SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 25 August 2005
Published in Soil Sci Soc Am J 69:1658-1665 (2005)
DOI: 10.2136/sssaj2005.0068
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Litaor, M. I.
Right arrow Articles by Shenker, M.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Litaor, M. I.
Right arrow Articles by Shenker, M.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Litaor, M. I.
Right arrow Articles by Shenker, M.
Related Collections
Right arrow Phosphorus
Right arrow Soil Chemistry
Right arrow Wetland Soils

Wetland Soils

Sorption Characteristics of Phosphorus in Peat Soils of a Semiarid Altered Wetland

M. I. Litaora,*, O. Reichmannb, A. Haimb, K. Auerswaldc and M. Shenkerb

a Dep. of Biotechnology and Environmental Sciences, Tel-Hai College, Upper Galilee 12210, Israel
b Dep. of Soil Science, The Hebrew Univ. of Jerusalem, Rehovot, Israel
c Dep. of Grassland Science, Tech. Univ. Muenchen, Am Hochanger 1, D-85350 Freising-Weihenstephan, Germany

* Corresponding author (litaori{at}telhai.ac.il)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
We studied the adsorption characteristics of P in altered peat soils of the Hula Valley, Israel, which has undergone repeated drying and rewetting cycles. As a result, the water quality of Lake Kinneret, the only freshwater lake in Israel, may have been adversely affected. Peat sample collection was based on pedogenic evaluation of the wetland's history and on the redox potential of aerobic (Eh = 450 mV) and anaerobic (Eh < –220 mV) conditions. Extractable citrate–bicarbonate–dithionite iron (FeCBD) was a dominant mineral phase in the aerobic layers (29 ± 5 g kg–1). Mössbauer spectra suggested that hematite, goethite, ferrihydrite, and magnetite are the main Fe minerals in these peat soils. The sorption maximum (Smax) of the aerobic layers ranged from 670 to 1750 mg P kg–1, with a mean value of 1250 mg P kg–1, whereas the anaerobic layers ranged from 625 to 975 mg P kg–1, with a mean value of 775 mg P kg–1. The equilibrium phosphorus concentration (EPC0) values in the most anaerobic peat layers were several orders of magnitude higher (0.31 mg L–1) than in the aerobic layers (0.01 mg L–1). Only a weak to moderate correlation was found between the sorption attributes and the Fe content due to precipitation of Ca-P phases. The main source of Ca in these soils is gypsum. Rewetting of the peat soils leads to a decrease in Smax and the buffer capacity, and an increase in EPC0, which could lead to higher P mobility. The increased potential of P mobility declined with a concurrent increase in Ca-P precipitation due to enhanced dissolution of gypsum.

Abbreviations: CBD, citrate–bicarbonate–dithionite • Eh, redox potential • EPC0, equilibrium phosphorus concentration • ICP–AES, inductively coupled plasma–atomic emission spectroscopy • MBC, maximum buffer capacity • OM, organic matter • Smax, sorption maximum


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
FRESHWATER WETLANDS function as nutrient sinks, which efficiently process and store P, thus reducing the potential of eutrophication in downstream waters (Richardson, 1985; Sharpley and Rekolainen, 1997). The draining of Hula Lake and the elimination of its surrounding wetlands to increase the arable land in northern Israel during the mid 1950s removed this crucial nutrient sink. As a direct consequence of the drainage, the peat soils were rapidly oxidized and this, coupled with the continuous internal conflagration of the oxidized peat soils, led to a subsidence of more than 2 m below the original surface that eventually induced partial reflooding of the drained area. To reverse some of the negative consequences of the drainage of Hula Lake and the surrounding wetlands, a small 100-ha lake (Agmon) was constructed in 1994, covering the least agriculturally productive peat soils in the Hula Valley (Fig. 1) . The creation of the new lake elevated the water table in the center of the Hula Valley by at least 60 cm (Tsipris and Meron, 1998) and occasionally, even to the surface, creating highly reduced conditions (Shenker et al., 2005).



View larger version (53K):
[in this window]
[in a new window]
  		Fig. 1. General location of the study area: (a) a historical superposition showing the drained Hula lake and wetlands (drained in the 1950s) and (b) the recent drainage canals and new lake Agmon; (c) the location of the soils in the deep peat (circles) and shallow peat soils (triangles).

 
Although P is extremely important in regulating algal growth in Lake Kinneret (Serruya and Berman, 1976), relatively little attention has been paid to the adsorption mechanisms of P in these altered peat soils. Current management practices may have facilitated an increase in P discharge into Lake Kinneret, which in turn may lead to accelerated eutrophication, a major water-quality issue for the only freshwater lake in Israel, which provides 25% of the country's drinking water. Hence, understanding the adsorption characteristics of P, especially in the rewetted peat soils of the Hula Valley, is crucial to improving management strategies for the upper catchment of the Jordan River. Our working hypothesis stated that the rewetting of the previously oxidized peat soils would dissolve the Fe hydroxides, oxyhydroxides, and oxides (subsequently all termed oxides) and gypsum that were formed in these soils during the drying period. Consequently, the concentration of P in the soil solution of the rewetted peat soils would be regulated by the release of P from the dissolving Fe oxides, seasonal changes in redox conditions that affect the state of Fe oxide dissolution, or precipitation of the released P as Ca-P minerals. The objectives of this research were (i) to quantify the sorption characteristics of the rewetted peat soils and (ii) to study the mechanisms that regulate P solubility in these altered wetland soils.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Site
The Hula Valley is in the northernmost segment of the Jordan-Arava Rift Valley, and is approximately 70 m above sea level. The Hula Valley is 175 km2 in size and is currently drained by a system of artificial canals, which empty into the Jordan River at the southern end of the valley. Following the draining of the swamp, a surface layer of peat soils, with an average thickness of 4 to 6 m, was oxidized. The peat soils of the Hula Valley are predominantly Histosols ({approx}1860 ha). Classification on the basis of decomposition identifies them as Medifibrists, Medihemists, Medisaprists, and Conflagrated Histosols, and each is further classified according to the occurrence and quantity of CaCO3: without lime, with minimal lime, and with lime (Department of Agriculture, 1986).

On the basis of this soil classification and mapping, we divided the study area into two regions: (i) deep peat soils (>4 m) located north of Lake Agmon where no rewetting has occurred, and (ii) shallow peat soils (<1 m) located around Lake Agmon where rewetting has occurred (Fig. 1). In the deep peat area, soil was sampled by excavating three soil pits to depths of 2 to 4 m and performing additional auguring, in triplicate, of four soil cores to a depth of 1.2 m. In the shallow peat area, we augured six soil cores to a depth of 1.2 m. The exact locations of the soil pits and cores were determined according to the groundwater level and flow direction at each site in relation to major drainage canals (Fig. 1). Each soil core was divided into increments of 20 cm or according to diagnostic layers observed in the open pits. Soil samples below 45 cm, exhibiting Eh values of –200 ± 50 mV throughout the year, were considered anaerobic and were immediately stored in tightly sealed plastic bags. The soil samples were kept at –15°C until chemical analysis.

General Soil Characterization
Soil pH was determined in 1:10 soil/water suspensions using a glass electrode. Organic matter (OM) content was determined by the dry combustion method (Nelson and Sommers, 1982), and CaCO3 content was determined with a calcimeter apparatus (Nelson, 1982). Extractable Fe, Al, and Mn oxides were determined by the CBD method described by Jackson et al. (1986). The sesquioxide contents in selected samples were also extracted by the ammonium oxalate procedure described by Ross and Wang (1993). Extractable Fe, Al, and Mn were measured by inductively coupled plasma–atomic emission spectroscopy (ICP–AES; Spectro, Kleve, Germany).

The identification and quantification of the various Fe oxides in selected soil samples were derived from Mössbauer spectra taken at room temperature and at 4.2°C with spectrometers operating in transmission geometry with a sinusoidal or constant-acceleration movement of the source of 57Co in a matrix of rhodium and a {alpha}-Fe foil for calibration. The spectra were fitted with sets of split Gaussian distributions of Lorentzian-shaped lines for the magnetic sextets and quadrupole doublets, as described by Friedl and Schwertmann (1996).

The ferrimagnetism was quantified by measuring magnetic susceptibility with a Bartington Magnetic susceptibility meter MS2. The measured susceptibility was calibrated against magnetite (FeIIFeIII2O4) and is given in percent weight of magnetite as indicated from Mössbauer spectroscopy. The magnetic susceptibility was measured in the size fractions 2000 to 800, 800 to 63, 63 to 2, and < 2 µm, and the total magnetite abundance in the bulk sample was calculated from those fractions.

Sorption Experiments
Since the altered peat soils have undergone repeated cycles of drying and rewetting, which, according to our hypothesis, affect P adsorption properties of the soils, we conducted the sorption analysis under both aerobic and anaerobic conditions. The sorption characteristics of P in the Histosols were evaluated using common isotherm models. The experimental sorption data were fitted by the Langmuir model. The aerobic soil sorption experiments were performed in triplicate at 25°C, using a soil-to-solution ratio of 2 g to 25 mL of 0.01 M KCl (Richardson and Vaithiyanathan, 1995). The sorption experiments were conducted using P concentrations ranging from 0 to 70 (0, 0.5, 1, 2, 4, 6, 8, 10, 20, 40, and 70) mg L–1, similar to the conditions set for the sorption experiment in Nair et al. (1984). The suspensions were equilibrated for 24 h in polyethylene tubes on a reciprocal shaker, followed by centrifugation at 1500 g for 20 min. A 0.45-µm aliquot of filtered supernatant was analyzed for P using the ammonium-molybdate colorimetric method (Murphy and Riley, 1962).

The anaerobic sorption experiments were conducted with freeze-dried soil samples under N2 atmosphere with a partial pressure of O2 < 0.0002 atm. Oxygen was removed from all solutions by N2 bubbling for 2 h before each experiment.

The sorption capacity was calculated using the Langmuir isotherm model:

[1]
where Smax is the maximum amount of solute adsorbed (mg kg–1), S0 is the initial P present in the adsorbed phase, ST is the total amount of P adsorbed (mg kg–1), Ct is the concentration of P in the solution after 24 h equilibration (mg L–1), and k is a constant related to binding energy (L mg–1). The maximum buffer capacity (MBC) of the soil, which is the increase in adsorbed P per unit increase in the final solution P concentration at the lowest part of the isotherm (as Ct approaches zero), was estimated from the product of the Langmuir constant k and Smax (Holford, 1979). S0, Smax, and k were estimated using the SigmaPlot curve-fitting routine.

The equilibrium dissolved P concentration at which net P adsorption is 0 (adsorption equals desorption) is defined as EPC0. This concentration is assumed to represent P concentration in the soil solution, and, as such, it is of great environmental importance. We derived the EPC0 from the Temkin isotherm model because, unlike the Langmuir equation, it explicitly estimates the EPC0 as follows:

[2]
where a is a coefficient related to the buffering capacity and b is a constant for a soil in its actual state. This constant represents the EPC0 at which S = 0, thus Eq. [2] can be rewritten as follows:

[3]
The use of the Temkin equation to assess P sorption parameters was recently advocated by Barrow (1999) and Vandenhove et al. (1998). The parameters a and EPC0 were estimated using the SigmaPlot curvefitting routine.

Because of the potential precipitation of Ca-PO4 minerals during the adsorption experiment due to the abundance of Ca in the soil's interstitial water (Litaor et al., 2003), we tested the aerobic adsorption results by conducting a sequential extraction experiment on four peat soil samples that were first subjected to the adsorption experiment described above. After each P addition, the soil sample was rinsed with 25 mL double distilled water for 30 min in polyethylene tubes on a reciprocal shaker, followed by centrifugation at 1500 g for 20 min. Next, the sample was subjected to a modified Hedley sequential extraction, as described by Tiessen and Moir (1993). The amount of P added to each soil fraction during the sorption experiments was computed as the difference between a given sample fraction and its P content before the sorption experiment. The supernatants collected from these batch-sorption experiments were also analyzed for the concentrations of Fe, Al, Ca, Mg, N, K, H4SiO4, S, and P using ICP–AES. The pH and alkalinity in the supernatants were determined using standard methods.

The Eh in the peat soils was determined by two automated in situ redox stations installed in the deep and shallow peat areas. The stations, described in detail by Shenker et al. (2005), consist of Eh electrodes installed at three depths, above and below the water table. The electrodes are inserted into PVC sleeves that fit snuggly around the electrode body for protection, leaving the very tip of the electrodes exposed to the soil and interstitial waters. Each electrode calibration was performed after the electrode had been inserted into its protective sleeve. The Eh electrodes (Cole-Palmer ORP Redox combination electrodes) were calibrated with standard buffer solutions mixed with quinhydrone (0.2 g per 100 mL). The electrodes were connected to a data logger (Multilog, Fourier Systems, Inc., Israel) equipped with internal batteries and recharged by an external solar panel.

Data analyses were conducted using SPSS/PC version 11. Empirical relationships between P sorption parameters and soil properties were established with nontransformed data using Pearson correlation procedure and linear and multiple stepwise regression analysis. Differences between groups of soils according to location and geochemical conditions were examined using ANOVA testing for all level of confidence. The saturation indices of the supernatants of the batch-sorption experiments with respect to various solid phases were computed with the geochemical model PHREEQC. The thermodynamic data for hydroxyapatite, brushite, octacalcium phosphate (OCP) and ß-Ca3(PO4)2 were taken from Lindsay (1979), and the calcium–bicarbonate–phosphate solid phases, Ca2(HCO3)2HPO4 (CBP2) and Ca3(HCO3)3PO4 (CBP3), from Salingar et al. (1993).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
General Soil Characteristics
No statistical differences in OM or sesquioxide contents of the aerobic layers were found between the deep and shallow peat soils (Table 1). None of the aerobic peat layers showed the fibric or hemic decomposition stages reported in an earlier soil survey (Department of Agriculture, 1986), which we explained by the continuous decline in OM content due to prolonged oxidation and intensive cultivation. There was also no statistical difference in OM or sesquioxide contents between the aerobic (<45 cm, Eh {approx} 450 mV) and anaerobic layers (>45 cm, Eh = –220 mV) of the shallow peat soils because the entire shallow peat profile was fairly aerobic before the rewetting in 1994 and had been undergoing intensive oxidation since the late 1950s. On the other hand, the amount of OM content in the anaerobic peat layers of the deep peat soils was significantly higher (P < 0.001) than that in the corresponding aerobic peat layers. The high level of OM in the anaerobic peat layer resembles the elevated OM content reported for the peat material before the drainage (Ravikovitch, 1945).


View this table:
[in this window]
[in a new window]
  		Table 1. Means and standard error of the mean of selected chemical and physical properties of the peat soils grouped by location and redox potential (aerobic > 450 mV, and anaerobic < –220 mV).

 
The vertical pattern of the extractable FeCBD and total P in the anaerobic layers of the deep peat soils was conversely related to that of OM content (Table 1). The extractable FeCBD and total P concentrations in the anaerobic layers were significantly lower (P < 0.001) than their counterparts in the aerobic peat layers. We postulated that the observed vertical pattern is a result of the prolonged oxidation of OM in the top soil layers that consequently released Fe, P, and other occluded elements, followed by the rapid oxidation of Fe oxides and hydroxides with subsequent adsorption of P. No such vertical pattern was observed in the shallow peat soils because the entire peat profile was oxidized until the rewetting in 1994, while the deeper section of the profile (>1 m) is mainly organic-rich marl, which did not decompose as quickly as the peat layers.

The pH of the anaerobic layer in the deep peat soil was significantly lower (P < 0.05) than that of the more aerobic peat layer. There was a moderate negative correlation (r = –0.65, P < 0.001) between OM content and the acidity in the anaerobic peat layers. Some of the anaerobic peat layers that exhibited high OM content (>450 g kg–1) were quite acidic (4.5–3.2), reminiscent of the acidic conditions prevailing in the wetlands before the drainage. The acidic peat layers contained no CaCO3 and the source of the acidity is probably the high content of humic (>30% of the OM) and fulvic (>3% of the OM) acids (Ravikovitch, 1992). The pH of the anaerobic layer of the shallow peat soil was moderately alkaline because of the high CaCO3 content in these horizons (Table 1).

The distribution of the physicochemical properties summarized in Table 1 also suggested that the spatial heterogeneity within each group and depth of sampling is small enough to adequately represent the study area. This inference is in agreement with an earlier study of more than 70 peat samples with unique georeferencing that were analyzed for the spatial pattern of the degree of P saturation using stochastic simulation techniques (Litaor et al., 2003). This earlier study clearly demonstrated that all peat samples across the current study area exhibited DPS values well below 10% without outliers.

Iron Mineralogy
The concentrations of Fe oxides were significantly higher than those of Al and Mn oxides (P < 0.001). The continuous decline in OM content has increased the proportion of the mineral phases, especially the Fe oxides, in these peat soils. Previous attempts at studying the mineral composition of the clay fraction in these peat soils had been unsuccessful because of the large amount of OM (Yaari-Cohen et al., 1971). The main incentive for studying the nature of the Fe oxides in the present work stemmed from their importance in the chemisorption of P. For example, Torrent (1987) found that crystalline oxides contribute at least 40% of the sorbence to P, while Hamad et al. (1992) found that Fe oxides are responsible for at least 30 to 40% of P sorption in calcareous soils.

The Mössbauer spectra suggested that hematite, goethite, and magnetite are present mainly in the top layers of the aerobic section of the soils (Table 2). The ferrihydrite occurred mainly at greater depths and was probably formed by oxygen penetration through the numerous cracks and macropores that were formed because of the repeated drying and rewetting cycles. This observation is in good agreement with published reports that ferrihydrite is commonly found as ochreous precipitate from ferriferrous water in bog environments (Schwertmann and Taylor, 1989). The magnetic susceptibility data (Table 2) also suggested that ferrimagnetic phases occur in higher concentrations in the top oxidized layers compared with the more anaerobic deeper peat horizons. The common occurrence of magnetite in soils is usually explained by lithological detritus from mafic rocks such as basalt and pyroclastic material, which could have originated from the Golan Heights east of the Hula Valley. However, x-ray diffraction analyses of the clay fraction provided no evidence of other mafic minerals (data not shown), while the large Fe/Ti ratios (>50) found in the ferrimagnetic particles of these soils are far above the ratio of lithogenic magnetites, which often range between 3 and 10 (Deer et al., 1992). Hence, we speculated that the occurrence of magnetite in these soils resulted from the peat fires that were quite common in the area before its reflooding and the creation of Lake Agmon. The formation of magnetite by natural burning events has been postulated by Le Borgne (1960), but is not yet clear how this process actually occurs (Stanjek, 2000).


View this table:
[in this window]
[in a new window]
  		Table 2. The mineralogy of Fe oxides in two representative soils sampled at the shallow peat area (depth < 1 m) and deep peat area (depth > 4 m).

 
Sorption Characteristics
Sorption of P by the peat soils was adequately described by the Langmuir equation with coefficient of determination (r2) values of 0.97 or higher for all samples, under both aerobic and anaerobic conditions. The vertical distribution of the Smax clearly indicates significant differences in maximum sorption capacity (P < 0.001) between aerobic and anaerobic layers (Table 3). The Smax of the aerobic layers ranged from 670 to 1750 mg P kg–1, with a combined mean value for the deep and shallow Histosols of 1250 mg P kg–1, whereas the anaerobic layers ranged from 625 to 975 mg P kg–1 with a combined mean value of 775 mg P kg–1. A similar vertical pattern of Smax values has been reported by Reddy et al. (1998) for peat soils in Florida and by several researchers working in more temperate wetlands such as Somerset Levels and Moors in England, Biebrza Valley in Poland, and Trebel Valley in Germany (Meissner and Leinweber, 2004). The EPC0 at which neither adsorption nor desorption occurs was significantly lower in the aerobic peat layers (0.01–0.03 mg L–1) than in the anaerobic layers (0.11 mg L–1), particularly at depths > 120 cm (0.31 mg L–1). There were no statistical differences for the Smax or EPC0 values between the aerobic layers of the shallow and deep peat soils.


View this table:
[in this window]
[in a new window]
  		Table 3. Means and standard error of the mean of various sorption attributes of the peat soils grouped by location and redox potential (aerobic > 450 mV, and anaerobic < –220 mV).

 
Holford (1979) suggested the use of the MBC term to describe the resistance to changes in P concentration in a soil solution or the labile pool in the solid phase. Recently, Villapando and Graetz (2001) demonstrated its usefulness in studying the sorption properties of Bh horizons in selected Florida Spodosols. As expected, high MBC values were associated with high P sorbing layers, such as the aerobic layers of the shallow and deep peat soils (7300 and 2850 L kg–1, respectively, Table 3). On the other hand, the highly anaerobic layers, especially from depths greater than 120 cm, exhibited extremely low MBC values that ranged from 110 to 340 L kg–1 with a mean of 180 L kg–1. The low MBC values coupled with the high EPC0 values could result in higher P mobility through these layers. The anaerobic layers of the shallow peat soils did not exhibit low MBC values (Table 3), probably because of the intensive oxidation before rewetting. With the advent of the recent rewetting of this area, a slow transformation of the altered peat layers into a permanent anaerobic condition could eventually lead to a decline in their sorption capacity and MBC values, an increase in their EPC0, and enhanced P mobility to groundwater and nearby waterways.

The sorption parameters presented in Table 3 were calculated from the batch-sorption experiment, which was conducted under aerobic conditions. Hence, results may not be representative of soil layers under permanent anaerobic conditions. To examine the magnitude of this potential error, we conducted an adsorption experiment in two anaerobic soil layers using aerobic and anaerobic conditions. In general, the mid-section of the isotherms under anaerobic conditions did not vary significantly from those obtained under aerobic conditions (Table 4). Small differences were observed at the highest added concentrations (C0 of 70 mg L–1) where P sorption was higher under anaerobic conditions, while the reverse was observed at the low added concentrations (C0 < 8 mg L–1), where P sorption was lower under anaerobic conditions by 20 to 30% compared with aerobic conditions. Since the low P additions are within the common range of P concentrations found in fertilized fields, this experimental difference has important environmental implications because it suggests that more P will be transported to waterways through the anaerobic soil layers.


View this table:
[in this window]
[in a new window]
  		Table 4. Sorption characteristics of P under aerobic vs. anaerobic experimental setting, using two depths representing reducing conditions in the deep peat area.

 
These results are not in accordance with those of Reddy et al. (1998) and Pant and Reddy (2001), who studied peat soils and estuarine sediments in Florida and found a significant decrease in Smax and increase in EPC0 in experiments performed under anaerobic conditions compared with an aerobic setting. They attributed the differences to a reduction in amorphous and poorly crystalline Fe oxides. We attribute the lack of difference between the two experimental conditions in the Hula peat soils to potential readsorption of P on organometallic complexes, P sorption on Al oxides that are unaffected by changes in Eh but mainly to Ca-P precipitation during the sorption experiments (see below).

Correlation of Phosphorus Sorption Attributes with General Soil Properties
The P Smax obtained from the Langmuir equation did not correlate well with most of the general soil parameters (Table 5). There were only weak to moderate correlations (P < 0.05) between Smax and EPC0, respectively, and the extractable FeCBD, FeOXAL, and AlOXAL. There was no correlation at all between the sorption attributes and the soil pH, or the OM, MnCBD, and CaCO3 contents. The moderate correlation between AlOXAL and P sorption is probably one of the reasons for the lack of significant differences in P sorption characteristics observed in aerobic and anaerobic experimental setting (Fig. 2) . Aluminum oxides are important P sorbents in many soil environments and are unaffected by changes in Eh. However, since the content of Al oxides in these soils is quite small relative to Fe (Table 1), their relative contribution to the overall P sorption capacity is limited. The lack of high correlation between sesquioxides and P sorption attributes was unexpected since several lines of evidence have suggested that these solid phases, especially the Fe oxides, should be the major sorbent of P in peat soils. First, the relatively large amount of extractable Fe (Table 1), its mineralogy (Table 2), and its ubiquitous presence in the field in the form of bright red coatings on soil aggregates and sediments deposited along the drainage cannels imply that Fe should be a strong participant in chemisorption processes. Second, during a prolonged microcosm experiment that simulated the redox change from oxidized peat soil to highly reduced conditions, ferrous Fe and P were released concurrently (Shenker et al., 2005). Third, there was good agreement between the ratio of oxalate-extractable P to Smax and the ratio of oxalate-extractable P to oxalate-extractable Fe+Al in 12 sampling locations at a depth of 50 cm across the Hula Valley (Litaor et al., 2003). Fourth, a high coefficient of determination exists between the P extracted by 0.5 M NaOH and extractable FeCBD, MnCBD, and AlCBD concentrations (r2 = 0.88, P < 0.001) (Litaor et al., 2004). The 0.5 M NaOH extract was designed to release P associated with sesquioxide solid phases (Tiessen and Moir, 1993). We further examined the latter relationships by running a correlation analysis between the sorption attributes and the various P concentrations extracted by Hedley's sequential extraction procedure (data tabulated in Litaor et al., 2004). The correlation analysis showed that Smax was moderately correlated with P concentration extracted by 0.5 M NaOH, 1 M HCl, and the sum of all P fractions (r = 0.5, P < 0.05; r = 0.42, P < 0.05; r = 0.51, P < 0.01, respectively). The EPC0 was highly correlated with the most labile P extracted by anion resin (r = 0.83, P < 0.001) and moderately correlated with the labile P extracted by NaHCO3 (r = 0.5, P < 0.03). These correlation results demonstrated the strong linkage between the labile P fraction and the point at which sorption equals desorption, which directly affects the P intensity factor. The moderate correlation between the 1 M HCl-extracted P and Smax implied that other soil processes, such as rapid Ca-P precipitation, might interfere with the batch-sorption experiment, obscuring the calculated P sorption capacity and resulting in a lower-than-expected correlation coefficient between Fe oxides and Smax.


View this table:
[in this window]
[in a new window]
  		Table 5. Correlation of P sorption attributes with selected soil parameters.

 


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. Sorption characteristics of peat soils under aerobic vs. anaerobic experimental conditions.

 
Pattern of Phosphorus Fractionation Following the Sorption Experiment
To test the importance of Ca-P precipitation during the sorption experiment, we sequentially extracted various P fractions following the addition of P at concentrations of 0, 1, 2, 8, 10, 30, 40, 50, 60, or 100 mg kg–1 using the common batch-sorption experiment (Table 6). The main finding of this experiment was the consistent increase in the P fraction extracted by 1 M HCl relative to the increase of other P fractions when C0 was 8 mg L–1 or higher. The level of equilibrium P (Ct) remained quite low, even with higher P additions, indicating the high capacity of the peat soil to hold the added P. Since the 1 M HCl represents mainly the Ca-P solid phases (Tiessen and Moir, 1993), the substantial increase in this fraction suggests that Ca-P precipitation occurred simultaneously with P sorption, even when C0 was as low as 8 mg L–1. This indicates that, under the experimental conditions, the process of adsorption cannot be distinguished from the process of precipitation. A similar experiment conducted in our laboratory using temperate peat soil collected from an altered wetland in the Trebel Valley, Germany, showed no change in the Ca-P fraction because the initial level of Ca in those peat soils was insignificant and all the added P was adsorbed by sesquioxides (Brielmann, 2004, unpublished data).


View this table:
[in this window]
[in a new window]
  		Table 6. Results of the P fractionation experiment using a modified Hedley's sequential extraction procedure after P addition at 0 to 100 mg L–1 in the batch-sorption experiment. We conducted four experiments using samples from the rewetted shallow peat, but for brevity, only one experiment is shown. The d(fraction) represents the increase in each fraction concentration compared with its concentration before the sorption experiment where the added P ranged between 0 to 100 mg L–1.{dagger}

 
The high content of OM in the soils studied here makes a detailed Ca-P mineralogical analysis difficult. To elucidate which Ca-P mineral might be formed during the batch experiment, we modeled the supernatants of successive P additions using elemental results of the pertinent geochemical variables and calculated the saturation index of selected Ca-P minerals known to affect P concentrations in soils (Table 7). The results of the geochemical modeling indicated that even under low P addition, the supernatants are supersaturated with respect to various Ca-P minerals, whereas under higher P addition (30 mg L–1), most Ca-P solid phases, except the most soluble mineral (brushite), will likely precipitate. The source of the Ca in these peat soils differs from other studies that attribute the relatively large Ca-P fraction to pedogenic CaCO3 as the primary geochemical agent in P sorption in selected Mediterranean and arid soils (Carreira et al., 1997; Lajtha and Bloomer, 1988). The main source of the Ca in our soils is gypsum, which precipitated in the peat layers following the drainage (Bein and Nielsen, 1988; Markel et al., 1998). The rewetting of the shallow peat soils and the seasonal groundwater fluctuations in the deep peat soils initiated the dissolution of the gypsum, resulting in mean Ca concentrations in the soil interstitial waters of 500 ± 300 mg L–1, and in the shallow groundwater of 850 ± 450 mg L–1. These results are quite similar to the concentrations found in the supernatants (Table 7). The fractionation analysis of the sorption experiment and the geochemical modeling results strongly suggest that Ca-P precipitation affected the Smax, thus obscuring the expected high correlation between the Smax and sesquioxides. The governing solid phase in the supernatant solutions was in the form of ß-Ca3(PO4)2 or other Ca-P phases with similar or higher stability. Since this precipitation was fast enough to occur during the experiment, and the residence time of percolating waters in this soil environment is even longer, we suggest that the Ca-P precipitation is an important P retention mechanism.


View this table:
[in this window]
[in a new window]
  		Table 7. Calculated saturation index (SI) of selected Ca-P solid phases using the supernatants from the sorption analysis (C0 = 0.2, 4, 10, 15, and 30 mg L–1). The input parameters were pH 6.2, a measured alkalinity of 250 mg L–1, and Ca concentrations that varied from 640 to 1350 mg L–1.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The results of the present study clearly confirmed the premises and sequential processes stated in our working hypothesis regarding the potential chemical behavior of P in peat soils of this altered wetlands. We hypothesized that the drainage of the wetlands initiated strong OM oxidation, followed by Fe transformation, which in turn resulted in P sorption on the Fe solid phases. Recent rewetting of these soils may desorb P from the dissolving Fe oxides, a process that may be partially mitigated by concurrent Ca-P precipitation due to gypsum dissolution. The vertical distribution of Fe oxides, Fe minerals, total P concentration, sequential extraction experiments of sorbed P, the high P sorption capacity in aerobic peat layers, and relatively lower sorption capacity in anaerobic peat layers all supported these premises. Moreover, only weak to moderate correlations exist between the sorption attributes and the sesquioxide contents, most likely due to the precipitation of Ca-P phases. The main source of Ca in this soil system is gypsum; hence, the solubility of P is greatly influenced by the dissolution and precipitation of gypsum which, in turn, is affected by the drying and rewetting cycles. The rewetting of the peat soils leads to a decrease in Smax and MBC, and an increase in EPC0, which could lead to higher P mobility. The potential mobility of P then diminishes via an increase in Ca-P precipitation due to increased gypsum dissolution. The current eco- and farming management practices that require the maintenance of high groundwater level in the center of the Hula Valley may result in higher P transport from the peat soils to waterways.


   ACKNOWLEDGMENTS
 
This research was supported in part by the EU project PROWATER, EVK1-CT1999-00036 and in part by a grant (GLOWA-Jordan River) from the Israeli Ministry of Science and Technology; and the German Bundesministerium fuer Bildung und Forschung (BMBF).

Received for publication March 2, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
S. Brand-Klibanski, M. I. Litaor, and M. Shenker
Overestimation of Phosphorus Adsorption Capacity in Reduced Soils: An Artifact of Typical Batch Adsorption Experiments
Soil Sci. Soc. Am. J., June 8, 2007; 71(4): 1128 - 1136.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
M. I. Litaor, G. Eshel, O. Reichmann, and M. Shenker
Hydrological Control of Phosphorus Mobility in Altered Wetland Soils
Soil Sci. Soc. Am. J., September 20, 2006; 70(6): 1975 - 1982.
[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 Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Litaor, M. I.
Right arrow Articles by Shenker, M.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Litaor, M. I.
Right arrow Articles by Shenker, M.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Litaor, M. I.
Right arrow Articles by Shenker, M.
Related Collections
Right arrow Phosphorus
Right arrow Soil Chemistry
Right arrow Wetland Soils

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
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome