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


     

Published online 31 August 2007
Published in J Environ Qual 36:1534-1544 (2007)
DOI: 10.2134/jeq2006.0490
© 2007 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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Surridge, B. W. J.
Right arrow Articles by Baird, A. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Surridge, B. W. J.
Right arrow Articles by Baird, A. J.
Agricola
Right arrow Articles by Surridge, B. W. J.
Right arrow Articles by Baird, A. J.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Phosphorus
Right arrow Nutrient Cycling

TECHNICAL REPORTS

Wetlands and Aquatic Processes

The Release of Phosphorus to Porewater and Surface Water from River Riparian Sediments

Ben W. J. Surridgea,*, A. L. Heathwaiteb and Andrew J. Bairdc

a Catchment Science Centre, Kroto Research Inst., The Univ. of Sheffield, Broad Lane, Sheffield, S3 7HQ, UK
b Centre for Sustainable Water Management, The Lancaster Environment Centre, Lancaster University, LA1 4YQ
c Dep. of Geography, Queen Mary, Univ. of London, Mile End Road, London, E1 4NS

* Corresponding author (b.surridge{at}sheffield.ac.uk).

Received for publication November 9, 2006.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
Sediments can be both a source and a sink of dissolved phosphorus (P) in surface water and shallow groundwater. Using laboratory mesocosms, we studied the influence of flooding with deionized water and simulated river water on P release to solution using sediment columns taken from a riparian wetland. The mesocosm incubation results showed that rather than retaining nutrients, sediments in the riparian zone may be a significant source of P. Concentrations of dissolved P in porewater reached more than 3 mg L–1 and in surface water over 0.8 mg L–1 within a month of sediment inundation. The reductive dissolution of P-bearing iron (Fe) oxides was the likely mechanism responsible for P release. Dissolved P to Fe molar ratios in anaerobic samples were approximately 0.45 when columns were flooded with water that simulated the chemistry of the adjacent river. This suggests there was insufficient Fe in the anaerobic samples to precipitate all P if the solutions were oxygenated or transported to an aerobic environment. If the anaerobic wetland solutions were delivered to oxygenated rivers and streams adjacent to the riparian zone, the equilibrium concentration of P in these systems could rise. The timing of P release was inversely related to the nitrate (NO3–) concentration in floodwater. This indicates that in riparian zones receiving low nitrate loads, or where NO3– loads are being progressively reduced, the risk of dissolved P release may increase. These findings present particular challenges for restoration and management in riparian areas.

Abbreviations: P, phosphorus • Pi, inorganic phosphorus • Po, organic phosphorus • MRP, molybdate reactive phosphorus • NO3–, nitrate • Fe2+, ferrous iron • DW, core incubations using deionized water as the flooding solution • SRW, core incubations using simulated river water as the flooding solution


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
THE river riparian zone, in an unaltered state synonymous with the river floodplain, is a critical interface in catchments. Located between aquatic and terrestrial ecosystems, it is a zone where the chemical, and in particular the nutrient, composition of shallow groundwater and surface water can be altered by a range of biogeochemical processes. Understanding these processes, and their impact on nutrient concentrations and loads, is a priority for protecting adjacent aquatic ecosystems, particularly in relation to the major issue of eutrophication that is central to the European-wide Water Framework Directive (Meyer, 1997; Butturini and Sabater, 1998; Hedin et al., 1998; Burt, 2005).

Research has shown that NO3– in shallow groundwater can be removed in riparian zones, predominantly through denitrification during heterotrophic microbial activity or through plant uptake (Lowrance et al., 1984; Peterjohn and Correll, 1984; Hill, 1996; Hill et al., 2000; Clement et al., 2003). Compared with NO3–, the influence of river riparian zones on P in shallow groundwater has received little attention. ‘Shallow groundwater’ is taken here to mean groundwater in the uppermost aquifer, within a maximum of 5 to 10 m of the ground surface. Most work concerned with P cycling in river riparian zones has focused on surface runoff, a major pathway for sediment-bound P transported from agricultural uplands to surface waters. Riparian zones can reduce particulate P concentration in surface runoff by promoting sedimentation, plant uptake, and adsorption processes (Dillaha et al., 1989; Lowrance et al., 2002; Lee et al., 2003; Kronvang et al., 2005). However, some evidence indicates that soils and sediments in riparian zones may act as sources of P to shallow groundwater (Jordan et al., 1993; Carlyle and Hill, 2001), or that P in shallow groundwater is not removed as effectively as NO3– (Osborne and Kovacic, 1993; Vanek, 1993). Recent evidence from organic-rich sediments has also highlighted the release of dissolved P to surface water and porewater. This so-called internal eutrophication is often associated with increased sulfate loading to these sediments as a result of anthropogenic sulfur pollution of the atmosphere, and of surface waters and groundwaters (Lamers et al., 1998, 2001, 2002; Lucassen et al., 2004). Dissolved inorganic P (Pi) limits primary productivity in many freshwater ecosystems (Hudson et al., 2000; Karl, 2000; Blake et al., 2005; Colman et al., 2005), and eutrophication of these systems has been predominantly linked to increased P loads and concentrations (Bowes et al., 2003; Jordan and Rippey, 2003). Therefore, P transformations in riparian zones, and in particular the release of dissolved P, may have important consequences for the chemical and biological status of the riparian environment itself, but also of adjacent aquatic ecosystems.

Our research focuses on riparian wetlands, many of which in the UK, Europe, and North America have been drained over the last few centuries to satisfy the demand for new agricultural land. To take advantage of their perceived ability to enhance water quality, there are now renewed efforts to conserve and restore riparian wetlands (Verhoeven et al., 2006). However, recent evidence has shown that restoration of riparian wetlands may lead to substantial P release to solution. These restored systems then represent P sources for adjacent aquatic ecosystems, even several years after restoration (Rupp et al., 2004; Zak et al., 2004). A better understanding of how P is transformed in riparian zones would be valuable for future wetland conservation and restoration strategies.

This article examines the release of P under changing hydrological conditions in a non-intensively drained riparian wetland. Although the total P content of wetland sediments not exposed to external nutrient sources is often relatively low, typically below 500 mg kg–1 total P (White and Reddy, 1999; Craft and Casey, 2000; Fisher and Reddy, 2001; Wright et al., 2001), water fluxes to riparian zones from adjacent rivers can deliver high P loads (Darke and Walbridge, 2000; Owens and Walling, 2002; Olde Venterink et al., 2003). A large proportion of the P transported by rivers is in particulate form (Foster et al., 1996; Walling et al., 1997), and because of reduced water velocity and sediment transport capacity in riparian zones, sedimentation and storage of particulate P can occur. Drainage of riparian wetlands to support agricultural activity may also lead to changes in the forms of P held within the sediment, for example because of mineralization of organic forms of P. These changes may significantly affect the long-term stability of the stored P.

This article investigates the potential for releases of P to surface water and porewater from stores of P held in riparian wetland sediments. Such releases could affect the internal nutrient status of the wetland. Given sufficient hydrological connectivity, releases of P from the sediment could also mean that riparian zones act as present-day sources of P to adjacent rivers and streams. The importance of these processes may grow as other sources of P are reduced, for example by investment in P-removal technology at sewage treatment works. The research questions addressed in the study were:

  1. Is P released to surface water and porewater from riparian sediments during inundated conditions which are common in these near-channel zones?
  2. If hydrological connectivity existed in the field, is there a potential for P in surface water and porewater to alter the equilibrium concentration of P in adjacent aquatic ecosystems?


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion
 Conclusions
 REFERENCES
 
The research was conducted using samples of peat collected from Strumpshaw Fen, a wetland complex covering 200 ha of the floodplain of the River Yare in Norfolk, UK (52°36' N, 1°26' E; Fig. 1 ). Major water sources to Strumpshaw Fen are rainfall and floodwater from the River Yare and the Lackford Run, a small tributary of the River Yare flowing along the northwest boundary of the wetland. Deeper groundwater from the major regional aquifer is prevented from entering the wetland by low permeability sands and clays at the interface between the base of the peat deposits and the underlying aquifer (Surridge, 2004).


Figure 1
View larger version (112K):
[in this window]
[in a new window]

  		Fig. 1. View of the Strumpshaw wetland complex in the floodplain of the River Yare looking down-river (Photograph courtesy of Mike Page, Royal Society for the Protection of Birds, RSPB). Note the ditch network running through the complex, the Lackford Run in the northwest of the wetland, and the hatched box showing the area from which cores were collected in this study.

 
The ditch network shown in Fig. 1 is used to manage two-way exchanges of water between the river and the fen to control water table position within the wetland. Management of the fen currently involves maintenance of surface inundation from late autumn/early winter through to late spring/early summer when evapotranspiration demand drives water tables below the peat surface. During mid to late summer, water tables can drop to over 50 cm below the ground surface. Surface-inundated conditions are re-established through changes in the precipitation-evapotranspiration balance, controlled influx of river water through the ditch network, or more rapidly as a result of over-bank flooding events from the River Yare and Lackford Run.

Mesocosm Collection and Pretreatment
A controlled laboratory experimental approach was used to investigate changes in P concentrations in surface water and porewater. The focus was on rewetting of the peat and subsequent flooded conditions. Substantial changes in osmotic and redox conditions are likely to occur during changes in water table position, with potential impacts on the concentration of dissolved P in surface water and porewater. In addition, rewetting of the peat in the field coincides with exchanges of water between shallow groundwater in the fen and surface water in the adjacent ditch network (Surridge, 2004). As a consequence, P released from the sediment during this time may be transported into the adjacent ditch network and ultimately to the River Yare.

The experiments were conducted by incubating intact cores of the upper 50 cm of peat (diameter = 24 cm) in polyvinyl chloride (PVC) cylinders (Fig. 2 ). Six cores were collected from random locations within a 20 by 20 m area located 10 m away from a section of the ditch network (see Fig. 1). Within the same area, a further nine 50 cm-deep intact cores were collected. These cores had a smaller diameter (10 cm) and were used for sediment P fractionation and analysis of other peat physicochemical properties. All cores were collected during June 2003 when water tables had been at least 40 cm below the peat surface for the preceding 20 d. The vegetation on the surface of the cores remained intact throughout the collection process and during the subsequent incubations.


Figure 2
View larger version (127K):
[in this window]
[in a new window]

  		Fig. 2. Core housing used in the incubation of larger diameter cores. Sampling ports with mini-piezometers installed are on the left side of the core housing (A), sampling ports on the right side (B) were not used in this study. Vertical tube (C) is connected to the housing through a port at the base (D) and gives the position of the water table within the core. Water is input to/drained from an identical port on the opposite side of the housing (hidden from view). Core housing is approximately 90 cm in length. Note that the image was taken at the end of the incubations after vegetation was removed from the core for analysis.

 
Before starting the incubations, porewater was allowed to drain under gravity from the cores, and evapotranspiration was allowed to occur from the peat surface for a period of 10 d as a standardization procedure. We do not believe this had a significant impact on P concentrations during the subsequent incubations because water tables were already at 40-cm depth when the cores were extracted from the field and had been for over 20 d, and water tables regularly fall to greater than 50-cm depth (the base of the cores) in the field, exposing sediment at this depth to aerobic conditions. Toward the end of the standardization period, mini-piezometers (diameter = 2 cm, length = 15 cm) were installed horizontally at 2.5, 10.0, 17.5, 32.5, and 47.5 cm below the peat surface to allow collection of porewater samples. The piezometers were constructed from inert PTFE (TEXfluor tubing, Parker TexLoc, Fort Worth, Texas) and their intakes drilled (40% total perforation) to allow porewater to enter the instrument easily. A PTFE mesh (Emflon mesh membrane, Pall Corporation, East Hills, New York) was placed around the intake of each piezometer to prevent the ingress of sediment.

Flooding of Cores and Water Sample Collection
Table 1 shows the matrix of treatments applied to the sediment cores during this study. Incubations were conducted on three of the six larger diameter cores following flooding with deionized water (hereafter DW, > 18 M {Omega}). On the three remaining larger diameter cores, and on three smaller diameter cores, incubations were conducted following flooding with water that simulated the chemistry of the River Yare (hereafter SRW). No water samples were collected from the smaller diameter cores during these incubations, but otherwise they were incubated in an identical manner to those of larger diameter. The remaining six smaller diameter cores were not flooded but were incubated in a drained state alongside the flooded cores during incubations using SRW. The water table in these six cores was maintained throughout the incubation at the level reached at the end of the preceding 10 d drainage period.


View this table:
[in this window]
[in a new window]

  		Table 1. Treatment matrix for cores incubated during this study. DW = deionized water, SRW = simulated river water. Pretreatment involved a 10 d drainage period as a standardization procedure. For details of porewater and surface water sample collection and analyses, and sediment P and sediment physicochemical analyses, see text.

 
The composition of the SRW was based on analyses of samples taken from the River Yare, and was made using standard salts dissolved in deionized water to give the final composition of (all mg L–1): Cl– = 70, SO42––S = 20, molybdate reactive P (MRP) = 0.46, Ca2+ = 100, Mg2+ = 4, Na+ = 60, K+ = 6, NO3––N = 10, alkalinity (as CaCO3) = 200, and pH = 7.80. Incubations using SRW did not include bacteria or algae in the flooding solution. Subsequent biologically mediated processes were dependent on those populations already present within the sediment cores. However, bacterial abundance can be extremely high in riparian sediments, at least of the same order of magnitude as that in river water samples and often in excess of that in groundwater (Alfreider et al., 1997; Brugger et al., 2001; Griebler et al., 2002; Feris et al., 2003). Therefore we do not believe the lack of added bacteria or algae in the flooding solutions significantly altered the processes occurring during our incubations compared to those which would occur naturally. All incubations were conducted in constant temperature fridges set to 15°C (±0.1°C). In the field, porewater temperatures are in the range of 8 to 17°C during late autumn/early winter, when flooding of the peat occurs (Surridge, 2004).

In both incubations water was introduced slowly from the base of the cores over a period of 5 to 6 h. When the water table reached the sediment surface, floodwater depth was adjusted by slow addition of water directly to the surface to give 10 cm of standing water. Samples were collected from the surface water and each of the mini-piezometers in larger diameter cores after 1, 6, 12, 24, 72, 144, 240, 360, 504, and 672 h (or 1, 6, 12, and 24 h and 3, 6, 10, 15, 21, and 28 d) of inundation. In incubations using SRW, additional samples were collected immediately after standing water appeared at the sediment surface (= 0 h), and additional samples were collected after 1032 h (43 d). Sampling protocols, involving the purging of all sampling apparatus with nitrogen and argon gas, were developed to maintain the redox integrity of the surface water and porewater solutions during collection.

Porewater and Surface Water Analysis
Water samples were processed in a glove bag under an inert N2 atmosphere to maintain redox integrity. Samples were analyzed for pH and dissolved oxygen (CD70 pH meter and OX60 dissolved oxygen meter; WPA, Cambridge, UK), and subsequently filtered through 0.45 µm cellulose nitrate acetate membrane filters. After discarding the first 10 mL to waste to minimize any contamination or interaction with the filter, a further 10 mL of the filtrate were acidified to pH < 2 using 1% v.v. 4 M H2SO4 and stored in argon-filled tubes. The acidified samples were analyzed for Fe using a 1,10 Phenanthroline monohydrate method, modified from the work of Stucki and Anderson (1981). The concentration of ferric iron (Fe3+) in all samples was below the detection limit (0.08 mg l–1), indicating that all iron was present in the ferrous form (Fe2+).

Acidified samples were also analyzed for major cations (Ca2+, Mg2+, Na+, K+, Mn2+, and Al3+) by atomic absorption spectrometry (AAS, 1100B Spectrometer; PerkinElmer, Wellesley, MA). The remaining unacidified filtrate was stored in argon-filled tubes under a N2 atmosphere until analysis for major anions (Cl–, NO3–, and SO42–) by ion chromatography (DX100; Dionex, Sunnyvale, California). The same unacidified filtrate was analyzed for MRP by automated colorimetry using flow injection analysis (FIAstar 5000; Foss Tecator, Warrington, UK). All sample dilutions were conducted using deoxygenated blank solutions within a glove bag filled with N2.

We used MRP determined on 0.45-µm filtrates as a surrogate for the orthophosphate ion. Previous work has shown that MRP determined in this way may not be composed solely of the orthophosphate ion but instead include colloidal P released under the analytical conditions of the MRP determination (e.g., Baldwin, 1998; Shaw et al., 2000; Zhang and Oldham, 2001; Hens and Merckx, 2002). However, sequential filtration of samples collected during our core incubations to <0.025 µm suggested that ≥95% of the MRP determined on 0.45-µm filtrates was likely to be present as the orthophosphate ion (data not presented).

Sediment-Bound Phosphorus
Changes in sediment-P fractions due to flooding were examined by comparing flooded and non-flooded replicates at the end of the incubations using SRW. Three smaller diameter cores that had been flooded, and three that had remained drained throughout the incubation, were cut into sections at 5-cm intervals and analyzed using the P fractionation scheme developed by Reddy et al. (1998) for organic soils. Non-dried subsamples (0.5 g dry weight equivalent) were sequentially extracted with 1 M KCl (labile P), 0.1 M NaOH (Fe- and Al-bound Pi, and alkali-extractable organic P (Po)), and 0.5 M HCl (Ca- and Mg-bound Pi). The residue from the sequential extraction was digested using concentrated H2SO4 and a selenium catalyst and analyzed for P: this fraction is termed residual Po (Reddy et al., 1998). Cores that had been flooded were assumed to be anaerobic and the fractionation scheme was altered to maintain redox integrity. Although widely used in studies of wetland sediments, the scheme uses operationally defined fractions on the basis of their chemical solubility. Our research did not aim to test the selectivity or efficiency of the scheme beyond previous work reported in the literature.

Separate subsamples of each 5-cm core section were analyzed for labile Pi and Po using 0.5 M NaHCO3 extraction after Ivanoff et al. (1998), and P associated with microbial biomass using CHCl3 treatment and 0.5 M NaHCO3 extraction after Hedley and Stewart (1982). Soil pH was determined on duplicate fresh soil samples from each 5-cm section using a deionized water to soil ratio (volume (cm3)/weight (g)) of 2.5:1.

On the final three cores of smaller diameter which had been incubated in a drained state alongside the flooded cores, selected physicochemical properties of the peat were examined. The cores were sectioned at 5-cm intervals and oven-dried at 70°C to constant weight. 1 M HCl extracts of oven-dry soil were analyzed for Ca2+ and Mg2+ by AAS and MRP by flow injection. Following Roden and Edmonds (1997), oven-dry soils were extracted with 0.5 M HCl followed by analysis for Fe by AAS. The 0.5 M HCl extraction has been shown to remove Fe2+ and Fe3+ from amorphous and poorly-crystalline phases, leaving crystalline phases largely unaltered (Tuccillo et al., 1999). Finally, further subsamples of oven-dry soil were extracted with 0.2 M acid ammonium oxalate (pH = 3) for non-crystalline aluminum oxides and hydroxides following Darke and Walbridge (2000).

Data Analysis
The Spearman rank correlation coefficient was used to test for correlations among concentrations of different elements during the incubations. Causal relationships between variables were assessed using linear regression. Two-factor ANOVA tests (flooded vs. non-flooded treatment and depth as the two factors) were used on raw, and where necessary, transformed data to assess if flooding had significant impacts on sediment-bound P fractions. Paired t-tests were conducted to check for significant differences between different fractions of the sediment-bound P in individual cores. The significance level used for all tests was p = 0.05.


   Results
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion
 Conclusions
 REFERENCES
 
Sediment-Bound Phosphorus Fractionation under Drained Conditions and the Release of Phosphorus after Flooding
Large quantities of total P (TP) were present under drained conditions in the upper 15 cm of peat at Strumpshaw Fen (>1200 mg kg–1), and concentrations decreased to more constant values at greater depth (Fig. 3 ). Similar depth profiles were observed for total organic P (TPo) and total inorganic P (TPi). The majority of TP was present as TPo, although up to 30% of TP was present as TPi in any given depth increment. Table 2 provides detailed information on the operationally-defined forms of P present in the sediment samples. TPi was dominated by NaOH and HCl fractions, TPo by residual Po. There were only small pools of labile Pi and Po, and no evidence was found for the hydrolysis of Fe-bound P during the bicarbonate extraction because KCl-Pi and NaHCO3–Pi were not significantly different (paired t test, p > 0.05). Between 6 and 9% of TP was present as microbial biomass.


Figure 3
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 3. Total phosphorus (TP), total organic phosphorus (TPo), and total inorganic phosphorus (TPi) in the surface 50 cm of peat in drained cores. Mean values from replicate cores are given (n = 3), error bars show ±1 standard deviation. TP is the sum of TPo and TPi, TPo the sum of NaOH-Po and residual-Po, and TPi the sum of KCl-Pi, NaOH-Pi, and HCl-Pi from the P fractionation scheme of Reddy et al. (1998).

 

View this table:
[in this window]
[in a new window]

  		Table 2. Fractionation of sediment-bound P in the upper 50 cm of peat at Strumpshaw Fen under drained conditions. Mean values from replicate cores are given (n = 3) with 1 standard deviation in parentheses. BD = below detection, Pi = inorganic P, Po = organic P.

 
Hydrolysis of Po during the sodium hydroxide step may result in erroneously high estimates of NaOH-Pi in organic-rich soils (Qualls and Richardson, 1995; Reddy et al., 1998). The similarity between TPi from the P fractionation scheme and Pi following 1 M HCl extraction can be used to test for this potential error (Reddy et al., 1995). These two determinants were strongly correlated in all samples from Strumpshaw Fen (r = 0.97, p < 0.01; TPi = 1.0036 x Pi(1 M HCl) + 9.1094), indicating that TPi from the P fractionation scheme did not differ significantly from 1 M HCl-extractable Pi. This suggests that there was only minimal hydrolysis of Po during the NaOH step of the P fractionation scheme. NaOH-Pi and 0.5 M HCl-Fe were also strongly correlated (r = 0.92, p < 0.01), suggesting that P extracted during the sodium hydroxide step represents P attached to solid-phase Fe. No significant correlation was found between NaOH-Pi and oxalate-extractable aluminum (p > 0.05).

During incubations using DW, the concentration of MRP in surface water began to rise soon after inundation and increased throughout the incubations, reaching final concentrations between 0.79 and 0.95 mg L–1 P (Fig. 4 .). Samples from 2.5 and 10 cm contained up to 0.1 mg L–1 P 1 h after flooding, and these concentrations remained relatively constant over the first 24 h (Fig. 5 ). Beneath 10 cm, porewater concentrations of MRP remained below the limit of detection during the first 24 h of flooding (<0.015 mg L–1 P). Large increases in porewater MRP concentration in all samples from incubations using DW began between 24 and 72 h, and releases of P continued throughout the incubation. The rate of MRP release to porewater increased between 24 and 72 h, and reached a maximum of 0.014 mg P L–1 h–1 in samples from 10 and 17.5 cm. Rates declined after 72 h and were an order of magnitude lower at the end of the incubation. Porewater MRP concentrations reached a maximum at 17.5-cm depth in all three cores, varying between 2.7 and 3.1 mg L–1 P. Concentration gradients suggest that diffusion of MRP occurred toward both the surface water and deeper porewater from 17.5-cm depth. The concentration gradients were maintained from the beginning of significant MRP release (24–72 h) until the end of the incubations.


Figure 4
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 4. Molybdate reactive phosphorus (MRP) concentration in surface water samples from replicate cores flooded with deionized water (DW) or simulated river water (SRW).

 

Figure 5
View larger version (28K):
[in this window]
[in a new window]

  		Fig. 5. The release of molybdate reactive phosphorus (MRP) into surface water and porewater during incubations using deionized water (DW) or simulated river water (SRW). Profiles showing average concentrations from replicate cores at each sampling time are given (n = 3). Error bars showing ±1 standard deviation are only shown for the final sampling time to maintain the clarity of the figure.

 
In incubations using SRW, MRP was initially removed from the surface water of each core, and releases only occurred after 504 h (Fig. 4). This contrasts with more rapid MRP release to the surface water during incubations using DW. Final surface water MRP concentrations during incubations using SRW were between 0.55 and 0.80 mg L–1 P. Figure 5 also shows the release of MRP to porewater during incubations using SRW, again contrasting with that during incubations using DW. Values from samples taken before 144 h have been omitted because all porewater MRP concentrations decreased to approximately 0.08 mg L–1 P within 12 h, and subsequently remained stable until 144 h. Increases in MRP concentration were observed in porewater before surface water, and the time of MRP release was delayed for shallower compared to deeper porewater. Concentration gradients developed between deeper and shallower porewater, and between porewater and surface water, but increased surface water MRP concentration was not observed until after 504 h. Increased rates of MRP release occurred particularly for samples at 17.5-, 32.5-, and 47.5-cm depth, with maximum rates of 0.009 mg P L–1 h–1. Maximum rates of MRP release coincided with initial increases in MRP concentration at each porewater sampling depth.

The average flux of MRP to the overlying water column during the DW incubations was 2.8 mg P m–2 d–1 (standard deviation = 0.35 mg P m–2 d–1), while during the SRW incubations it was 3.2 mg P m–2 d–1 (standard deviation = 0.57 mg P m–2 d–1) between 504 and 1062 h.

Changes in Sediment-Bound Phosphorus during Flooding
Compared to non-flooded replicates, there is clear evidence that the NaOH-Pi fraction in the sediment of cores that were flooded was significantly lower at the end of the incubations (Table 3). Results from the ANOVA tests showed that depth in the cores was also a significant factor affecting the magnitude of the decrease in NaOH-Pi. Below 25-cm depth, NaOH-Pi was almost fully depleted during the incubations, being reduced by up to 95% in flooded compared to non-flooded replicates. Relatively large quantities of P, up to 78% of the concentration in non-flooded replicates, remained in the NaOH-extractable fraction in flooded replicates between 1- and 25-cm depth at the end of the incubation. The magnitude of the changes in NaOH-Pi was correlated with the depth profiles of MRP concentration at the end of the incubation.


View this table:
[in this window]
[in a new window]

  		Table 3. Changes to sediment P pools due to flooding with simulated river water. Differences are based on mean values from replicate cores (n = 3). Negative values indicate pool was larger in non-flooded compared to flooded replicates. Significance determined by two-factor ANOVA tests at p = 0.05 level. BD = below detection, Pi = inorganic P, Po = organic P.

 
Other sediment P fractions were not altered to the same extent as a result of flooding. The KCl-Pi, NaHCO3–Pi, and NaHCO3–Po data did indicate significant decreases in these fractions in the near-surface peat, but the changes in concentration were small and toward the detection limit of the method (1 mg kg–1). HCl-Pi, NaOH-Po, residual Po, and microbial biomass P did not change significantly between flooded and non-flooded replicates (p > 0.05). Absolute changes in the HCl-Pi, NaOH-Po, residual-Po, and microbial biomass P fractions as a result of flooding were not consistently positive or negative at all depths (Table 3), and were not correlated with concentrations of MRP in porewater.

Release of Solutes other than Molybdate Reactive Phosphorus
Changes in the concentration of dissolved species other than MRP were examined alongside changes in sediment P fractions to understand the controls on MRP release. Spatial and temporal patterns of MRP release were significantly and positively correlated with Fe2+ release in all cores during incubations using DW and SRW (r = 0.93, p < 0.01). The dominant P to Fe molar ratio in porewater samples during incubations using SRW was 0.45 (Fig. 6 ). The molar P to Fe ratio in porewater from incubations using DW decreased from 2.21 at 2.5-cm depth to 0.11 at 47.5-cm depth (data not presented). No significant correlations were observed between MRP and any other dissolved cations potentially responsible for the sorption of MRP in the drained cores (Ca2+, Mg2+, Al3+, Mn2+).


Figure 6
View larger version (16K):
[in this window]
[in a new window]

  		Fig. 6. Correlation between molar concentrations of molybdate reactive phosphorus (MRP) and ferrous iron (Fe2+) in all porewater samples from incubations using simulated river water.

 
Dissolved oxygen was below the detection limit in all porewater samples within 12 h of inundation, and increases in Mn2+ were observed throughout all cores between 12 and 24 h after inundation during DW and SRW incubations (data not presented). Concentrations of NO3––N in cores from incubations using DW did not exceed 0.3 mg L–1. However, in cores flooded with SRW, which each contained 10 mg L–1 NO3––N at the beginning of the incubation, the concentration of NO3– decreased over time, and Fe2+ (and MRP) were only detected in solution after concentrations of NO3– approached zero (Fig. 7 ). The time at which NO3– concentration approached zero varied with depth in the cores, occurring earlier at greater depth. Decreases in sulfate concentration were observed in all samples, beginning after but then occurring alongside increases in Fe2+ concentration (data not presented). We did not collect samples for dissolved gas analysis and so cannot report changes in these species.


Figure 7
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 7. Nitrate-nitrogen (NO3––N) and ferrous iron (Fe2+) concentration in surface water and porewater (10 and 32.5 cm depth) from incubations using simulated river water, normalized to concentration at time = 0 h. Mean concentrations from replicate cores are given (n = 3), error bars show ±1 standard deviation.

 

   Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
Conclusions
 REFERENCES
 
Release of Phosphorus and Potential for Changes in Phosphorus Concentration in Aerobic Receiving Waters
The maximum concentration of TP in sediment from Strumpshaw Fen (~1450 mg kg–1) is toward the upper end of concentrations reported for organic deposits that have been affected by nutrient loading from anthropic activities (e.g., Reddy et al., 1998; Mayer et al., 1999; Fisher and Reddy, 2001; Newman and Pietro, 2001; Soto-Jimenez et al., 2003). Depth profiles showing TP and TPo concentrations increasing toward the core surface are likely to reflect historical increases in P loads from both the River Yare and Lackford Run, associated with population growth alongside industrial and agricultural intensification within the catchment (Moss, 1979). Similar depth profiles for TPi may also reflect increased Pi loads from the river or adjacent agricultural land. Alternatively, lower water tables during the mid-1800s due to agricultural activity on the fen may have resulted in the mineralization of some Po in the near surface sediments, leading to an accumulation of Pi in this zone. In recent decades, groundwater and surface water levels in many lowland areas of Europe have decreased significantly (Lamers et al., 1998). Therefore, accumulations of potentially soluble Pi due to mineralization of Po in organic-rich sediments may be a relatively widespread phenomenon.

The data reported here show that under controlled laboratory conditions riparian wetland sediments may act as sources of P to both porewater and surface water. Depending on the composition of the flooding solution these releases can occur extremely rapidly. For example, porewater MRP concentrations above 1.0 mg L–1 P developed within 6 d of flooding during incubations using DW, and within 15 d in incubations using SRW. Maximum porewater MRP concentrations above 3.0 mg L–1 P were reached during incubations using both flooding solutions, and releases of MRP occurred concomitantly with increases in dissolved Fe2+ concentration. Based on samples collected from the field in a range of wetland types, Zak et al. (2004) suggest that on delivery to an aerobic environment, increased equilibrium concentrations of P would only be expected if P to Fe molar ratios in iron-rich anaerobic solutions are above approximately 0.33. If lower than 0.33, sorption of P onto precipitating Fe in the aerobic environment would be expected to remove all P from solution. The dominant P to Fe molar ratio in porewater during incubations using SRW (0.45) was greater than 0.33, which coupled with the high MRP concentrations, suggests that porewater from Strumpshaw Fen could lead to increased equilibrium concentrations of P if delivered to adjacent surface waters. Our findings emphasize the need to consider both the surface and subsurface as potential routes for MRP transport to receiving waters. Because MRP includes the fractions of total P that are directly available and rapidly consumed by bacteria and algae (Zhang and Oldham, 2001; Hens and Merckx, 2002), the transport of MRP from porewater to adjacent surface waters may drive changes in the biological status of these aquatic ecosystems. Additional information, including field validation of our laboratory evidence for MRP release, and quantification of the volumes of water exchanged between porewater and adjacent surface water, is required if these pathways for MRP transport are to be investigated further.

Factors Controlling Phosphorus Release during the Incubations
In incubations using both DW and SRW, MRP and Fe2+ were released concomitantly, and were significantly correlated in all porewater samples. Significant decreases in NaOH-extractable Pi were observed from all depths in the sediment as a result of flooding with SRW. Concentrations of dissolved cations other than Fe2+ were not significantly correlated with MRP, and sediment-bound P, other than that in the NaOH-extractable fraction, was largely stable during flooding. These findings support the classical model in which the reductive dissolution of Fe3+–P under anaerobic conditions leads to P and Fe2+ release to solution (Mortimer 1941, 1942; Patrick and Khalid, 1974; Lamers et al., 1998). Substantial quantities of NaOH-Pi remained in the sediment at 0 to 25 cm in flooded cores at the end of incubations using SRW, suggesting there was the potential for continued release of P from this fraction and the development of higher porewater and surface water concentrations of P.

Results from incubations using SRW showed that the existence of concentration gradients per se did not result in P release to surface water. The pH of samples during these incubations was always less than 7, meaning that P was most likely prevented from entering the surface water by sorption to positively charged surfaces on oxidized Fe-oxyhydroxides at the redox interface (Buffle, 1988; Buffle et al., 1989; Zak et al., 2004). Only after the reduction of Fe-oxyhydroxides at the porewater-surface water interface, shown by the release of Fe2+ to solution, did concentrations of MRP in the surface water rise. This suggests that P transport into the surface water was dominated by diffusion, rather than by gas ebullition or the action of invertebrates which are not expected to be dependent on redox conditions. Sorption of MRP at the oxidized porewater-surface water interface means that calculating fluxes of P into surface water using only concentration gradients and diffusion coefficients, as reported previously in some literature, may be open to error.

Increased sulfate loadings to aquatic ecosystems have been shown to promote P release to solution. Mineralization, and subsequent P release are enhanced by the presence of sulfate as an electron acceptor and by increased alkalinity as a result of sulfate reduction, and at high concentrations sulfate may compete with phosphate for anion adsorption sites. In addition, sulphide, produced through sulfate reduction, can reduce Fe3+–P and subsequently form insoluble iron sulphides, thereby decreasing the potential for binding of phosphate to reduced solid-phase Fe (Caraco et al., 1989; Lamers et al., 1998, 2001, 2002; Lucassen et al., 2004; Smolders et al., 2006). However, concentrations of MRP in porewater during incubations using DW (no added sulfate) were comparable to those during incubations using SRW (20 mg L–1 SO42–S), in contrast to what might have been expected if sulfate was promoting P release. In incubations using both DW and SRW, we also observed increases in MRP concentration in porewater before decreases in sulfate (presumably indicating sulphide production) began. Our findings suggest that reductive dissolution of Fe3+–P generated high porewater concentrations of MRP without the prior need for sulfate reduction, although later sulfate reduction may have contributed to MRP release and the maintenance of high porewater MRP concentrations. This suggests that re-adsorption of MRP by solid phase reduced Fe following initial Fe3+–P reduction was not sufficient to prevent concentrations of MRP in porewater from increasing substantially.

Previous work has shown that changes in soil moisture conditions, for example due to sediment inundation during our incubations, may alter the proportion of P stored in microbial biomass (Qui and McComb, 1995; Olila et al., 1997; Turner and Haygarth, 2001; Wright et al., 2001). Rewetting of a soil can lead to the direct release of P to solution from nucleic acids and phospholipids in cells, associated with the death of microbes due to osmotic shock and cell rupture (Webley and Jones, 1971; Salema et al., 1982; Turner et al., 2003; Nguyen and Marschner, 2005). However, the absence of significant decreases in microbial P in our incubations (Table 3) and the lack of substantial increases in MRP concentrations in porewater immediately following rewetting both suggest that cell lysis did not contribute significantly to MRP release. Because alternating flood-dry cycles occur in the field at Strumpshaw Fen, the microbial community may have developed resistance to changes in soil moisture conditions, for example, through the formation of resting stages during water table drawdown and sediment oxidation (Baldwin and Mitchell, 2000). Unlike some previous studies which have demonstrated P release from the microbial biomass, we maintained our columns intact during drying and did not expose the sediment to artificial oven or air drying. This may more closely simulate sediment drying under field conditions, and consequently more accurately reflect the potential for P release from the microbial biomass following rewetting. The possibility that bacterial utilization of polyphosphate under anaerobic conditions (Boström et al., 1988; Gächter et al., 1988; Gächter and Meyer, 1993; Wright et al., 2001) led to MRP release to porewater is also unsupported by the observation of stable microbial biomass P during our incubations.

Nitrate as a Redox Buffer Delaying Phosphorus Release
In comparison to incubations using DW, the timing of MRP release was delayed and varied with depth during incubations using SRW. The time at which the concentration of NO3– in solution was reduced to approximately zero also varied with depth in the cores during incubations using SRW, and the removal of NO3– from solution corresponded with the onset of Fe2+ and MRP release. Assuming that reductive dissolution of Fe3+–P led to the release of Fe2+ and MRP, potential mechanisms responsible for delayed Fe2+ and MRP release in the presence of NO3– include (i) the preferential use of NO3– compared with Fe3+ because it is an energetically more favorable electron acceptor in the microbial oxidation of organic carbon (Christensen et al., 2000); (ii) enzymatic re-oxidation of Fe2+ during NO3– reduction, resulting in the formation of oxidized Fe phases and re-adsorption of dissolved P (Straub et al., 1996; Weber et al., 2001, 2006); and (iii) outcompeting of Fe-reducing bacteria by NO3––reducing bacteria for a limiting resource, such as labile organic carbon. Further work would be required to indicate which of these possible mechanisms was responsible. It is also likely that diffusion from the surface water supported higher NO3– concentrations in shallow porewater, leading to the delayed release of Fe2+ and MRP compared to porewater at greater depth.

These findings extend those reported by Lucassen et al. (2004) who focused on groundwater-fed wetlands, and research in lacustrine systems (Andersen, 1982; Jensen and Andersen, 1992), and suggest that NO3– in surface-derived floodwater may also act as an important redox buffer in freshwater wetlands. Management of riparian wetlands in the UK has often involved isolation of the wetland from NO3––enriched river water. With reduced input of river water, and in systems such as Strumpshaw Fen which are disconnected from deeper groundwater (Surridge, 2004), rainwater with lower NO3– concentration becomes a more dominant water source. Our data suggest that this may promote more rapid releases of P compared to systems in which river water or groundwater, enriched in NO3–, remain important water sources.


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
REFERENCES
 
Riparian wetland sediments can contain large quantities of P in potentially soluble form, due to loads from adjacent nutrient-enriched channel water or because of mineralization of Po during sediment oxidation. Mesocosm incubations showed that P was rapidly released from these sediments to both surface water and porewater following flooding, and concentrations of MRP reached over 3 mg L–1 in porewater and over 0.8 mg L–1 P in surface water within 1 month of inundation. Releases of P were most likely controlled by the reductive dissolution of Fe3+–P under anaerobic conditions. The presence of NO3– in floodwater acted as an effective redox buffer, substantially delaying, although not preventing, the release of MRP and Fe2+. These findings confirm that riparian sediments can act as an internal P source to the wetland itself, even if external nutrient loadings are reduced by, for example, isolation of the riparian zone from adjacent rivers and streams. The sediments may also provide additional sources of P enrichment for adjacent surface waters because of the high concentrations of MRP released after flooding, and the high P to Fe molar ratios in anaerobic porewater. Therefore, it is possible that, rather than buffering adjacent aquatic ecosystems against nutrient delivery, riparian zones may become a source of nutrients under present-day conditions.

Our findings also have important implications for wetland restoration which often aims to restore the hydrological integrity of riparian zones, resulting in higher average water tables. In situations where a historical store of P has built up in the riparian sediments, this rewetting may release substantial quantities of P to solution. Further, it is common for the water quality of adjacent rivers and streams to be deemed to be too poor to be used in riparian wetland restoration, for example because of high NO3– concentration. Under these circumstances, higher water tables may be maintained using water sources which are low in NO3–, such as groundwater or rainwater. Our findings have shown that this may exacerbate the release of P to solution, and this in turn may have negative chemical and biological consequences for both the restored wetland and for adjacent aquatic ecosystems.


   ACKNOWLEDGMENTS
 
We are grateful to the Royal Society for the Protection of Birds and their reserve warden, Tim Strudwick, for allowing us to use Strumpshaw Fen as a research site, for background data, and for assistance during fieldwork. We also wish to thank the Environment Agency for providing data for use in this research. The research was supported by a studentship (to BWJS) from the University of Sheffield.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
C. C. Hoffmann, C. Kjaergaard, J. Uusi-Kamppa, H. C. B. Hansen, and B. Kronvang
Phosphorus Retention in Riparian Buffers: Review of Their Efficiency
J. Environ. Qual., August 24, 2009; 38(5): 1942 - 1955.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
E. M. Berryman, R. T. Venterea, J. M. Baker, P. R. Bloom, and B. Elf
Phosphorus and Greenhouse Gas Dynamics in a Drained Calcareous Wetland Soil in Minnesota
J. Environ. Qual., August 24, 2009; 38(5): 2147 - 2158.
[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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Surridge, B. W. J.
Right arrow Articles by Baird, A. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Surridge, B. W. J.
Right arrow Articles by Baird, A. J.
Agricola
Right arrow Articles by Surridge, B. W. J.
Right arrow Articles by Baird, A. J.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Phosphorus
Right arrow Nutrient Cycling

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