redox active) minerals could influence PO4 dissolution during reduction. X-ray absorption near-edge structure (XANES) spectroscopy is a more direct method for differentiating phosphate associated with Fe(III) or Al(III) minerals (e.g., strengite [FePO4·2H2O] and variscite [AlPO4·2H2O]), or phosphate sorbed to Fe- or Al-oxide mineral surfaces (Hesterberg et al., 1999; Khare et al., 2004; Beauchemin et al., 2003).

While there seems to be agreement on the importance of iron oxides on P sorption capacity of soils, it seems to be at variance with increased P dissolution during soil reduction. For example, Khalid et al. (1977) showed a decrease in dissolved PO4 under reducing conditions for various soils incubated for 15 d, while Holford and Patrick (1981) observed fluctuations in dissolved PO4 and Fe over a 45-d reduction period for a silty loam soil. These studies showed that dissolution of Fe and PO4 may vary over time when soils are anoxically incubated.

The rate of microbial reduction of a soil is a function of the rate of oxidation of organic carbon by soil microbes and depends on temperature, carbon source, presence and type of electron acceptors, and the microbial population present in the soil (Coyne, 1999, p. 158–169). These parameters would be different for aerobic soils, poorly drained soils (under sustained anoxic conditions), and soils that are rapidly flooded by irrigation water or animal waste effluent. For example, microbial activities will vary between aerobic or anaerobic soils. In aerobic soils, aerobes and facultative anaerobes will use oxygen as a terminal electron acceptor until it is depleted. Reduction of other electron acceptors, such as NO–3 and Mn- and Fe-oxides, may not occur unless the soil becomes anaerobic (Ehrlich, 1995). Also, a decreased efficiency of organic matter oxidation under anaerobic conditions causes production of water-soluble metabolites and a diminished production of CO2(g) (Ehrlich, 1995; Jeon and Park, 2000), resulting in increased concentrations of DOM (Fiedler and Kalbitz, 2003). Increased concentrations of DOM, sorption of DOM on mineral surfaces, and dissolution of Fe-oxides may affect the solid-phase speciation and dissolution of phosphorus as a soil becomes reduced. These changes could result in different PO4 dissolution mechanisms and rates (Roden and Edmonds, 1997; Willet, 1985).

Given that pH and DOM may increase during microbial reduction of soils, proposed P dissolution mechanisms include (i) reductive dissolution of Fe(III) minerals with associated PO4, (ii) competitive adsorption of DOM and PO4 by ligand exchange on mineral surfaces, (iii) DOM-enhanced dissolution of surface Fe or Al with concomitant release of PO4, (iv) formation of aqueous ternary DOM–Fe–PO4 or DOM–Al–PO4 complexes, and (v) decreased PO4 sorption with increasing pH. Reductive dissolution of Fe(III)-associated PO4 includes dissolution of Fe(III) oxides (Roden and Edmonds, 1997) with adsorbed, occluded, or coprecipitated P, and dissolution of Fe(III) phosphates such as strengite (FePO4·2H2O) (Lindsay, 1979; Willet, 1985; Miller et al., 1993). It is documented that carboxylate anions (a common functional group of many plant root exudates and soil organic matter) compete with P for sorption sites on soil minerals (Earl et al., 1979; Lopez-Hernandez et al., 1986; Violante et al., 1991). Carboxylate anions can also solubilize Fe and Al from mineral surfaces (Gerke, 1993). As surface Fe or Al is released through reductive dissolution or complexation with organic acids, aqueous binary DOM–Fe or DOM–Al complexes may form (Gerke, 1997). These complexes may enhance the dissolution of phosphate by forming ternary, aqueous complexes of phosphate bound to DOM through Fe and Al bridges (Gerke and Hermann, 1992). For soils receiving inputs of readily metabolizable organic carbon (e.g., from animal waste amendments), P dissolution mechanisms involving DOM may be particularly important during reduction. Soil pH typically rises during reduction and may increase after amendment with alkaline wastes, such as poultry litter (Vadas and Sims, 1998). Phosphate sorption on oxide minerals decreases with increasing pH (Oh et al., 1999), although this change is not great under acidic conditions. Also, DOM typically increases with increasing pH due to deprotonation of acidic functional groups (Swift, 1996), which may increase the dissolution of phosphate as described above.

To quantitatively predict P dissolution and transport in reduced soils, such as those amended with high organic carbon inputs (e.g., from livestock waste), a better understanding is needed of P dissolution mechanisms during microbial reduction of soils. The objectives of this study were to determine (i) how the rate of microbial reduction affects P dissolution from the surface horizon of an acidic, P-enriched Coastal Plain soil and (ii) the relative importance of various P dissolution mechanisms during the reduction of this soil.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil Collection and Handling
Samples were collected from the surface horizon (0–7 cm) of a Cape Fear sandy clay loam soil in a cattle-grazed pasture on the NCDA Tidewater Research Station near Plymouth, North Carolina. The pasture had received past applications of swine lagoon effluent (exact history unknown). The soil samples were double-bagged in low-density polyethylene bags (LDPE), placed in 4-L Pyrex glass jars on ice, and transported to the lab. To reduce the possibility of photochemical redox reactions and changes in soil properties, all sample handling in the laboratory was done in a glove box under a N2(g) atmosphere and in the absence of ultraviolet radiation (using a red-filtered safe light with an emission spectrum of 760 to 630 nm wavelength). The moist soil samples were sieved to <2 mm through a stainless steel sieve, thoroughly mixed in an LDPE bucket, and stored in the dark at 4°C in 4-L Pyrex glass jars.

Soil Characterization
Total P was determined on samples of the whole soil, silt + clay fraction (each 0.25 g dry wt.), and sand fraction (1.5 g dry wt.) using the acid digestion method of Kuo (1996) with the following exceptions: a total of 3 mL of HF was added and unreacted F– was complexed with saturated boric acid. Mehlich-3 P and inorganic forms of P such as NH4Cl-P, NH4F-P, and NaOH-P were determined on the whole soil (Pierzynski, 2000). Dissolved reactive P was determined in filtrates by the molybdenum blue method (Olsen and Sommers, 1982).

Acid ammonium-oxalate (dark)-extractable Fe and Al (Feox, Alox) and citrate–bicarbonate–dithionite (CBD)-extractable Fe and Al (FeCBD, AlCBD) were determined on the whole soil (0.25 and 0.5 g dry wt., respectively) and silt + clay fraction (0.25 g dry wt.) according to Jackson et al. (1986). All digests and extracts were filtered through 0.2-µm Isopore polycarbonate membranes (Millipore, Billerica, MA). Iron and Al were measured in the filtrates using flame atomic absorption spectrometry (FAAS). Soil pH was measured using a 1:5 soil to water ratio (Thomas, 1996).

To separate the silt + clay fraction and determine particle size distribution, the sand was wet-sieved through a 53-µm sieve. A subsample of the silt + clay fraction was further separated by centrifugation and oven-dried at 110°C before weighing (Dixon and White, 1999). Moisture contents of the whole soil and silt + clay fractions were determined gravimetrically on triplicate samples dried for 24 h at 110°C. Results of chemical analyses were calculated on a dry mass basis.

Reduction and Aerobic Control Experiments
Suspension Preparation
A preliminary experiment showed that continuously stirring a suspension of whole soil led to abrasion of the magnet and glass vessel by sand particles. Therefore, we used only the silt + clay fraction of the soil in the reduction experiments. The silt + clay fraction was separated by sonicating 200 g of soil in degassed, deionized water for 10-min intervals in an LDPE bucket, while purging the headspace with N2(g). Two sequential sonications were done with 4:1 and 2:1 water to soil ratios. The sand and silt + clay fractions were separated as described above, and brought into a stock suspension using degassed, deionized water (65 g silt + clay kg–1 suspension). Stock suspensions were stored for up to 5 d at 4°C before the start of each reduction experiment.

Reactor Design
The reduction and aerobic control experiments were conducted in a 1.5-L Cytostir (Kimble-Kontes, Vineland, NJ) glass bioreactor (with a suspended magnetic stirring device) wrapped with Tygon tubing connected to a temperature-controlled water bath to maintain the suspensions at 25°C. The overall reactor system was designed, in part, based on the batch reactor system of Patrick et al. (1973). Upstream from the reactor, N2(g) (0.2 or 0.5 L min–1) was split, and half was purged through 200 mL of a standardized 0.005 mol L–1 NaOH solution acting as a CO2(g) trap. To prevent evaporation of the suspension, the other half was water-saturated by passing through deionized water, and then into the reactor through a gas-dispersion tube. Effluent gas from the reactor was purged through a CO2(g) trap (200 mL of 0.005 mol L–1 NaOH). For an aerobic control experiment, zero-grade air [CO2(g) (0.0003 mol mol–1)] was purged through 2000 mL of 0.2 mol L–1 NaOH solution (0.2 L min–1) upstream from the reactor to trap residual CO2(g). Effluent gas from the reactor flowed through a 200-mL CO2(g) trap (0.005 mol L–1 NaOH).

The reactor vessel cap was modified to accommodate a gas dispersion tube as an inlet for N2(g), a sampling tube, and a port for a pH or Eh electrode. Periodic measurements of pH and Eh were taken directly in the silt + clay suspensions using calibrated combination pH and redox electrodes. The working condition of the redox electrode was determined by measuring the potentials of pH 4.0 and 7.0 buffer solutions containing 30 mg of quinhydrone (Microelectrodes, Bedford, NH). Measurements of Eh for the Ag/AgCl reference electrode were corrected to the standard hydrogen electrode (+199 mV) (Patrick et al., 1996).

Suspension Treatments and Sampling Method
Duplicate reactors containing 600 mL (control treatment) or 1200 mL of soil silt + clay suspensions (65 g kg–1 solids) were simultaneously reacted for 40 d while purging with N2(g). We studied the effects of three treatments: (i) no added dextrose (control treatment); (ii) addition of 2 g dextrose kg–1 solids (D-glucose, anhydrous; ACS grade, Fisher Scientific, Hampton, NH) at time 0 d (0-d treatment); and (iii) three additions of 0.67 g dextrose kg–1 solids (at 0, 12, and 26 d) to yield a total addition of 2 g dextrose kg–1 solids (spiked-addition treatment). An aerobic control experiment was also performed with 2 g dextrose kg–1 solids added at time 0 d. Suspensions were sampled periodically (over 40 d), and the total volume of suspension decreased only 20% before the termination of the experiments. To prevent oxidation of Fe(II) during sampling, approximately 40 mL of suspension were removed from the reactor as four 10-mL subsamples using a glass syringe. Each of the four subsamples was transferred anoxically into four 15-mL evacuated Pyrex test tubes fitted with rubber stoppers, and centrifuged at 27000 x g for 10 min.

To prevent exposure to oxygen, supernatant solutions from the four centrifuged test tubes were filtered (and combined) through a 0.2-µm polycarbonate membrane in a glovebox under a N2(g) atmosphere. A vacuum was drawn through the receiver flask (beneath the filter) of a Millipore polycarbonate 47-mm filter holder while flowing N2(g) through the cover (above the filter). Iron(II) in a 2.5-mL subsample of filtrate was immediately complexed by adding phenanthroline reagent (0.1 mL of concentrated HCl, 1 mL of 0.1% 1,10-phenanthroline, 0.5 mL of ammonium acetate buffer solution, and 1 mL of degassed, deionized water; Clesceri et al., 1989; Olson and Ellis, 1982). The remaining filtrate was acidified and stored at 4°C for other analyses.

Chemical Analysis of Aqueous Reactor Samples
Within 24 h of complexing, Fe(II) was measured colorimetrically ({lambda} = 510 nm) in duplicate subsamples using a UV-visible spectrophotometer. Duplicate or triplicate filtrate samples were analyzed for DRP colorimetrically ({lambda} = 840 nm) using the molybdenum blue method (Olsen and Sommers, 1982). To avoid interference of absorbance readings due to varying concentrations of dissolved organic carbon (DOC), blanks of each filtrate sample were prepared by replacing the coloring reagents [i.e., 1,10-phenanthroline for Fe(II) analysis and molybdate "reagent B" for PO4 analysis] with an equal amount of deionized water and subtracting absorbance readings for the blanks from those of samples.

Sample filtrates were analyzed for DOC using a TOC analyzer. Total dissolved Fe and Al were measured on acidified samples using FAAS. Concentrations of CO2(g) (maximum 0.0025 mol L–1) were determined in 0.005 mol L–1 NaOH traps by titrating the NaOH solutions for unreacted alkali with standardized HCl after the addition of 1 mL of 1.5 mol L–1 BaCl2 to precipitate CO32– as BaCO3 (Anderson, 1982). Concentrations of reactor CO2(g) titrated downstream from the reactor were corrected by subtracting the concentration of CO2(g) in the inlet gas subsample captured upstream. Results of titrated CO2(g) for the reduction experiments should be considered relative, as the CO2(g) traps may not have been 100% efficient due to the low molarity of the NaOH used to obtain greater sensitivity in titration.

X-Ray Absorption Near-Edge Structure Spectroscopy
Phosphorus K-XANES analysis was used to qualitatively determine whether Fe(III)-associated P was present in the silt + clay sample used in the reduction experiments. A moist sample (time 0 d) was analyzed at Beamline X19A (National Synchrotron Light Source, Upton, NY) following methods and data normalization procedures described by Khare et al. (2004). The normalized data were then compared with data for standards of PO4 sorbed to ferrihydrite or noncrystalline Al-hydroxide collected in that study.

pH Experiment
A separate 48-h batch experiment was conducted to determine the effect of pH on DOC and dissolved PO4, Fe, and Al in silt + clay suspensions under aerobic conditions. Aliquots of a stirred, aqueous stock suspension of silt + clay were weighed into tared, 250-mL polycarbonate centrifuge bottles to yield 65 g kg–1 solids in the final 60-g sample. While stirring, duplicate samples were adjusted in random chronological order to pH values between 5.0 and 7.5 using standardized 0.5 mol L–1 KOH or HCl. The samples were brought to the final mass with deionized water and 0.5 mol L–1 KCl (to yield a 5 mmol L–1 K background) and equilibrated for 48 h by shaking at 25°C in a temperature-controlled water bath. Periodically, sample pH was adjusted and K+ concentration corrected. Redox potential was measured on each sample at time 0 and 48 h. After 48 h, suspension pH was measured and samples were centrifuged at 16000 x g for 20 min. Filtered (0.2 µm) supernatant solutions were analyzed as described above for reactors.

Citric Acid Experiment
A separate batch experiment was conducted to determine the effect of a complexing organic acid (citrate) on dissolved PO4, total Fe, and total Al in the silt + clay suspension under aerobic conditions. Aliquots of a stirred, aqueous stock suspension of silt + clay were weighed into tared, 30-mL polycarbonate centrifuge tubes to yield 10 g kg–1 solids in the final 30-g sample. While stirring, a 0.25 mol L–1 citric acid monohydrate (H3C6H5O7·H2O) solution neutralized with 0.75 mol L–1 KOH was added to each duplicate sample in random chronological order, yielding final input citrate concentrations between 0 and 500 mmol kg–1 solids. Samples were adjusted to pH 7.0 with 0.25 mol L–1 KOH or HCl, diluted with deionized water, and brought to final mass by the addition of 0.25 mol L–1 KCl solutions. Samples were equilibrated for 19 h by shaking periodically, and after the 19-h equilibration period, suspension pH was measured and samples were centrifuged at 22000 x g for 20 min. Redox potential was measured on each sample at 0 and 19 h. Filtered (0.2 µm) supernatant solutions were analyzed as described above for reactors.

Statistical Analysis
Statistical analyses were performed with SAS Version 8.2 (SAS Institute, 1999) using PROC CORR, PROC REG, or PROC GLM.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil Characteristics
Selected chemical and physical properties of the whole soil and silt + clay fraction used in these experiments, and inorganic P parameters for the whole soil, are shown in Table 1. Both samples were slightly to moderately acidic (pH 5.9 and 6.4). While the silt + clay fraction constituted 51% of the soil mass, results in Table 1 indicate that the silt + clay fraction accounted for about 85, 71, 100, 91, and 83% of the total P, Feox, FeCBD, Alox, and AlCBD, respectively. The Feox to FeCBD ratio for the silt + clay was 0.6, consistent with soils containing elevated amounts of poorly crystalline Fe-oxides due to active redox processes (Schwertmann, 1985).


View this table:
[in this window]
[in a new window]
  		Table 1. Selected properties and sequential extraction data of whole soil samples from the Cape Fear soil and the silt + clay fraction used in this study.{dagger}

 
Mehlich-3 P (related to plant available P) was sevenfold greater than what is considered to be optimum for plant growth and crop yields (Pierzynski, 2000), indicating that the soil had excess P loading. Chemical extraction results (Table 1) targeting "labile P" (NH4Cl-P), "Al-associated P" (NH4F-P), and "Fe-associated P" (NaOH-P) for the whole soil show extractable "Al-associated P" to be 1.8 times greater than extractable "Fe-associated P," which is consistent with other results for soils from the North Carolina Coastal Plain (Novais and Kamprath, 1978).

Reduction and Aerobic Control Experiments
Figure 1 shows changes in Eh, CO2(g) trapped in 0.005 mol L–1 NaOH, and pH for duplicate redox reactor suspensions subjected to the three dextrose treatments. The redox potential (Eh) decreased over time for the three anaerobic treatments, and a concomitant evolution of CO2(g) indicated that the Eh declined as a result of microbial oxidation of organic carbon (Coyne, 1999, p. 158–169). Declining Eh trends for the control and spiked-addition treatments were strongly correlated with each other (Pearson correlation coefficient; r = 0.98) and were significantly different (p < 0.01) from the 0-d treatment (Pearson correlation coefficient; r = 0.61 and 0.65), indicating that the rate of reduction for the 0-d treatment was significantly different. This is evident in Fig. 1 for the 0-d dextrose treatment, which showed a rapid decline in Eh from 340 to 99 mV between 0 and 7 d and subsequent stabilization at approximately 70 mV for the remainder of the 40-d reduction period. The overall average Eh for all treatments decreased from 410 ± 30 mV (initial) to 100 ± 30 mV (40 d). With declining Eh, the pH increased significantly (p < 0.05) toward neutrality (from 5.9 ± 0.2 to 6.7 ± 0.4), as has been observed when acidic soils are reduced (Patrick et al., 1996). There was no significant difference in cumulative, trapped CO2(g) (2.0 ± 0.3 mmol) between the three reduction treatments (Fig. 1). In contrast, results from the aerobic control experiment showed that cumulative trapped CO2(g) was 5.33 ± 0.09 mmol, Eh ranged from 339 ± 6 to 425 ± 15 mV, and pH decreased from 5.52 ± 0.04 to 5.4 ± 0.1 (data not shown).



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 1. Trends in redox potential (Eh), pH, and cumulative evolved CO2(g) over a 40-d anaerobic reduction period for silt + clay suspensions with control, 0-d, and spiked-addition treatments. Arrows denote times of dextrose addition, and error bars represent standard deviations for duplicate reactors.

 
Figures 2 and 3 show changes in the concentrations of DRP, DOC, Fe(II), total Fe, and total Al in duplicate redox reactor suspensions subjected to the three dextrose treatments, along with selected measurements for the aerobic control experiment. For all anaerobic treatments, DRP increased by up to sevenfold, with a grand mean ranging from 1.5 ± 0.2 mg L–1 at 0 d to 10 ± 2 mg L–1 after 40 d of reduction. Similarly, DOC increased up to fivefold, ranging from 34 ± 9 mg L–1 at 0 d to 155 ± 45 mg L–1 at 40 d. When data for all three treatments were combined, a significant (p < 0.001) linear relationship was found between DRP and DOC (R2 = 0.79, n = 34). For all three treatments, DRP and DOC increased at a rate consistent with the rate of decreasing Eh and increasing trapped CO2(g) (Fig. 1), implying that microbial reduction led to the changes in solution chemistry. These trends are particularly evident for the 0-d dextrose treatment between 0 and 15 d, where a sharp increase in DRP and DOC (Fig. 2B) corresponded with a sharp decrease in Eh (Fig. 1B). Dissolved reactive P and DOC then stabilized for the remainder of the 40-d reduction period for this treatment. Concentrations of DRP and DOC for the aerobic control treatment showed minor, but significant (p < 0.05 and p < 0.01, respectively) increases (0.86 ± 0.04 to 1.2 ± 0.1 mg L–1 and 17.2 ± 0.9 to 21.1 ± 0.5 mg L–1, respectively), and were less than initial concentrations of DRP and DOC for the anaerobic experiment with a 0-d treatment (Fig. 2B).



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 2. Trends in dissolved organic carbon (DOC) and dissolved reactive phosphorus (DRP) over a 40-d anaerobic incubation period for silt + clay suspensions with control, 0-d, and spiked-addition treatments. Included are data for DOC and DRP from an aerobic control experiment for the 0-d treatment. Arrows denote times of dextrose addition, and error bars represent standard deviations for duplicate reactors.

 


View larger version (27K):
[in this window]
[in a new window]
  		Fig. 3. Trends in dissolved organic carbon (DOC), Fe(II), total Fe (FeT), and total Al (AlT) over a 40-d anaerobic incubation period for silt + clay suspensions with control, 0-d, and spiked-addition treatments. Data for AlT at 26 and 40 d were not obtained due to insufficient sample volume. Included are data for FeT from the aerobic control experiment for the 0-d treatment. Arrows denote times of dextrose addition, and error bars represent standard deviations for duplicate reactors.

 
For all three reduction treatments, the concentration of dissolved Fe(II) increased with decreasing Eh and increasing trapped CO2(g) (Fig. 3). Combining data for all three treatments, significant linear relationships between DOC and molar concentrations of dissolved Fe(II), total Al, or total Fe (p < 0.001, 0.01, and 0.01, respectively) were found, with R2 values of 0.60, 0.30, and 0.23, respectively (correlation plots not shown). The same was true for significant linear relationships (p < 0.001) between DRP and molar concentrations of Fe(II), total Al, and total Fe, having R2 values of only 0.51, 0.41, and 0.36, respectively (correlation plots not shown). This latter result indicated that less variation in DRP could be explained by dissolved metals than by DOC. For the aerobic control experiment, concentrations of total Fe ranged from 0.20 ± 0.06 to 0.15 ± 0.02 mg L–1, and dissolved Fe(II) and total Al were not detected (<0.028 and 0.073 mg L–1, respectively). Note that there was not sufficient supernatant solution to analyze total Al in the 0-d treatment after 26 and 40 d of reduction due to loss of supernatant solution during filtration. Dissolved Fe(II) and total Fe for the spiked-addition treatment after 40 d of reduction were lower than those in the control and 0-d treatments. Measured concentrations of dissolved Fe(II) greater than total Fe in some samples may be due to interference in the colorimetric and/or FAAS measurement due to increased DOC levels (Loeppert and Inskeep, 1996).

X-Ray Absorption Near-Edge Structure Spectroscopy
Figure 4 shows the pre-edge region (low-energy side of the white line peak) of a P K-XANES spectra for PO4 sorbed on ferrihydrite or noncrystalline Al-hydroxide, and the silt + clay sample used in redox experiments. As reported by Khare et al. (2004), the distinct pre-edge feature can be seen for PO4 sorbed on ferrihydrite, whereas no pre-edge feature is evident for the noncrystalline Al-hydroxide. The spectrum for our silt + clay sample shows a pre-edge feature intermediate between those of the standards, indicating that PO4 was associated in part with Fe(III). The fact that the pre-edge feature in the silt + clay XANES spectrum was weak implies that PO4 was more likely adsorbed on Fe-oxide minerals, rather than occurring as an Fe(III)-phosphate mineral (Hesterberg et al., 1999; Khare et al., 2004).



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Normalized X-ray absorption near-edge structure (XANES) spectra comparing the pre-edge regions for a silt + clay sample versus standards of PO4 sorbed to ferrihydrite or noncrystalline Al-hydroxide.

 
pH Effects on Phosphorus Dissolution
Figure 5 shows changes in DOC, DRP, total Fe, and total Al as affected by pH under aerobic conditions (Eh = 535 ± 10 mV). Dissolved total Fe and total Al increased slightly as pH increased, while DOC and DRP exhibited a minimum at pH 6.5. Dissolved organic carbon increased significantly (p < 0.05) from a minimum of 35.8 ± 0.4 to 50 ± 2 mg L–1.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 5. Dissolved reactive phosphorus (DRP), dissolved organic carbon (DOC), total Fe (FeT), and total Al (AlT) as affected by pH for silt + clay suspensions reacted in a batch experiment for 48 h. Error bars represent standard deviations in measurements between duplicate samples.

 
Citrate Effects on Phosphorus Dissolution
Figure 6 shows results of the batch experiment on PO4 dissolution from the silt + clay as affected by citrate under aerobic conditions (Eh = 430 ± 20 mV). Dissolved reactive P, total Fe, and total Al all increased significantly (p < 0.001) with increasing additions of citrate; concentrations of DRP, total Fe, and total Al increased from 0.41 ± 0.01 to 3.70 ± 0.03 mg DRP L–1, 0.07 ± 0.01 to 7.9 ± 0.1 mg total Fe L–1, and 0.7 ± 0.3 to 11.1 ± 0.3 mg total Al L–1. Geochemical modeling (Gustafsson, 2004) of aqueous solution data (for all citrate inputs) for this citrate batch experiment predicted 69 to 99% of aqueous Fe(III) to be complexed as Fe(citrate)0, and 87 to 100% of aqueous Al(III) to be complexed as Al(citrate)0 and Al3–2. This suggests that if our reduced reactor suspensions contained complexing organic acids analogous to citrate, a significant proportion of dissolved Fe(III) and Al(III) would be present as DOM–Fe or DOM–Al complexes.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 6. Dissolved reactive phosphorus (DRP), total Fe (FeT), and total Al (AlT) as affected by increasing concentration of dissolved organic carbon (DOC) for silt + clay suspensions reacted with different inputs of citrate for 19 h at pH 6.93 ± 0.06. Error bars represent standard deviations in measurements between duplicate samples.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Results from redox reactor studies indicated that the rate of microbial reduction of the silt + clay suspensions was similar for the control and spiked-addition treatment, but more rapid for the 0-d treatment. While concentrations of DRP were similar for all anaerobic treatments after 40 d of reduction (grand mean = 10 ± 2 mg PO4 L–1), more rapid reduction in the 0-d treatment resulted in the maximum DRP concentration being reached at least 15 d earlier than for the other treatments. Because DRP did not increase as much in the aerobic control experiment, the results for the anaerobic treatments are not an artifact of using a continuously stirred reactor. These results suggest that phosphate dissolution may occur when soils become water-saturated soon after easily metabolizable carbon sources are applied. Increased dissolution of organic carbon, Fe(III) [calculated as total Fe – Fe(II)], Fe(II), and total Al during reduction was observed for all treatments except for Fe(III) in the spiked-addition treatment. Increased dissolution of Fe(III) may have resulted from enhanced dissolution of Fe from mineral surfaces by DOM. It is not clear why the concentration of Fe(III) is greatest in the anaerobic control treatment.

Elevated levels of aqueous Fe(II) were measured in reactor supernatant solutions (for all treatments) when Eh was between 120 and 150 mV (at pH 6.5 ± 0.4), indicating that the Fe(III)–Fe(II) redox couple was reached (Patrick et al., 1996). Both XANES analysis of the silt + clay fraction (Fig. 4) and chemical extraction data for a whole soil sample (Table 1) indicated that some PO4 was associated with Fe(III). Therefore, reductive dissolution of Fe(III) (e.g., Fe-oxide minerals) and associated PO4 is one possible mechanism explaining the increase in DRP in our reactor experiments (Holford and Patrick, 1981; Jensen et al., 1998; Maguire et al., 2000; Phillips, 1998; Sallade and Sims, 1997b; Willet, 1989). If reductive dissolution of Fe(III) from Fe-oxides and concomitant release of associated PO4 was the only PO4 dissolution mechanism, and no competing Fe(II) or PO4 uptake mechanisms are important, then one would expect the molar ratio of dissolved Fe(II) to PO4 to be greater than or equal to one. This hypothesis includes the possibilities that PO4 may be sorbed as either a monodentate (Persson et al., 1996) or bidentate (Tjedor-Tjedor and Anderson, 1990; Arai and Sparks, 2001) surface complex on Fe-oxides, and that surface Fe(III) ions other than those binding PO4 may also be reductively dissolved. Our observed ratios of dissolved Fe(II) to PO4 ranged from 0.1 to 0.3, suggesting that reductive dissolution of Fe(III) and associated PO4 was not the only operative mechanism.

The positive linear relationship between DOC and PO4 is consistent with mechanisms of competitive adsorption between DOM and PO4 for mineral surfaces, and the formation of aqueous ternary DOM–Fe–PO4 and DOM–Al–PO4 complexes. These mechanisms are consistent with the observed molar ratios of (total dissolved Fe + total Al) to DRP of 3.0 and 3.2 for the 0-d treatment at 15 d (when DOC and DRP were near maximum) and the control treatment at 40 d, respectively. However, for the spiked-addition treatment at 40 d the (total Fe + total Al) to DRP molar ratio was only 0.3, suggesting that the mechanism of competitive adsorption between PO4 and DOC (e.g., ligand exchange) may better explain the increase in DRP.

We hypothesized that an increase in DOC in our reactors resulted from either microbial metabolism of organic C (Japenga et al., 1992) or from increasing pH as the suspensions were reduced. Dissolved PO4, DOC, total Fe, and total Al at pH 7 from the aerobic pH batch experiment (Fig. 5) were 9-, 4-, 65-, and 155-fold lower, respectively, than dissolved concentrations of these constituents at the end of the anaerobic redox reactor experiments. These results indicated that the increase in pH during microbial reduction, in itself, had a minimal effect on increasing DOC or the dissolution of PO4, total Fe, or total Al. Increases in concentrations of DOC, DRP, Fe(III), and total Al in the reduced reactor solutions were more likely a result of increasing DOM generated by microbial reduction.

Dissolved organic matter is often composed of organic acids from root extracts, microbial extracts, and fulvic–humic macromolecules that contain an abundance of carboxylic acid functional groups (Swift, 1996). These carboxylate anions are known to compete with PO4 for surface binding sites (Bolan et al., 1994; Violante et al., 1991) and chelate Fe and Al (Berrow et al., 1982; Duff et al., 1963; Hue et al., 1986). Fox et al. (1990) showed that the extraction of 10 g of a Spodosol soil (Bh horizon) with 100 mL of 1.0 mmol L–1 oxalate resulted in the dissolution of 3.5% of the total inorganic soil P (2.1 mmol P kg–1). Berrow et al. (1982) showed that the microbial by-product 2-ketogluconic acid, derived from microbial glucose reduction, chelated 17 mmol Fe kg–1 from a poorly drained soil containing 646 mmol Fe kg–1. Duff et al. (1963) extracted 13% of Fe from ß-ferric oxide, 0.8% Fe from goethite, and 2% of aluminum from aluminum hydroxides by incubating 5 to 500 mg of these minerals for 8 to 14 d (25°C) with glucose and ketogluconic acid produced by bacteria. These results give further evidence that increases in DOM during microbial reduction induced dissolution of Fe(III), Al, and PO4 in our anaerobic reactors.

Aerobic experiments with citrate provided additional insights on DOM effects on PO4 dissolution. Citrate is a tri-carboxylic acid that is often found in soil solutions (excreted by plant roots), and it forms aqueous complexes with Fe(III) and Al(III) (e.g., log k = 11.5 and 10.9 for 1:1 Fe–citrate and Al–citrate complexes) (Martell and Smith, 1977; Gerke, 1997). Figure 6 shows that concentrations of DRP, total Fe, and total Al increased with increasing additions of DOC as citrate. At the maximum citrate input of 360 mg DOC L–1, the concentrations of DRP, total Fe, and total Al were significantly greater (p < 0.01) than concentrations of these constituents in the pH batch experiment (Fig. 5). The changes in DRP, total Fe, and total Al in Fig. 6 appear to be due to ligand exchange of citrate for adsorbed PO4 (Violante et al., 1991; Earl et al., 1979; Lopez-Hernandez et al., 1986) and citrate-induced dissolution of soil minerals or organic-matter-bound Fe and Al (Gerke, 1993). We propose that similar mechanisms contributed to enhanced PO4 dissolution from microbially generated DOM in our reduction experiments. Once PO4, Fe, and Al are in solution, aqueous DOM–Fe or DOM–Al complexes can bind phosphate (Bloom, 1981; White and Thomas, 1981; Gerke and Hermann, 1992; Gerke, 1993), potentially preventing it from partitioning back into the solid phase.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Dissolved reactive P increased up to sevenfold during microbial reduction of silt + clay from a Cape Fear soil sample following three different dextrose treatments. The maximum overall concentration of DRP (10 ± 2 mg PO4 L–1) was similar for all treatments after 40 d of reduction. Compared with the control or spiked-addition treatments, a single addition of dextrose caused a more rapid initial increase in DRP, and the maximum DRP concentration was reached approximately 15 d earlier. Our results indicated that multiple mechanisms contributed to phosphate dissolution during microbial reduction of the soil material studied. Although reductive dissolution of Fe(III)-oxide minerals may have increased DRP, a strong correlation between DRP and DOC (R2 = 0.79, n = 34) and Fe(II) to PO4 ratios of ≤0.3 indicated that P dissolution mechanisms involving increasing DOM produced during microbial reduction were more important. Specific mechanisms for enhanced P dissolution that could not be differentiated from our experiments include increased ligand exchange of DOM for mineral-adsorbed PO4; DOM-induced dissolution of Fe(III)- and Al(III)-oxide minerals with concomitant release of sorbed PO4, and formation of ternary aqueous DOM–Fe(III)–PO4 or DOM–Al(III)–PO4 complexes. Our laboratory results suggest that the addition of easily metabolizable carbon sources (e.g., animal waste) to Coastal Plain soils may enhance dissolution of phosphate under wet (reduced) soil conditions through reduction of Fe-associated P or interactions with increasing concentrations of DOM. Formation of ternary aqueous complexes of DOM–Fe(III)–PO4 or DOM–Al(III)–PO4 in these soils, as suggested by our data, could potentially increase phosphate leaching.


   ACKNOWLEDGMENTS
 
The authors are grateful to Dr. Suzanne Beauchemin for insightful discussion concerning chemical analyses and experimental approaches, Dr. Wolfgang Caliebe for help in collecting XANES data, Ms. Nidhi Khare for help with XANES data analysis and for providing standards, and Ms. Elizabeth Hutchison for laboratory assistance. Thanks are extended to Guillermo Ramirez for advice concerning sample analysis. Funding was provided by USDA NRI Grant 2001-35107-10179 and the North Carolina Agriculture Research Service.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 1589-1599. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
Z. He, C. W. Honeycutt, T. Zhang, and P. M. Bertsch
Preparation and FT-IR Characterization of Metal Phytate Compounds.
J. Environ. Qual., July 1, 2006; 35(4): 1319 - 1328.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
G. C. Murray and D. Hesterberg
Iron and Phosphate Dissolution during Abiotic Reduction of Ferrihydrite-Boehmite Mixtures
Soil Sci. Soc. Am. J., June 21, 2006; 70(4): 1318 - 1327.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hutchison, K. J.
Right arrow Articles by Hesterberg, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hutchison, K. J.
Right arrow Articles by Hesterberg, D.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hutchison, K. J.
Right arrow Articles by Hesterberg, D.
Related Collections
Right arrow Redox Processes
Right arrow Phosphorus
Right arrow Other Environmental Contamination
Right arrow Nutrient Management
Right arrow Water Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome