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


     

Published online 31 May 2006
Published in J Environ Qual 35:1181-1192 (2006)
DOI: 10.2134/jeq2005.0213
© 2006 American Society of Agronomy, Crop Science Society of America, and 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 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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Christophoridis, C.
Right arrow Articles by Fytianos, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Christophoridis, C.
Right arrow Articles by Fytianos, K.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Christophoridis, C.
Right arrow Articles by Fytianos, K.
Related Collections
Right arrow Phosphorus
Right arrow Laboratory Column Studies
Right arrow Nutrients
Right arrow Sorption/Exchange
Right arrow Other Environmental Contamination

TECHNICAL REPORTS

Wetlands and Aquatic Processes

Conditions Affecting the Release of Phosphorus from Surface Lake Sediments

Christophoros Christophoridis and Konstantinos Fytianos*

Environmental Pollution Control Laboratory, Department of Chemistry, Aristotle University of Thessaloniki, GR-541 24 Thessaloniki, Greece

* Corresponding author (fyti{at}chem.auth.gr)

Received for publication May 25, 2005.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Laboratory studies were conducted to determine the effect of pH and redox conditions, as well as the effect of Fe, Mn, Ca, Al, and organic matter, on the release of ortho-phosphates in lake sediments taken from Lakes Koronia and Volvi (Northern Greece). Results were evaluated in combination with experiments to determine P fractionation in the sediment. The study revealed the major effect of redox potential and pH on the release of P from lake sediments. Both lakes showed increased release rates under reductive conditions and high pH values. The fractionation experiments revealed increased mobility of the reductive P fraction as well as of the NaOH–P fraction, indicating participation of both fractions in the overall release of sediment-bound P, depending on the prevailing environmental conditions. The results were assessed in combination with the release patterns of Fe, Mn, Ca, Al, and organic matter, enabling the identification of more specific processes of P release for each lake. The basic release patterns included the redox induced reductive dissolution of P-bearing metal oxides and the competitive exchange of phosphate anions with OH– at high pH values. The formation of an oxidized surface microlayer under oxic conditions acted as a protective film, preventing further P release from the sediments of Lake Volvi, while sediments from Lake Koronia exhibited a continuous and increased tendency to release P under various physicochemical conditions, acting as a constant source of internal P loading.

Abbreviations: DOC, dissolved organic carbon • LOI, loss of ignition • TOC, total organic content • TP, total phosphorus


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE concentration and forms of nutrients in lake systems vary widely, depending on nutrient sources (internal and external), as well as environmental, climatic, morphological, and hydrological factors. The main external sources responsible for P loading include drainage from intensively cultivated areas, sewage–municipal waste, and runoff from livestock farming (Reynolds and Davies, 2001; Kaiserli et al., 2002). The increased and continuous addition of P to lake ecosystems has led to rapid eutrophication in numerous lakes (Boström et al., 1982).

Lake sediments are composed of complex aggregations of minerals, organic and inorganic species, as well as water. Depending on various physicochemical and biological factors, sediments act as either sinks or sources of a large number of organic and inorganic compounds, especially P in the form of readily diluted phosphates (Boström et al., 1988; Eckert et al., 1997; Kim et al., 2003; Søndergaard et al., 2003). When acting as a source of P, lake sediments increase the P concentration in the water column, a process called internal loading. This can lead to increased eutrophication. In some cases internal loading can contribute up to 80% of the total P (TP) input of a lake (Penn et al., 2000). Although improved catchment management has reduced the extent of external P loading, in many cases lake P concentrations have not responded as expected (Koussouris et al., 1992). Instead, they decreased only slightly or did not decrease at all due to the internal loading of the lakes (Boström et al., 1988).

Phosphorus can be found in lake sediments in various forms. Some fractions are virtually permanently bound in the sediments, while others are potentially mobile, leading to release of P under appropriate conditions (Boström, 1984; Søndergaard et al., 2003). Inorganic P, which occurs in its readily soluble form as the orthophosphate anion, can be found in sediments either as part of a mineral or as precipitated phosphate salt, such as hydroxyapatite Ca5(PO4)3(OH), FePO4, AlPO4, and Ca3(PO4)2 (Reynolds and Davies, 2001). More often phosphates are fixed as adsorbed anions on the surface or interior of various metal oxides and hydroxides, which constitute the sediment particles (especially those of Fe, Mn, and Al). Phosphorus can also be sorbed by sedimentary organic matter or exist in the form of organic phosphate esters, but generally at low concentrations.

The release of P from lake sediments is a highly complex phenomenon which involves a number of physical, chemical, and biological processes such as: desorption, ligand exchange mechanisms, dissolution of precipitates, mineralization processes, release from living cells, and autolysis of cells. The first two seem to be the prevalent release mechanisms in lake sediments, with diffusion as the main transport process (Boström et al., 1988). Related environmental factors that have been identified as controlling sediment P release are: temperature, pH, redox potential (Eh), nitrate and sulfate concentration, bioturbation and mixing, as well as biological activity (Kleeberg and Kozerski, 1997). Concentrations of Ca, Mg, Mn, and Fe both in the sediment and water also influence the P retention capacity. The influence of the above mentioned factors is analyzed below.

An increase in temperature results in a decrease in P adsorption and indirectly to an enhancement of biological activity (Kleeberg and Kozerski, 1997; Perkins and Underwood, 2001).

Under increased pH values (e.g., high primary production) the P-binding capacity of Fe and Al compounds is decreased, due to ligand exchange reactions in which hydroxide ions replace orthophosphate ions (Andersen, 1975; Boström et al., 1982; Boers, 1991), while the extent of dissolved P reprecipitation to iron(III) compounds will be limited (Boström et al., 1988; Koski-Vahala et al., 2001; Olila and Reddy, 1995).

At low redox potentials (below +200 mV), part of the insoluble oxidized Fe(III) oxides and hydroxides are reduced to their soluble Fe(II) forms, thus releasing P sorbed on the surface and interiors of these compounds. The extent of this release is controlled by the redox value, the Fe/TP ratio (availability of P binding sites), and the presence of other redox active compounds such as nitrates and sulfates (Boström et al., 1982; Perkins and Underwood, 2001; Søndergaard et al., 2003). Manganese oxides and hydroxides act in a similar way although their role is not absolutely understood. In general, anaerobic conditions favor the release of P to the overlying water. Stratification of deep lakes as well as microstratifications occurring in shallow lakes, can lead to formation of anaerobic zones that favor P release mechanisms.

Increased nitrate concentration acts as a buffer to the redox potential of the surface sediments, keeping it at an increased level, thus preventing the anaerobic dissolution of Fe-bound P. On the other hand, it can increase the transfer of P from the sediments to the overlying water, acting as an alternative electron acceptor in biological processes (Boström et al., 1988; Gächter et al., 1988). Sulfates also influence P release indirectly, under low redox conditions, forming sulfides and enhancing Fe(III) dissolution by forming FeS (especially under prolonged anaerobic conditions) (Caraco et al., 1993).

Microbial activity can cause mobilization of sediment-bound P through mineralization processes (Marsden, 1989), depending on the P content of the organic substrate and growth yield of mineralizing bacteria. Bacterial activity can mobilize the P pool indirectly, by nitrate consumption and sulfide and methane formation, thus affecting the redox potential, pH, and other factors that are significant to P equilibrium (Boström et al., 1988).

Phosphorus can also be found sorbed in Ca and Mg compounds, mainly carbonates. The main factor controlling their retention capacity of P is pH. Humic material acts as a sink of P, due to its tendency to chelate Fe and therefore P. Moreover, it can adsorb organic P forms (Boström et al., 1988).

Determination of the total P content of the sediment does not provide information related to the conditions under which P can be fixed or released by the sediment. For this reason, a number of sediment P fractionation schemes have been developed, based on the needs of individual studies. Various methods separate P fractions according to the physicochemical properties characterizing its binding form (Psenner et al., 1984; Psenner et al., 1988; de Groot and Golterman, 1990; Hupfer et al., 1995; Pardo et al., 1998; Rydin, 2000), their bioavailability (Zhou et al., 2001), and their ecological importance.

Sediment P retention is highly variable between lakes, not only due to different morphometry and water renewal rates, but also due to differences concerning edaphic characteristics and the history and the magnitude of the external P load (Boström et al., 1988). In this study, experiments were conducted to obtain a better understanding of the physicochemical conditions affecting sediment P release and the chemical changes taking place on the sediment surface of two lakes located in Northern Greece, thus determining the extent of internal P loading and long-term eutrophication potential.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Study Area
Two lakes were studied, Volvi and Koronia, both located in Northern Greece, northeast of Thessaloniki. Both lakes are of great ecological and financial importance to the surrounding area. They host numerous flora and fauna species, and they are protected by the Ramsar Convention as well as the MedWet and Natura 2000 environmental networks, as sites of international importance for the ecological value of the wetland habitat. Both lakes are water-supplied by rainfall, ground water, and stream flow (Kaiserli et al., 2002). The main morphometric characteristics of the studied lakes are given in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Main morphometric characteristics of the studied lakes.

 
Volvi is a meso-to-eutrophic lake. It is the largest and deepest of the two lakes. Water stratification takes place during the summer months, when DO levels are decreased in the hypolimnion and anoxic conditions can be observed in the deep part of the lake (Koussouris et al., 1992). The major environmental impacts on Lake Volvi are due to agricultural runoff containing nutrients, intensive water abstraction, and the transport of significant quantities of particulate matter by contributing streams, which tend to decrease the depth of the lake.

On the other hand, Lake Koronia is a hypereutrophic shallow lake, which suffers from long periods of drought. Its surface area and its active depth have been diminished over the last 5 yr due to the exhaustive pumping of lake water for agricultural and industrial usage. Additionally, agricultural runoff from the cultivated areas in the catchment (causing fertilizer and pesticide pollution), untreated domestic effluents, and industrial wastewater have significantly degraded the water quality. Increased nutrient input, in combination with the constantly decreasing available water volume, has converted the lake to a hypereutrophic aquatic system, suffering periodic algal blooms, depletion of DO, and toxic effects on the fauna. Various contaminants have accumulated over the years in the remaining water and sediments of this shallow lake, which are easily disturbed and mixed by water movement and wind (Koussouris et al., 1992).

Sediment Sampling and Analysis
Intact sediment cores were collected from one sampling site in each lake at the end of the winter period (January 2003), at an approximate depth of 12 m for Lake Volvi and 3 m for Lake Koronia (three replicates for each sampling site). Water samples were obtained from the water column 50 cm above the sediment sampling site. The sites were selected based on past measurements of nutrient content of sediment. They were not located near effluent streams, small rivers, or known point-pollution sources. A map of the studied area with the sampling sites and the major streams is provided in Fig. 1.


Figure 1
View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. Map of the studied area with sampling sites.

 
All samples were kept cool in the field (4°C). They were transferred to the laboratory, where they were immediately used for column experiments so that minimal transformations would occur (intact sediment cores, 6 cm diameter, 10 cm deep). A small part of the surface of the sediment cores (5 g total taken from three replicate cores) was slowly dried at low temperature (50°C), homogenized, and sieved keeping the silt/clay (<2000 µm) fraction (three replicates). Particle-size distribution, pH, and total P (Psenner et al., 1984) were determined on this fraction of the sediment. Also water content, loss of ignition (LOI), and carbonate content were measured based on weight losses after drying and combustion of the sediments at 105, 550, and 900°C, respectively (Kaiserli et al., 2002). Total organic content (TOC) was also determined using the Walkley-Black method (Tan, 1996). The TOC along with LOI and carbonate content are different approaches for the determination of the organic fraction, which provides useful information on the ability of the sediment to adsorb P on organic matter. Sediments were also analyzed for total metal content (Al, Mn, Fe, Ca) as well as for exchangeable metal content (extraction with CH3COONH4) (APHA, 1985; Tan, 1996). Measurements of metal concentrations were conducted with an atomic absorption spectrophotometer (PerkinElmer 2380). Phosphorus fractionation was also performed on the sediment samples, according to the schemes of Psenner et al. (1984) with modifications by Hupfer et al. (1995) and Rydin (2000) (Fig. 2).


Figure 2
View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. Fractionation scheme selected for this study, based on the scheme of Psenner et al. (1984) with slight modifications by Hupfer et al. (1995) and Rydin (2000).

 
The experiments were divided into two main sections: the first comprised column experiments showing the extent of P release from the sediments to the overlying water column under control of certain physicochemical parameters. The second series of experiments included the application of P fractionation schemes on sediment samples, before and after the series of P release experiments. This provided a range of information on the fractions and binding forms of P as well as the changes taking place after the release.

Column Experiments–Phosphorus Release
Release measurements were conducted on the intact sediment cores collected at the two sampling stations located at Lakes Koronia and Volvi. Past bibliographical references state that in most cases there is no statistically significant variation in P release rates with location, as long as the sampling sites are of related geochemical composition and they are not influenced by point-sources (Penn et al., 2000). Nevertheless, three replicates of each core were used for each sampling site to increase the validity of the results. Degradation of the sediment is not significant, since the cores were used immediately for the column experiments. The experimental design provided a relatively realistic simulation of the possible physicochemical conditions that could occur at these lakes (Fig. 3).


Figure 3
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3. Experimental apparatus used in the release experiments.

 
The sediment cores were put in glass cylinders, of approximately 70 cm length and 6 cm diameter. One liter of overlying lake water obtained from the same location was added (passed through a 0.45-µm membrane filter to retain algae). The glass tubes were stoppered with lids that contained outlets, enabling the insertion of pH and redox (WTW sentix, Ag/AgCl) electrodes. Gas lines (CO2, air, N2, and CO2–free air) were included in the experimental design. Gases were introduced with bubbling devices about 5 cm above the sediment surface, thus slightly mixing the water column and inhibiting the formation of P concentration gradients. To simulate the actual conditions that appear in a lake system throughout the year, separate cores were used for the combinations of three values of pH (7, 8, 9) and three values of redox potential (Eh = +200 mV, +100 mV, and –200 mV)—three replicates for each condition. The selection of these values was based on past studies indicating these values as more common in actual lake systems (Boers, 1991; Seitzinger, 1991; Maine et al., 1992; Moore and Reddy, 1994; Gomez et al., 1999; Penn et al., 2000).

Redox potential and pH values were controlled by gas purging to remain close to the values selected (±0.1 for the pH values and ±10 mV for Eh). Purging with CO2 or CO2–free air helped to control pH while maintaining carbonate and Ca chemistry. As a consequence, the simulation of actual conditions was more successful. Moreover, the presence of carbonate and bicarbonate ions enabled the control of pH at very low values of redox potential (Moore and Reddy, 1994). The temperature of the columns was not controlled, although room temperature was maintained at approximately 20°C. Aliquots of overlying water were obtained daily by a drain cock situated at 10 cm above the sediment surface. Aliquots were obtained until the concentration of orthophosphates became stable. The water samples were passed through a 0.45-µm filter before analysis for orthophosphates, Fe, Mn, Al, Ca, and DOC content (Shimadzu V-csh TOC Analyzer). Experiments lasted until P concentration reached an equilibrium in the overlying water.

Phosphorus Fractionation
Samples from the surface of the sediments were taken at the beginning and at the end of the incubation period with a long steel spatula and filtered through 0.45-µm filters to eliminate the interstitial water. The disruption to the core is minimal since the fractionation method requires a small amount of sample (1 g). Phosphorus fractionation was performed on these samples, according to the scheme shown in Fig. 2. The scheme proposes an initial extraction with NH4Cl 1 M (loosely sorbed P fraction), followed by 0.11 M Na2S2O4/0.11 M NaHCO3 (reductive P fraction), which represents the redox sensitive fraction of P bound on the surface of Fe(III) and Mn oxides and hydroxides. The reduction of ferric hydroxide to the soluble form of ferrous hydroxide, leads to the dissolution of the sorbed fraction of P. The next step includes treatment with 1 M NaOH, which extracts the P fraction sorbed on the surface of aluminum hydroxides (NaOH–P) and the interior of ferric oxides of the sediment particles. Desorption of P is caused by substitution of phosphates by hydroxide ions. The last of the extraction series is 0.1 M HCl, which removes the Ca-bound fraction (HCl–P) and represents the amount of P found in Ca and Mg minerals. Each extraction procedure was repeated three times so as to obtain credible results. Table 2 provides a summary of the fractions determined according to the selected scheme.


View this table:
[in this window]
[in a new window]
  		Table 2. Phosphorus fractions in the sediment and their descriptions based on the scheme of Psenner et al. (1984) with slight modifications by Hupfer et al. (1995) and Rydin (2000).

 

   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
General Features and Chemical Content of Sediment and Water Samples
The results of the physicochemical analysis of the sediment and water samples are presented in Table 3 (mean values of three replicates). The fractional composition of sediment samples taken from both lakes was dominated by the sand fraction (75–2000 µm), although in the case of Lake Koronia, the silt–clay fraction (<75 µm) was higher, enhancing the ability of this sample to adsorb various chemical species including P.


View this table:
[in this window]
[in a new window]
  		Table 3. Physicochemical properties of lake sediment and water (mean values, n = 3).

 
Sediment from Lake Koronia shows high water (49.4%) and organic content, the latter expressed as loss of ignition (4.2% dw), as well as TOC (2.3% w/w). Iron and Mn concentrations are higher, while P and Ca concentrations are almost double compared to those from Lake Volvi, as a result of the long historic external nutrient loading of the lake.

The mean concentrations of the different P fractions determined in the sediment samples as well as the relative contribution of each fraction to the sedimentary inorganic P fraction are shown in Fig. 4. Fractionation results reveal relatively similar distribution of P among the different fractions in both lakes. The predominant fraction seems to be HCl–P (Ca bound), exhibiting higher relative and absolute concentrations in Lake Koronia. The NaOH–P and redox sensitive P (BD–P) participate strongly in the overall P content. The sequence of fractions shows the following order: HCl–P > NaOH–P > BD–P > NH4Cl–P (Fig. 4). The residual P fraction remains relatively stable under most conditions, while the loosely sorbed fraction (NH4Cl–P) is readily soluble under all conditions. In some lakes NaOH–P or BD–P have been reported as the dominant fractions of the sediments. The fractionation pattern obtained here is typical of calcareous sediments with a strong presence of Fe and Mn sorbed P (Kaiserli et al., 2002).


Figure 4
View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Comparison of P fractionation results: (a) mean concentration of P fractions, (b) relative % contribution of each P fraction to the sedimentary inorganic P fraction.

 
Column Experiments
Phosphorus Release
The influence of pH and redox potential on the release of P from the sediment cores was clear. The general pattern is one of increased P release under reductive redox conditions and high pH values. The release of P is expressed as the change in P concentration in the overlying water.

In both lakes, P release is favored under extremely reductive conditions (–200 mV), and it is double at pH 9 compared to pH 8 (Fig. 5). Lake Koronia shows a similar pattern of release to Lake Volvi, although the amount of P released is not as large as expected, based on the analysis of total sedimentary P (Fig. 4a). In Lake Koronia at pH 7 and 8 the release pattern is not clear until after Day 10 of the experiment when maximum release is at pH 9 followed by pH 8 and 7, respectively. The release pattern is similar under medium reductive conditions of +100 mV. All experiments demonstrated that after a certain period of time the P concentration in the overlying water stabilizes, probably due to the equilibrium between P desorption and fixation on the sediment surface (Penn et al., 2000).


Figure 5
View larger version (11K):
[in this window]
[in a new window]
  		Fig. 5. Phosphorus release under extremely reductive conditions (–200 mV) as a function of pH, (a) Lake Volvi, (b) Lake Koronia.

 
In the case of Lake Volvi, P release is increased only slightly when the prevailing conditions are converted to extremely reductive. Variations of pH have a greater effect on the release of P. On the other hand, P release in the case of Lake Koronia seems to be more strongly related to redox changes, especially from +100 mV to more reductive conditions, presumably because of the increased mobility of the BD–P fraction of this sediment sample (Fig. 6).


Figure 6
View larger version (11K):
[in this window]
[in a new window]
  		Fig. 6. Phosphorus release at pH 9 as a function of redox potential, (a) Lake Volvi, (b) Lake Koronia.

 
This general release pattern is in accordance with past studies (Rydin, 2000; Kaiserli et al., 2002). Phosphorus is fixed at the surface of hydroxides of various metals such as Fe and Mn. At low redox potential, the release of P increases due to the apparent reduction of insoluble Fe(III) to readily soluble Fe(II) hydroxides resulting in immediate P liberation. In the case of Lake Volvi, where pH variations exert greater influence on P release than redox potential (except for Eh = +300 mV), P is deposited mainly on the Al and Fe oxide fraction and it is released under high pH values, due to ion exchange between OH– anions and adsorbed phosphates (Boström et al., 1982; Gomez et al., 1999; Reynolds and Davies, 2001). These hypotheses are investigated further and compared with sediment P fractionation (see Phosphorus Fractionation in Sediments).

Under oxic conditions (+300 mV) the release of P from the sediment of Lake Volvi is negligible, regardless of pH, as shown in Fig. 7a. Under these conditions, the formation of a brown–yellow microlayer was observed on the surface of the sediments, which is attributed to the formation of iron(III) hydroxide on the sediment surface. This layer protects the lower surface layers from further P desorption. It also adsorbs dissolved P forms (Penn et al., 2000).


Figure 7
View larger version (11K):
[in this window]
[in a new window]
  		Fig. 7. Phosphorus release under oxic conditions (+300 mV) in relation to pH, (a) Lake Volvi, (b) Lake Koronia.

 
Under oxic conditions, Lake Koronia (Fig. 7b) is of particular interest, since there was no formation of an oxidized protective layer. Moreover, release rates remained high, probably because of the activation of lower sediment layers and the contribution of the organic fraction, which is a potential source of sorbed P (Rzepecki, 1997; Gomez et al., 1999). In addition, an important factor affecting the extent of P release is the saturation status of the available sorption sites on the surface of the metal hydroxides (Clasen et al., 1982; Hoyer et al., 1982; Perkins and Underwood, 2001). In the case of sediment P sorption sites being oversaturated, the release of P is, to an extent, independent of pH and redox conditions. After a certain time, the release is stabilized, mainly because of the achievement of equilibrium between release and sorption of phosphate ions. Additionally, Löfgren and Boström (1989), has shown that limited P release could occur from "aerobic" sediment surfaces following the formation of anoxic microlayers at the sediment–water interface due to enhanced microbial activity.

During spring, algae could contribute to the rise of pH up to values around 9, a process induced by photosynthetic activity, withdrawing CO2 from the water and shifting the CO2/HCO–3/CO3= equilibrium that controls pH (Boers, 1991). Lake Koronia could probably contribute to the internal P loading even under oxic conditions, unlike Lake Volvi where a protective oxidized layer is formed on the sediment preventing P release. In Lake Volvi, even if large P quantities were liberated at low redox conditions, lake turnover in spring would bring about oxic conditions, leading to the formation of a new protective layer and further P fixation. In contrast, sediment in Lake Koronia appears to be contributing to the internal loading regardless of the physicochemical conditions. Bearing in mind the past environmental pressures inflicted on this lake and the increased frequency of anoxic incidents, it can be assumed that this lake is also threatened by a significant internal source of nutrient pollution.

Both lakes show similar P release patterns under reductive conditions, with a few interesting differences. In general Lake Koronia shows greater release rates compared to Lake Volvi. At extremely low redox values (–200 mV) P is readily liberated, due to increased mobility of the reductive P fraction. In contrast, at medium reductive values (+100 mV) maximum release occurs at pH 8, not pH 9 as expected. We can assume that in the case of Lake Koronia, the fraction mainly responsible for P liberation is reductive P and not NaOH–P, meaning that redox potential controls P release to a greater extent than pH. This is indicated in Fig. 6b and proved by the fractionation results (Fig. 4). The increased organic fraction of Lake Volvi, could contribute to these changes, since it has been proved that the organic fraction of the sediments can serve as P sorption sites. Also, the increased contribution of fine particles in the overall particle-size distribution of the sediment of Koronia (Table 3) is a factor explaining the increased sorption capacity of P in this sediment.

The mean daily P release rates were calculated and presented in Table 4. For purposes of comparison, Table 5 provides a list of P release rates reported in the literature, taking into account the surface area of the sediment exposed to the water column, volume of water, and time period of each experiment. The values reported in this study are lower in comparison with other meso-hypereutrophic lakes, possibly because of the calcareous content of the sediments which is high compared to other lakes. Phosphorus is fixed in the HCl–P (Ca–P) fraction of the sediments, which does not easily release P under common pH and redox variations.


View this table:
[in this window]
[in a new window]
  		Table 4. Mean P release rates (mg P m–2 day) for Lake Volvi and Lake Koronia.

 

View this table:
[in this window]
[in a new window]
  		Table 5. Phosphorus release rates for lake and river sediments from the literature.

 
Role of Iron and Manganese in Phosphorus Release
The reductive dissolution of Fe and Mn hydroxides is followed by liberation of adsorbed quantities of P. In this study, aliquots obtained from column P release experiments were also analyzed for soluble Fe and Mn, to determine the extent of the dissolution of their oxides and hydroxides, which are significant P-bearing compounds.

In Lake Volvi, the dissolution of Fe from the sediment is enhanced as the redox values are decreased, showing a maximum at extremely reductive conditions (–200 mV). On the other hand, the maximum Fe release is observed at the lowest pH (pH 7), obviously due to direct hydroxide dissolution (Fig. 8a). At lower redox values P release is increased by activation of the reductive P fraction. As pH rises, the prevailing process of P release is the activation of the NaOH–P fraction and the exchange of P by OH– ions at the sorption sites. If the major parameter controlling P release was redox potential, then maximum P liberation would be observed at a combination of low redox values and low pH values (pH 7). On the contrary, maximum liberation of P occurs at higher pH values (pH 9) and reductive conditions, meaning that for Lake Volvi, P is mainly mobilized by ion-exchange processes (NaOH–P fraction). It can be assumed that pH and redox have a synergistic effect on the various mobilized fractions of P, with the effect of pH prevailing in the case of Lake Volvi. Therefore, both the fractions of BD–P and NaOH–P are mobilized to a different extent. Under oxic condition only a limited amount of Fe is dissolved, due to the protective oxidized layer of ferric hydroxide on the sediment surface.


Figure 8
View larger version (11K):
[in this window]
[in a new window]
  		Fig. 8. Iron release in relation to pH, Eh = –200 mV, (a) Lake Volvi, (b) Lake Koronia.

 
Dissolution of FePO4 probably does not take place in Lake Volvi, since it has been shown that the dissolution of this mineral starts below –250 mV for pH 7 (Williams and Patrick, 1973). Roden and Edmonds (1997) on the other hand, have shown that this dissolution could take place through biologically mediated processes at more positive redox values. For Lake Volvi this process is likely to be insignificant.

At –200 mV Lake Koronia shows a similar Fe release profile to Lake Volvi, but at increased rates and amounts, due to a more significant contribution from the BD–P fraction (Fig. 8b). Part of the Fe dissolution is attributed to the increased amount of the exchangeable sedimentary Fe, as shown in Table 2. Even under oxic conditions, considerable amounts of Fe are released proving the activation of lower sediment layers with lower redox potential, resulting in P liberation. Nevertheless, the contribution of BD–P at oxic conditions cannot account for the total amount of P release. Lake Volvi, on the other hand, exhibits little or no Fe release at all under oxic conditions.

It has been shown that the Fe/TP ratio of the sediment may provide a measure of free sorption sites for orthophosphate ions on hydroxyl-oxide surfaces; therefore, if this ratio is kept above certain values then the surface remains undersaturated with sorbed P, which can be fixed under oxic conditions (Jensen et al., 1992). Different values have been proposed for this value (Kleeberg and Kozerski, 1997; Perkins and Underwood, 2001), but a realistic and useful value is 10 to 15 (by weight) (Jensen et al., 1992; Søndergaard et al., 2003). The Fe/TP ratio in sediment (by weight) in Lake Koronia is approximately 6, meaning that sorption sites appear to be closer to saturation, liberating P even under oxic conditions. In contrast, Lake Volvi has a ratio of approximately 10, showing limited P saturation and therefore limited internal loading under oxic conditions.

A correlation of the released concentration (mmol) of Fe and P provides valuable information on the processes of P release and on the mobilized P fractions (Hoyer et al., 1982). In the case of Lake Volvi, at high pH (pH 8–9) and reductive conditions (Eh = –200 mV) the relationship is logarithmic and not linear, as would be expected if P were fixed entirely on the redox-sensitive fraction and released in stoichiometric proportion to Fe (Fig. 9a). Instead, at high pH, P is quickly released at first (evidently with the contribution of the NaOH–P fraction) and BD–P is gradually mobilized. The release tends to be determined by the redox-sensitive fraction and more specifically by the Fe hydroxyl-oxides that contain P. At neutral pH (pH 7), the relationship is linear, showing a proportional release of P and Fe, mainly due to the enhanced or total participation of the BD–P fraction.


Figure 9
View larger version (13K):
[in this window]
[in a new window]
  		Fig. 9. Relationship of released mmol P and mmol Fe (Eh = –200 mV), (a) Lake Volvi, (b) Lake Koronia.

 
The relationship between the released Fe and P is linear in the case of Lake Koronia (except for pH 9), indicating higher contribution of the BD–P fraction to the release of P from this sediment sample (Fig. 9b).

Manganese can be liberated under the same reductive conditions and similar mechanisms, leading to P liberation. The difference is that it reduces and dissolves far more rapidly and at higher redox values than Fe(III). Manganese is a very sensitive redox indicator and therefore gives a good indication of the start of oxygen depletion. Manganese release patterns for Lake Volvi seem similar to those of Fe, thus providing us with quantitative information concerning P liberation (Fig. 10a). Due to the faster reduction and dissolution of Mn than that of Fe, it could slow or inhibit the reduction of Fe, slowing the mechanisms of P liberation (Boström et al., 1982; Hoyer et al., 1982). This could be assumed to be the reason of the limited initial P release at Lake Volvi.


Figure 10
View larger version (10K):
[in this window]
[in a new window]
  		Fig. 10. Manganese dissolution related to pH at Eh = +100 mV, (a) Lake Volvi, (b) Lake Koronia.

 
In contrast, Lake Koronia shows increasing Mn release only under extremely reductive conditions. At +100 mV (Fig. 10b) and +300 mV, Mn release decreases with time. Though the total Mn content of the sediment is high, the exchangeable portion is limited. As far as Lake Koronia is concerned, the contribution of Mn to the P release cannot be substantiated and quantitatively related. It has been shown in the past that the concentration of P in the sediments is controlled by Fe and to a restricted extent by Mn, which serves as a highly sensitive indicator for the formation of anaerobic conditions (Boström et al., 1982).

Role of Calcium, Aluminum, and Dissolved Organic Carbon
Measurements of Ca and Al concentrations were conducted to determine their role in sediment P equilibrium and to obtain information to support the P fractionation study.

Concentrations of Ca released from the sediment to the overlying water are generally unaffected by the variations of pH and redox. At pH 7 only there is a considerable amount of Ca dissolved (Fig. 11). This does not seem to relate to P release from the sediment, since Ca concentration remained unchanged even at redox +300 mV, where P release is low (Lake Koronia) or zero (Lake Volvi). In general, the fraction of HCl–P does not seem to be affected by the physicochemical changes at different redox potentials. A limited decrease of Ca concentrations is observed at Lake Koronia at higher pH, indicating minimal participation of the calcareous fraction of P and therefore not affecting the overall P release. Under oxic conditions and low pH, the slight Ca dissolution could be related to the dissolution of P. At higher pH the calcareous fraction does contribute to the release of P, since a precipitation of Ca was observed. The calcareous fraction of both lakes seems to be stable and is increased only by precipitation procedures. Although it serves as a major P retention portion of the sediment, it acts as a sink of P rather than a P source.


Figure 11
View larger version (12K):
[in this window]
[in a new window]
  		Fig. 11. Changes in Ca concentration as a function of pH (Eh = –200 mV) in the case of Lake Koronia.

 
The analysis of Al and DOC during the column experiments was supplemental to the main study. The Al results do not provide additional information regarding the influence of Al on the release of P. This suggests that the Al-bound P is released only through ion-exchange mechanisms, without affecting the crystal lattice of the Al-oxides, where P is fixed.

The possible dissolution of organic matter related to P release was investigated with DOC measurements of the water aliquots. In the case of Lake Koronia, DOC has a slight tendency to increase at low redox but not to the same extent under oxic conditions. It can be assumed that certain bacteria mineralize a part of the sediment organic matter under reductive conditions, contributing to a partial liberation of loosely bound P (Boström et al., 1982; Kim et al., 2003). Lake Volvi does not exhibit the same tendency. Only a small amount of DOC is dissolved but it is probably not related to the P equilibrium, as it is stable for both reductive and oxidative conditions.

Phosphorus Fractionation in Sediments
In the case of Lake Volvi, maximum P liberation occurs under reductive conditions (Fig. 12a), as a result of the BD–P fraction dissolution. This supports the findings in the column experiments relating Fe dissolution and P release. As pH is increased, it is apparent that more P is released, due to the activation of the NaOH–P fraction. The mobilization of BD–P seems to be stable both at extremely reductive as well as slightly reductive conditions (Fig. 12). It has already been discussed that reductive Fe dissolution occurs under +100 mV, which means that P release from the BD–P fraction eventually reaches the same level. The P fractionation scheme shows the synergistic effect of high pH and low redox values on the overall liberation of P from the sediments, which is maximum at pH 9 and Eh = –200 mV. At pH 8 the HCl–P fraction is slightly increased mainly due to partial sorption of P on the calcareous part of the sediment.


Figure 12
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 12. Lake Volvi: P fractionation before and after the column experiments, (a) Eh = –200 mV, (b) Eh = +100 mV.

 
Under oxic conditions and pH 9, the fractionation pattern remains relatively stable (Fig. 13). This occurs due to the formation of the oxidized microlayer, mentioned previously. Only a slight increase of the NaOH–P and BD–P fractions is observed, possibly due to further P fixation and the changes in the fractional distribution. It is also noticeable that the NaOH–P fraction is more effectively mobilized under both high pH and low redox, probably because of the gradual formation of the protective oxidized microlayer as there is a shift to more positive redox values. Also a large amount of P is fixed by the calcareous portion of the sediment.


Figure 13
View larger version (33K):
[in this window]
[in a new window]
  		Fig. 13. Lake Volvi: P fractionation before and after the column experiments, pH 9.

 
The P sediment fractionation results prove the previous assumptions that were based on the column experiments. The data related to P release, as well as Mn and Fe dissolution, are in agreement with the P fractionation results.

In the case of Lake Koronia, the readily soluble P amount is greater than that of Lake Volvi. The BD–P fraction reduction is almost double in Lake Koronia compared to that of Lake Volvi, liberating increased P amounts even under medium reductive conditions +100 mV (Fig. 14a). The results agree with the extensive Fe dissolution observed during the column experiments for Lake Koronia, proving that the P release in this case is controlled mainly by changes in redox conditions. Worth noticing is that the decrease of the NaOH–P fraction is limited as pH shifts to lower values. At the same time, the BD–P fraction shows a tendency to decrease, compensating in terms of P release for the lack of extensive NaOH–P mobility. In combination, these results lead to the conclusion that the BD–P fraction is the determining factor controlling P release for Lake Koronia under reductive conditions, with supplementary participation of the NaOH–P fraction.


Figure 14
View larger version (23K):
[in this window]
[in a new window]
  		Fig. 14. Lake Koronia: P fractionation before and after the column experiments, (a) Eh = +100 mV, (b) Eh = +300 mV.

 
Under oxic conditions (Fig. 14b), the fractionation results of the sediment of Lake Koronia, show a limited participation of the three major mobile fractions and do not provide statistically significant data to fully clarify the origin of the released P. At +300 mV Lake Koronia exhibited limited but continuous P release, which could be attributed to the mobilization of the BD–P fraction of lower sediment layers and to the oversaturation of the available P sorption sites. The fractionation results after the column experiments for Lake Koronia at oxic conditions reveal a relatively small participation of the BD–P fraction in P release, which is constrained, because of the partial Fe oxidation on the surface (decreased Fe dissolution during the column experiments) and the resorption of P on BD–P. Sediments from Lake Koronia contain higher organic content than Lake Volvi and they have a relatively increased fraction of fine sized particles (<75 µm) (Table 3); therefore, they are capable of retaining P and liberating it even under oxic conditions. It is worth noticing that at all pH values, part of the dissolved P reprecipitates to the HCl–P fraction, which seems to increase, especially at pH 8. This is not accompanied by Ca precipitation during the column experiments, leading to the conclusion that P is partly sorbed on this fraction rather than precipitating with Ca minerals.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Phosphorus liberation from sediments in both lakes is greatly influenced by variations of pH and Eh. The results of the column experiments in relation to the sedimentary P fractionation provide information on the extent and the major processes of internal P loading of the lakes.

Lake Volvi exhibits lower P release rates than Lake Koronia. The major mechanisms leading to internal P loading for Lake Volvi seem to be (i) P desorption due to the reductive dissolution of Fe oxides and hydroxides at low redox potential, and (ii) ion-exchange with OH– at higher pH. The second mechanism seems to be of major importance in Lake Volvi, as the combination of column experiments and sedimentary inorganic P fractionation has indicated. Variations in pH have a greater influence on P release rates rather than redox itself; therefore, the NaOH–P fraction is expected to be of greater importance to the overall release. This is proved by the P fractionation results. Furthermore, the formation of a protective oxidized microlayer of iron(III) hydroxide on the surface of the sediment under oxic conditions seems to prevent P dissolution or desorption, thus inhibiting further internal loading under these conditions. Phosphorus fractionation results agree with those obtained by the column experiments, showing stability of all P fractions due to the formation of the oxidized layer.

In Lake Koronia, the activation of the reductive P fraction, triggered by the decrease of Eh, has been proved to be of major importance and is supported by quantitative results related to the major parameters affecting the equilibrium of P at the sediment–water interface. This is confirmed by the Fe dissolution at these redox values, which remains high, as well as the changes in the fractionation patterns indicating high mobilization of this fraction. Phosphorus release through ion-exchange mechanisms occurs at high pH values but in the case of Lake Koronia, it is not the major driving force of the release. At lower pH, where ion-exchange is limited, increased participation of BD–P in the total P pool is observed and proved by the fractionation schemes. Unlike Lake Volvi, P release in Lake Koronia occurs even at +300 mV (oxic conditions), but to a reduced extent (compared with reductive conditions), possibly because of the oversaturation of P sorption sites on the surface. The smaller particle size composition and the increased organic content indicate a higher capacity for P adsorption, which could be released even under oxic conditions.

Manganese is liberated in a similar way as Fe, although in some cases it could slow or inhibit the dissolution of Fe and therefore the liberation of P. The results are not clear as far as its quantitative role is concerned. Calcium dissolution does not seem to affect the equilibrium of P significantly in these lakes, since the major mechanisms of P release are different.

These two lakes demonstrate low P release rates compared with other European and international cases. This could be attributed to the comparably shorter time of exposure to external P sources and the increased calcareous content of the lakes, which constitutes a stable P pool.

Because of their strong impact on lake water concentrations, it is clear that knowledge of sediment–water interactions and the processes behind retention and release of P is fundamental for understanding the function of shallow lakes. The combination of the results from the column experiments and the sedimentary P fractionation provides information toward a complete understanding of the P release pathways. These findings could contribute to more effective management of the studied lakes. It is clear that management strategies directed toward preservation of aerobic conditions at neutral pH will minimize release rates and therefore internal loading. Nevertheless, an important prerequisite for obtaining success and long-term minimization of eutrophication effects is elimination of the remaining external P loading.


   ACKNOWLEDGMENTS
 
The authors wish to thank the Greek National Foundation of Scholarships for the financial assistance during the course of the study.


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





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 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 ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Christophoridis, C.
Right arrow Articles by Fytianos, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Christophoridis, C.
Right arrow Articles by Fytianos, K.
GeoRef
Right arrow