| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Limnology and Environmental Protection, P.O. Box 27, FIN-00014, University of Helsinki, Finland
b Department of Applied Chemistry and Microbiology, P.O. Box 27, FIN-00014, University of Helsinki, Finland
Corresponding author (jukka.koski-vahala{at}helsinki.fi)
Received for publication March 31, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
Abbreviations: EPC, equilibrium phosphorus concentration in bath solution
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
Sediment resuspension has been found to increase the total amount of P in the water column (Peters and Cattaneo, 1984; Hamilton and Mitchell, 1988; Kristensen et al., 1992), whereas the results regarding changes in soluble reactive phosphorus (SRP) have varied, even to the point of being contradictory: sediment resuspension has either decreased (Fitzgerald, 1970; Gunatilaka, 1978; de Groot, 1981), increased (Rippey, 1977; Søndergaard et al., 1992; Reddy et al., 1996), or had no effect on the soluble P in the water column (Peters and Cattaneo, 1984). Furthermore, the mechanisms behind various results have not been fully understood. The lack of useful methods and the heterogeneity of the settling matter (i.e., seston) complicate the phenomenon (e.g., Gächter and Mares, 1985). Thus, to assess the role of resuspended sediment in controlling P fluxes in lakes, more knowledge of the factors contributing to and useful approaches to simulate the resuspension phenomenon are needed.
At present, P exchange between the solid and solution phases is understood to be dependent on the P equilibrium between the phases. Even if a true equilibrium cannot be reached, the desorption–sorption isotherm method (White and Beckett, 1964), based on equilibrium theory, has been used to determine P release from soil or sediment to ambient water (e.g., Taylor and Kunishi, 1971; Yli-Halla et al., 1995). Although this method simplifies the whole solid–solution system, desorption–sorption has been found to control P exchange in mineral soils. In comparison with microbial processes, chemical mechanisms are found to be of major importance in P retention in sediment rich in inorganic material as well (Kairesalo et al., 1995), and the contribution of the bacterial production to nutrient fluxes has been found to be low in a lake with intensive resuspension of inorganic sediment (Koski-Vähälä et al., 2000). In fact, sorption capacity of suspended sediment has been observed to be higher than that of surface soil particles because intensive erosion selectively transports finer particles with higher P sorption capacity (Sharpley, 1980; Sharpley et al., 1981). Furthermore, eroded fine particles are also higher in P than surface soil.
The dynamic nature of resuspension has often been ignored in resuspension research. Phosphorus exchange has been investigated at a constant solid to solution ratio, and the desorption from and sorption to the solid phase has been expressed as a function of the P concentration in the equilibrium solution (after the event). However, in resuspension studies, this approach does not yield any direct information about the quantity of P to be released or sorbed when the sediment material is mixed with water layers of varying P concentration (before the event). Furthermore, the solid to solution ratio continuously decreases with intensifying mixing, and the resuspended material will encounter water layers of dissimilar chemical characteristics (e.g., pH, ionic strength, and oxygen concentration).
In the present study, the factors contributing to the P loading risk were investigated by a desorption–sorption isotherm technique and an extraction test. The progressive mixing was simulated by using decreasing solid concentrations, and the changes of environmental conditions in different water layers were simulated by varying solution pH, redox potential, and ionic strength. To estimate the extent of the P exchange when resuspended solids are mixed with the different water layers, the isotherms were depicted as a function of P in the bath solution.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
The water content of the sediment was determined by drying a subsample in an oven at 105°C and loss on ignition at 550°C. The P reserves were characterized using sequential fractionation analysis as developed by Chang and Jackson and modified by Hartikainen (1979). Total carbon and nitrogen were determined with a Leco (St. Joseph, MI) CHN-900 analyzer. Exchangeable cations (Ca, Mg, Na, and K) were extracted with four portions of 1 M NH4Cl at a sediment to solution ratio of 1:15 (w/v) and analyzed with an atomic absorption spectrophotometer (AAS). The effective cation exchange capacity (ECEC) was calculated as the sum of exchangeable cations. Amorphous hydrated oxides of iron, aluminum, and manganese were extracted for 2 h with 0.05 M ammonium oxalate (pH 3.3) at a sediment to solution ratio of 1:20 (Schwertmann, 1964) and measured by AAS. To measure the P amount in the interstitial water the sediment was centrifuged (2350 x g), the supernatant was filtered through a Nuclepore 0.2-µm membrane filter (Whatman, Maidstone, UK), and the dissolved P (PO4–P) concentration of three replicates was measured by a molybdenum blue–ascorbic acid method (Murphy and Riley, 1962) with an autoanalyzer (Lachat [Milwaukee, WI] QuickChem 8000).
Desorption–Sorption Isotherms and Their Interpretation
For the study of P exchange between resuspended sediment and the surrounding solution phase, a set of triplicate sediment samples was allowed to react with solutions containing various concentrations of a standard KH2PO4 (0 to 500 µg P L-1). Deionized water was used as a background solution. The suspensions were stirred on an orbital shaker (150 rpm) in the dark at 20 ± 1°C for 24 h, which has been shown to be long enough for the main part of exchange reactions to take place (Barrow, 1983; Froelich, 1988; House et al., 1995). After shaking, subsamples of 30 to 40 mL were filtered through a 0.2-µm membrane filter (Nuclepore) and analyzed for dissolved P by a molybdenum blue–ascorbic acid method with an autoanalyzer (Lachat Quick 8000). To simulate the effect of gradual mixing on the P fluxes, solid concentrations of 1700, 170, and 17 mg L-1 were used.
When studying the effect of pH on P exchange, isotherms were prepared by using Tris buffer (Trizma base) (Sigma–Aldrich, St. Louis, MO) as the background solution in order to maintain a constant pH in the suspension during shaking. The ionic strength of the Tris solution, I = 0.001, was equal to that in Lake Tuusulanjärvi (
1 mmol L-1). The pH of the solutions was adjusted to 7 or 9 by adding H2SO4, and in order to make them equal in ionic strength, K2SO4 was added to the solution of pH 9. The highest solid concentration of 1700 mg L-1 was not used in the isotherms with the buffer solutions because Tris buffer was not able to maintain a constant pH at this solid concentration.
To monitor the P exchange in anoxic conditions, sediment and the Tris solution of pH 7 were bubbled with N2 gas for 24 h to remove any oxygen from these media. All additions were completed under an N2 atmosphere, and the O2 concentration was checked using Winkler's method (SFS 3040, 1990). To prevent reoxidation, the sample bottles were shaken in Minicrip plastic bags (Amerplast, Tampere, Finland) filled with N2 gas. After shaking, subsamples of 20 mL were filtered through 0.2-µm membrane filters (Nuclepore) by pressfiltration (Kemira Systems, Helsinki, Finland) in an N2 atmosphere. The filtrates were collected in plastic bottles with 0.5 mL ascorbic acid (60 g L-1) to avoid the reoxidation of soluble Fe(II) and to stabilize the dissolved P concentration.
Desorption from or sorption to the suspended material was measured from the changes in phosphate concentration in the solution and calculated in mg P kg-1 of dry sediment (Q). The negative values of Q represent desorption and the positive values sorption. When assessing the role of resuspension in the internal P loading, it is more informative to know to what extent suspended solids are able to release or retain P in the water column they encounter than to know the P exchange as a function of the final P concentration to be attained. Therefore, the results were depicted as a function of dissolved P concentration in the bath solution to which the sediment was added (C) and the data were fitted to a nonlinear model:
 |
The parameters of this isotherm model were determined using iteration by the NLIN procedure of SAS (SAS Institute, 1989). At the solid concentration of 1700 mg L-1 a linear model Q = bC - Q0 (b = constant) was used as all observations were in the linear range (see Hartikainen, 1991).
The paramater Q0 stands for the potential labile P in the resuspended material to be desorbed under the experimental conditions used. The x intercept gives the P concentration in the solution where no net desorption or sorption can be expected to take place. This equilibrium P concentration (EPC) is the same as that used in the conventional Q/I plots to express the P concentration that soil or sediment can maintain without an additional P supply under the prevailing conditions (White and Beckett, 1964; Taylor and Kunishi, 1971; Hartikainen, 1991). Thus, in a solution of this P concentration suspended solids do not cause any P release or act as a sink for P. Qmax, the maximum amount of P that the sediment material can retain, is only of theoretical significance because normally an extremely high P concentration is needed to obtain the maximum sorption capacity. Nevertheless, this parameter is useful for assessing the total number of sorption sites on the solid material. The parameter K determines the shape of the isotherm graph and depicts the sorption power of the material.
Extraction at Different pH
The effect of pH on the P release was investigated in more detail by an extraction experiment where sediment was added to H2SO4 or NaOH solutions of different concentrations. The ionic strength of the solutions was adjusted to 1 mmol L-1 (I = 0.001) by adding CaCl2. The suspensions of different solid concentrations (1700, 170, and 17 mg L-1) were stirred on an orbital shaker (150 rpm) in the dark at 19 ± 1°C for 4 h. Thereafter, a subsample of each suspension was filtered, and dissolved P concentrations in the filtrates (0.2-µm Nuclepore) were determined as described above. The P desorption was calculated from the increase in P concentration in the solution (five replicates) and expressed in mg P kg-1 of dry sediment.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
|
|
The maximum P sorption capacity (Qmax) decreased considerably at pH 9 concomitantly with an increase in the Q0 and EPC parameters (Table 2). As compared with the Tris solution of pH 7, water of low ionic strength had a similar effect. Under anoxic conditions, on the other hand, the Qmax values were highest. The shape of the desorption–sorption graphs at elevated pH and under anoxic conditions were dissimilar (Fig. 1). This denotes that the P mobilization mechanisms were possibly dissimilar. The parameter K seemed to vary irregularly between the treatments (Table 2).
|
Effect of pH on Phosphorus Mobilization
The separate extraction experiment revealed that the effect of pH on the P mobilization was markedly dependent on the concentration of suspended solids in the system (Fig. 2). At the solid concentrations of 17 and 170 mg L-1, the P release increased sharply below pH 5 and above pH 7. At the highest concentration of 1700 mg L-1, on the contrary, no effect was observed at pH ranging from 3 to 8.5, and no significantly enhanced P mobilization was found until above pH 9. When the concentration of suspended solids diminished, the acid-evoked increase in the P release (pH < 5) was more distinct than the base-evoked release (pH > 7).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
The EPC value (the intercept on the x axis), on the other hand, dictates the critical P concentration below or above which the suspended solids will release or retain P, respectively. Thus, EPC describes the actual P loading risk. The isotherms prepared at various solid concentrations provided evidence of the dynamic nature of the resuspension process: the larger the volume of solution the sediment material will be mixed with, the higher the P concentration in the surrounding solution that allows desorption to occur. This essential property of resuspended sediment is due to the P buffering capacity of solid particles. In the bottom deposits, a high concentration of solids enhances the sorption reactions and decreases EPC. Furthermore, the ionic strength (i.e., salt concentration) is higher in the interstitial water than in the upper water layers. This promotes P sorption to and decreases P desorption from solid surfaces (Ryden et al., 1977; Barrow et al., 1980; Pardo et al., 1992; Yli-Halla and Hartikainen, 1996). The effect of ionic strength is based on changes in the net surface charge of variably charged particles and the electrical potential of oxide surfaces (Parker et al., 1979; Barrow et al., 1980; Barrow and Ellis, 1986; Bolan et al., 1986). A low ionic strength, in turn, promotes the dispersion of aggregated mineral particles and, consequently, enhances the release of colloidal P.
The effect of ionic strength in the solution phase on P mobilization has received only sporadic attention in sediment studies. However, the comparison of the results obtained in the water and Tris treatments show that a rather slight increase in the ionic strength in the background solution can markedly reduce EPC in the system. This response is in accordance with previous observations of Yli-Halla and Hartikainen (1996) showing that at low ionic strengths the salt effect on the P release is more pronounced than at higher ionic strengths. Ionic strength affected EPC most markedly at the low solid concentration, which emphasizes the loading risk of resuspended material as compared with the calm sediment. In general, there is only little variation in the ionic strength between and within lakes, whereas in estuaries the ionic strength varies markedly and, hence, can affect the P mobilization. However, in order to apply the results of laboratory studies to field conditions the sediment studies should be carried out at ionic strengths relevant to natural water systems.
As for the other chemical factors studied, EPC increased in the order: pH 7 < pH 7 anoxic conditions < pH 9, which is in accordance with earlier observations in sediment (Istvánovics, 1994). This sequence indicates that elevated pH, which is common in eutrophic lakes during the summer, favors the role of resuspended solids as a P source more than a low redox potential does. Concomitantly with rising pH, intensive primary production decreases the dissolved P concentration in the water column, which creates conditions favorable for internal P loading.
An increase in EPC due to a low redox potential can be attributed to changes in sorption components. According to classical theory, iron(III) is reduced under anaerobic conditions to iron(II), leading to dissolution of Fe-bound P (Einsele, 1938; Mortimer, 1941, 1942). On the basis of the ligand exchange theory introduced by Hingston et al. (1967), the intensified P release under anaerobic conditions is due to the reduction and/or dissolution of the hydrated iron oxides acting as a sorption component. However, some of the P released under anoxic conditions will be resorbed by aluminum oxides, which diminishes the net mobilization of P (Hartikainen, 1979). At high pH, on the other hand, the intensified P release has been interpreted as an increase in negative charges on hydrous oxide surfaces and as competition between OH- and H2PO-4 for sorption sites (Andersen, 1975; Lijklema, 1977, 1980; Boström et al., 1982). The more marked increase in EPC at high pH can be explained by the fact that at high pH both Fe- and Al-bound P will be released. It is also possible that high pH increased the release of organic matter and thus, some organic P contributed to P concentrations measured. Thus, the dissimilar shapes of the isotherms at high pH and anoxic conditions are obviously due to these differences in reaction mechanisms.
The effect of pH on EPC in the isotherm study and on P release in the extraction test became more pronounced when the amount of solids in the suspension decreased. This reaction pattern emphasizes the role of pH as a factor controlling the fate of P during resuspension. However, no attention was paid in previous studies to the fact that the effect of pH on the P release depends on the solid concentration. Because the studies have been carried out at higher solid concentrations, enhanced P desorption from sediment has been observed only at much higher pHs than in the present study (e.g., Rippey, 1977). This retarded mobilization can be seen also in the base extraction curve in Fig. 2, showing that because of the high pH buffering at the highest solid concentration, the P mobilization did not become apparent until at very high pH levels. The more pronounced pH effect at low sediment to solution ratios can be explained by the higher H+ and OH- load per dry solid mass and sorbed phosphate ions in the system, rendering the chemical reactions more effective. In other words, when the amount of suspended solids decreases, relatively more OH- or H+ is available for ligand exchange and dissolution reactions, respectively.
The results of the extraction test agree with those obtained earlier in closed systems with increasing acid or base dosages (Murrmann and Peech, 1969; Rippey, 1977); the release of P was enhanced at both high and low pH with a minimum release at pH of 5 to 6. However, Hartikainen and Simojoki (1997) showed that in an open elution system where the soil is repeatedly treated with a solution of a constant acid concentration, the release of P from the solid material regularly decreases. In batch titrations, the acid concentration in the closed system becomes abnormally high and starts to dissolve P, especially from Ca-bound reserves (Hartikainen and Simojoki, 1997). In lakes, pH values below 5 are uncommon and, thus, dissolution-induced P release is obviously of minor importance in internal P loading.
In theory, the redox parameter (the sum of pH and electron activity pe) tends to be constant (Lindsay, 1979). Thus, lowered redox potential often brings about increased pH, which intensifies the effect of anoxic conditions that release both Fe- and Al-bound P. This is supported by sediment investigations that show higher P release under anaerobic conditions at all pH levels studied (Boström et al., 1982; Istvánovics, 1994). These separate desorption mechanisms could enhance the internal P loading in waters where anoxic sediment is mixed with the water column of elevated pH. However, the phenomenon is far from simple, because the P release is affected by the P to Fe ratio of the sediment (Lijklema, 1977), by resorption of released P from the Al hydroxides (Hartikainen, 1979), and by increased reprecipitation of P with Ca carbonate at high pH in Ca-rich waters (Gunatilaka, 1982).
The maximum P sorption capacity (Qmax) decreased considerably at pH 9 due to the increased competition of OH- ions for sorption sites. Its increase under anoxic conditions is, on the contrary, difficult to explain, because the anoxic environment decreases the number of sorption sites through reduction and dissolution of Fe oxides. The result might be attributed to the uncertainty in the regression procedure, when the parameter Qmax is calculated outside the range for which experimental data are available.
|
PRACTICAL IMPLICATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONPRACTICAL IMPLICATIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. L. Bjorneberg, D. T. Westermann, J. K. Aase, A. J. Clemmens, and T. S. Strelkoff Sediment and Phosphorus Transport in Irrigation Furrows J. Environ. Qual., April 3, 2006; 35(3): 786 - 794. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Liikanen, M. Puustinen, J. Koskiaho, T. Vaisanen, P. Martikainen, and H. Hartikainen Phosphorus Removal in a Wetland Constructed on Former Arable Land J. Environ. Qual., May 1, 2004; 33(3): 1124 - 1132. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
= pH 9, x = pH 7 N2, 
= deionized water.