Phytate as a novel phosphorus-specific paleo-indicator in aquatic sediments

(1)

DOI 10.1007/s 10933-008-9283-6 ORIGINAL PAPER

Phytate as a novel phosphorus-specific paleo-indicator in aquatic sediments

Benjamin L. Turner • Kaarina Weckstrom

Received: 26 February 2008/Accepted: 17 November 2008/Published online: 16 December 2008

Abstract A reliable geochemical paleo-indicator for phosphorus remains elusive, despite the importance of understanding historical changes in the nutrient status of aquatic ecosystems. We assessed the potential of phytate (salts of mvo-inositol hexakisphosphate) as a novel phosphorus-specific paleo- indicator by measuring its concentrations in dated sediments from an embayment in Helsinki, Finland, with a known 200-year history of trophic changes.

Phytate was extracted in a solution containing sodium hydroxide and EDTA and detected by solution ^:P NMR spectroscopy with spectral deconvolution.

Concentrations varied markedly with sediment depth and paralleled previously determined changes in diatom assemblages and geochemical indicators linked to trophic status. In contrast, total sediment phosphorus did not reflect phosphorus inputs to the embayment, presumably due to the mobilization of inorganic phosphate under anoxic conditions during

B. L. Turner (El)

Smithsonian Tropical Research Institute, Apartado 0843-03092, Balboa, Ancon, Republic of Panama e-mail: [email protected]

K. Weckstrom

Department of Quaternary Geology, Geological Survey of Denmark and Greenland, 0. Voldgade 10, 1350 Copenhagen, Denmark

K. Weckstrom

Institute for Limnology, Austrian Academy of Sciences, Mondseestrasse 9, 5310 Mondsee, Austria

periods of high pollutant loading. Importantly, phytate appeared to be stable in these brackish sediments, in contrast to other organic and inorganic phosphates which declined abruptly with depth. We therefore conclude that phytate represents a potentially impor- tant indicator of historical changes in phosphorus inputs to water bodies, although additional studies are required to confirm its stability under conditions likely to be encountered in lakes and coastal ecosystems.

Keywords Inositol phosphate • Paleo-indicator Phosphorus • Phytate • Sediment

NMR spectroscopy

Solution 31P

Introduction

A geochemical indicator of phosphorus inputs to sediments remains elusive, despite the importance of understanding historical changes in the trophic status of aquatic ecosystems (Anderson and Rippey 1994;

Engstrom and Wright 1985; Vaalgamaa 2004).

Phosphorus in lake and coastal sediments changes relatively rapidly on paleoecological timescales, so total phosphorus is not generally considered to be a reliable indicator of the phosphorus status of a particular site at the time the sediment was deposited.

For example, anoxia in buried sediments changes the solubility of inorganic phosphate associated with iron

(2)

oxides, a major fraction of the total phosphorus in many sediments, which leads to transfer of phosphate away from the site of deposition (e.g. Anderson and Rippey 1994). Similar problems occur in marine systems, in which a significant proportion of the sedimentary phosphorus is re-mobilized during sul- fate reduction (Jensen et al. 1995). The usefulness of 'biogenic' phosphorus compounds, such as deoxyribonucleic acid (DNA) or polyphosphate, as paleo- indicators is also limited, due to their relatively short turnover times in lake sediments (Ahlgren et al. 2005;

Watts et al. 2002).

In the absence of a phosphorus-specific paleo- indicator, historical rates of phosphorus loading are usually inferred from changes in diatom assemblages and quantitative diatom—total phosphorus transfer functions (Anderson 1997; Bennion et al. 2004;

Bradshaw and Anderson 2001; Hall et al. 1997).

However, diatom-based inference models rely on calibration against large modern data sets of surface- sediment diatoms and associated water chemistry, which typically involve >50 sites and are therefore time consuming and laborious to collect. A number of such datasets are available publicly for down-core phosphorus reconstructions (e.g. EDDI 2001), but should only be used for comparable sites in the same geographical region.

Here we report the potential suitability of phytate (salts of wryo-inositol hexakisphosphate) as a phosphorus-specific tracer in aquatic sediments. This compound, also known as phytic acid (the free-acid form), occurs widely in the environment (Turner et al. 2002). It is abundant in plant seeds and constitutes the main form of organic phosphorus in many soils and sediments, where it accumulates due to its strong interaction with clays and other abiotic soil components (for recent reviews see Turner et al.

2007). There is less information for aquatic ecosystems, although phytate appears to occur widely in sediments (McKelvie 2007). For example, Suzumura and Kamatani (1995b) reported that total inositol hexakisphosphate (i.e. phytate plus three other stereoisomeric forms) constituted ~ 10% of the organic phosphorus in river sediments, while Sommers et al.

(1972) found similar values for inositol penta- and hexa-kisphosphate in the sediments of four freshwater lakes.

The abundance and stability of phytate in the environment suggests that it could be a useful

indicator of phosphorus deposition in sediments.

We used a novel solution ³¹P NMR spectroscopy procedure with spectral deconvolution to quantify historical changes in phytate concentrations in a well- studied sediment core from an embayment in Hel- sinki, Finland, that has undergone marked changes in phosphorus loading in the recent past. By comparison with diatom assemblages and known changes in nutrient loading to the embayment, our aim was to determine the suitability of phytate as a paleo- indicator for phosphorus in sediments.

Study site and trophic history

Toolonlahti is a small (surface area 0.2 km ), shallow (mean depth ca. 2 m) and turbid (transparency

< 1 m) embayment located in the centre of Helsinki.

Although similar in many respects to a lake, the embayment is connected to the open sea through a narrow system of straits formed following post-glacial isostatic uplift and construction work since the nineteenth century. This has restricted water exchange and increased the residence time, but the salinity (4.8%o) remains similar to the outside archipelago.

Wastewater input to the embayment increased markedly following extensive urbanization of the catchment in the late nineteenth and early twentieth century. Point-source loading began in 1823 following the construction of a sugar mill, but increased markedly in the late nineteenth century due to untreated municipal wastewater from the first enclosed sewage pipe completed in 1878. The sugar mill closed in 1965, but its wastewater had been previously diverted elsewhere. Two wastewater treatment plants were installed in the catchment in 1910 and 1915 to serve the rapidly expanding population of Helsinki (which increased from 120,000 to 350,000 during the first half of the twentieth century) until their closure in 1935 and 1959, respectively (Tikka- nen et al. 1997). The plants removed organic matter from sewage to reduce its biological oxygen demand, but did not remove dissolved nutrients.

The trophic history of the embayment has been studied in detail (Korhola and Blom 1996; Tikkanen et al. 1997; Weckstrom 2006). There have been marked shifts in diatom life-forms, which reflect closely the known history of nutrient loading (Fig. 1).

Increased turbidity associated with eutrophication

(3)

Planktonic diatoms (%) Number of diatom species

2000 2000

Fig. 1 Changes in the abundance of planktonic diatoms and the number of diatom species in sediment from Toolonlahti, Helsinki, Finland, indicating the trend in the trophic status of the embayment (data from Weeks trom 2006)

favours planktonic diatoms over benthic species (Andren et al. 1999; Bennion et al. 2004; Cooper and Brush 1991). At Toolonlahti the percentage abundance of planktonic diatoms increased with sewage loading from the late nineteenth century, peaking in 1956 when such organisms represented up to 90% of all taxa (Fig. 1). A concurrent decline in species richness to

~20 species was observed during this time (Fig. 1), consistent with commonly-observed declines in algal diversity following eutrophication (Wetzel 2001).

Clear signs of recovery were evident in the diatom data following the closure of the second treatment plant in 1959; the abundance of planktonic diatoms decreased to around 55% and species richness increased to around 50 species in the early 1980s (Fig. 1). However, Toolonlahti remains eutrophic, with an annual mean total phosphorus concentration of ~70 ug P 1~ and a total dissolved nitrogen concentration of ~600 (j.g N 1_1 (Weckstrom 2006).

The changes in diatoms are supported by geochemical indicators such as the concentrations of cadmium, copper, zinc, and other elements, which increased markedly during the first half of the twentieth century and declined thereafter (Tikkanen et al. 1997). High concentrations of cadmium and zinc are typical for urban wastewaters (Blomqvist et al. 1992), confirming that the changes in trophic status in Toolonlahti determined from diatom

assemblages largely reflect the impact of nutrient loading from sewage. In addition, the organic matter content of pre-1800 sediment from a deeper core, taken close to the core studied here, clearly indicated that most disturbance at the site occurred after 1800 (Tikkanen et al. 1997). We assume therefore that the earliest sediment samples analyzed here are repre- sentative of relatively unimpacted conditions.

Materials and methods

The core was sampled with a mini-Mackerefh corer (Mackereth 1969) in October 2003 from the deepest area of the embayment and sectioned into 1-cm intervals. A single core was considered representa- tive, given the small size and simple bathymetry of the embayment, as well as previous studies on multiple cores (Anderson 1998). The main characteristics of the obtained sediment stratigraphy were described and sub-samples stored at 4°C in small plastic bags. The chronology of the core was determined by correlation with a core from a previous study (dated by Pb, 7Cs and spheroidal carbona- ceous particles) using the distinctive changes in loss on ignition through the profile (Tikkanen et al. 1997).

Sediment (1.50 ± 0.01 g) was first treated with 30 ml of 1 M HC1 by shaking horizontally for 1 h at 22°C to remove cations that can interfere with subsequent extraction of organic phosphorus (Turner et al. 2005). After centrifugation (3,000g, 15 min) the supernatant was decanted and retained for determination of phosphate by automated molybdate colorimetry. The sediment was washed by shaking for 10 min with deionized water, centrifuged again, and the supernatant discarded. The sediment was then extracted by shaking with 30 ml of a solution containing 0.25 M NaOH and 50 mM Na2EDTA (ethylenediaminetetraacetate) for 16 h at 22°C (Cade-Menun and Preston 1996). The procedure is assumed to quantitatively recover organic phosphorus, including phytate, from soil (Turner et al. 2005) and we assumed this would also be the case for aquatic sediment. The technique has been used widely in recent years to extract organic phosphorus from a range of freshwater, brackish, and marine sediments (Ahlgren et al. 2006; Cade-Menun et al.

2005; Carman et al. 2000; Hupfer et al. 1995, 2004;

Turner and Newman 2005; Turner et al. 2006).

(4)

The NaOH-EDTA extracts were centrifuged (8,000g, 30 min) and a 20 ml aliquot was spiked with 1 ml of a solution containing 50 u.g P ml^-1 of methylene diphosphonic acid as an internal standard (chemical shift 17.61 ± 0.03 ppm). The spiked extracts were frozen at —35°C, lyophilized (~24 h), and homogenized by gently crushing to a fine powder.

For solution P NMR spectroscopy, each lyophilized extract (~100mg) was re-dissolved in 0.1 ml of deuterium oxide and 0.9 ml of a solution containing 1.0 M NaOH and 0.1 M Na2EDTA, and then trans- ferred to a 5-mm NMR tube. Solution ³¹P NMR spectra were obtained using a Bruker Avance DRX 500 MHz spectrometer operating at 202.456 MHz for P. Sam- ples were analyzed using a 6 u.s pulse (45°), a delay time of 2.0 s, an acquisition time of 0.4 s, and broadband proton decoupling. Approximately 30,000 scans were acquired for each sample and chemical shifts of signals were determined in parts per million (ppm) relative to an external standard of 85% H3PO4.

Spectra were plotted using 5 Hz line broadening and signals were assigned to phosphorus compounds based on literature reports (Turner et al. 2003a). Signal areas were calculated by integration and concentrations of phosphorus compounds were calculated from the integral value of the internal standard. Concentrations of phytate-phosphorus were determined by spectral deconvolution of the phosphate monoester region and quantification of the four signals from phytate (Turner et al. 2003b).

Total carbon and nitrogen were determined by combustion and gas chromatography using a Flash NCI 112 Soil Analyzer (CE Elantech, Lakewood, NJ). Total phosphorus was determined by ignition (550°C x 1 h) and extraction in 1 M H2S04 (1:50 soil/solution ratio, 16 h), with phosphate detection by automated molybdate colorimetry. Total phosphorus extracted in NaOH-EDTA was determined by induc- tively-coupled plasma optical-emission spectrometry (Optima 2100; Perkin-Elmer Inc., Shelton, CT, USA) on neutralized extracts following a 20-fold dilution.

Results

Total element concentrations

Total carbon increased markedly after the late 1800s and then again after the mid-1900s, reaching a maximum of 10.3% in sediment dated to 1958—one year before the closure of the second sewage plant (Fig. 2). Total nitrogen followed a similar pattern, reaching a maximum concentration of 0.76% in the 1958 sample (Fig. 2). There were corresponding changes in the C/N ratio through the core, from a minimum of 8.5 in sediment dated to 1851 to a maximum of 13.6 in sediment dated to 1958 (data not shown). Concentrations of carbon and nitrogen in the most recently deposited sediment (0-1 cm depth) were 6.7 and 0.60%, respectively, with a C/N ratio of 11.2.

Fig. 2 Changes in total carbon, nitrogen, and phosphorus in a sediment core from Toolonlahti, Helsinki, Finland

Carbon (%)

J 1 1 1 L

2000

1950

1900

1850

1800

Nitrogen (%)

j 1 i_

Phosphorus (%)

1 1 i_

- 2000

1950

1900

(5)

Total phosphorus showed a different pattern, with a maximum concentration of 0.33% in sediment dated to 1887 (Fig. 2). Concentrations then declined steadily in younger sediments, with the most recent sediment containing 0.15% phosphorus, a value similar to the pre-1887 sediment.

Identification of signals in solution P NMR spectroscopy

Broad groups of signals were identified as follows (Fig. 3, top spectrum). Phosphate was identified as the strong signal at 6.43 ± 0.08 ppm (mean ± standard deviation of the chemical shift in all spectra).

C

rn—i—i—i—i—I—i—i—i—1—|—i—i—i—i—I—i—i—i—1—|—i—i—i—i—I—i—i—i—i—|—i—i—

20 10 0 -10

Phosphate monoesters were identified as the complex group of signals between 3.5 and 6.0 ppm. Two phosphate diesters were identified: phospholipids occurred as broad and weak signals between 0.5 and 2.0 ppm, while DNA was identified as the broad signal at —0.16 ± 0.05 ppm. Pyrophosphate, an inorganic polyphosphate with a chain length of two, was identified at —3.96 ± 0.08 ppm, although long- chain polyphosphate was not detected in any sample.

Phosphonate, which occurs at ~ 20 ppm, was detected in only the 1951 sample (~2 mg P kg-1).

The phosphate monoester region was examined in detail to quantify individual compounds (Fig. 3, bottom spectrum). Phytate was identified from its characteristic series of four signals in a 1:2:2:1 pattern (Turner et al. 2003b), which occurred in these samples at 6.04 ± 0.04 ppm (C-2 phosphate), 5.15 ± 0.05 ppm (C-l and C-3 phosphates), 4.77 ± 0.05 ppm (C-4 and C-6 phosphates) and 4.64 ± 0.04 ppm (C-5 phosphate). Other signals were assigned to the degradation products of phospholipids (5.36 ± 0.05 ppm and 5.02 ± 0.05 ppm), which are formed during extraction and analysis in strong alkali (Turner et al.

2003a). Two clear but unidentified signals were detected in all spectra at 2.96 ± 0.06 ppm and 6.91 ± 0.06.

^4^\fW*vVV1uw

u jr* u

^XuAvvVrt/ VWv-iwJ'

7 6 5 4 3 Chemical shift (ppm)

Fig. 3 Solution 31P NMR spectra of a NaOH-EDTA extract of sediment from a core dated to 1958 (30-31 cm depth) from Toolonlahti, Helsinki, Finland. The top spectrum shows the broad groups of signals: A methylene diphosphonic acid (internal standard for signal quantification); B phosphate; C phosphate monoesters; D deoxyribonucleic acid; E pyrophosphate. Not indicated are the phospholipids, which occur between 0.5 and 2.0 ppm. The bottom spectrum shows the expanded phosphate monoester region, with signals assigned as follows: a unidentified inositol phosphate; b phosphate; c the four signals from phytate, occurring in a 1:2:2:1 pattern; d degradation products of phospholipids in alkali; e unidentified organic phosphate

Phosphorus concentrations in NaOH-EDTA extracts

The NaOH-EDTA extracts contained between 356 and 596 mg P kg-1 of total phosphorus, which represented between 14 and 34% of the total sediment phosphorus (data not shown). Organic phosphorus (the sum of phosphate monoesters, phosphate diesters, and phosphonates) ranged between 139 mg P kg-1 in the deepest sediment layer dated to 1817 and 317 mg P kg-1 in the sample dated to 1958 (Table 1). These values represented between 5 and 17% of the total sediment phosphorus (Table 1). The concentration in the youngest sediment was 220 mg P kg-1, which constituted 15% of the total sediment phosphorus.

Phytate concentrations were low (<20 mg P kg- ) in the youngest (post-1980) and oldest (pre-1900) sediments (Fig. 4), including the sample containing the greatest total sediment phosphorus concentration dated to 1887 (Fig. 2). After 1887, phytate concentrations increased markedly to a maximum of 77.1 mg P kg^-1 in the sample dated to 1958

(6)

Table 1 Organic phosphorus compounds extracted in NaOH-EDTA from a sediment core from Toolonlahti, Helsinki, Finland Depth (cm) Date Organic Pa Phytate DNA Phospholipids Unidentified (% organic P) Other monoesters

(mg P kg ) (% organic P) (% organic P) (% organic P)

A (2.96 ppm) B (6.91 ppm) (% organic P) 0-1 2003 220(15) 6.7

5-6 1992 252(14) 8.3 10-11 1983 223(15) 8.1 19-20 1972 195(12) 11.8 30-31 1958 317 (17) 24.3 35-36 1951 241(13) 21.6 39-40 1943 246(13) 22.1 50-51 1914 238(10) 18.3 59-60 1887 161 (5) 10.3 70-71 1851 152(9) 7.0 79-80 1817 139(8) 9.6

23.1 20.4 20.0 15.3 10.2 12.2 8.9 14.0 9.4 10.0 7.5

4.3 6.1 7.0 7.2 6.0 6.4 8.2 7.4 11.4 9.6 12.0

2.0 2.6 2.4 4.0 4.6 4.4 4.6 5.7 4.3 4.7 4.1

1.0 1.2 0.4 0.8 2.3 1.5 1.5 0.8 3.0 0.3 1.7

628 61.4 62.1 61.0 525 53.1 54.8 53.9 61.7 68.4 65.2 Values in parentheses are organic phosphorus as a proportion (%) of the total sediment phosphorus

Phytate DNA Phospholipids Pyrophosphate

2000

1950

1900

1850

1800

0 25 50 75

—i—

50 25

2000

1950

1900

1850

1800 50

Fig. 4 Changes in the concentrations (mg P kg Helsinki, Finland

25 50 0 10 20 0 Phosphorus (mg P kg"1 dry wt)

dry weight) of phosphorus compounds in a sediment core from Toolonlahti,

(Fig. 4). In this sample, phytate constituted 24.3% of the organic phosphorus, while in older (pre-1887) and younger (post-1980) sediments it constituted <10%

(Table 1). Other phosphate monoesters (i.e. exclud- ing phytate) constituted between 93 and 174 mg P kg^-1, being greatest in the sample dated to 1958 (Table 1).

DNA concentrations declined steadily with depth from the sediment surface, with only a small increase in the sample dated to 1914 (Fig. 4). The maximum concentration was 51 mg P kg-1 in surface sediment, declining to ~ 10 mg P kg-1 in the oldest sediment.

Expressed as a proportion of the extracted organic phosphorus, DNA declined steadily from 23.1% in

(7)

the surface sediment to ~ 7.5% in the oldest sediment (Table 1).

Phospholipids, in contrast, increased from <10 mg P kg-1 in surface sediment to ~20 mg P kg-1 in sediment dated to 1943 (Fig. 4). Concentrations then remained relatively constant through the rest of the core. As a proportion of the extracted organic phosphorus, phospholipids increased from <5% in surface sediment to 12% in the oldest sediment (Table 1).

Pyrophosphate concentrations were greatest in sediment deposited since 1972 (Fig. 4). The maximum concentration was 53 mg P kg-1 in the youngest sediment, declining sharply to only 7 mg P kg^- in the sample dated to 1958 and remaining low throughout the deeper part of the core.

The two unidentified signals at 2.96 and 6.91 ppm followed a similar trend to phytate and total carbon, occurring at maximum concentrations in sediment dated to 1958 (Table 1). However, the 6.91 ppm signal also showed a clear peak in the sample dated to 1887, coincident with the maximum concentration of total sediment phosphorus. The 2.96 ppm signal occurred at up to 14.6 mg P kg- and represented between 4.4 and 5.7% of the organic phosphorus, while the 6.91 ppm signal occurred at up to 7.0 mg P kg- and represented up to 3.0% of the organic phosphorus (Table 1). In the most recent sediment these compounds were present at about 4 and 2 mg P kg-1, respectively.

Discussion

Phytate as a paleo-indicator

Total sediment phosphorus is of limited use as an indicator of past trophic status due to the mobility of inorganic phosphate in anoxic bottom waters and sediments (Vaalgamaa 2004) and the rapid degradation and recycling of most 'biogenic' phosphorus compounds (Ahlgren et al. 2005). This was confirmed here, because total phosphorus concentrations did not correspond to the known history of nutrient loading to the embayment and therefore provided a poor record of trophic status. In particular, the greatest total phosphorus concentration occurred in sediment dated to 1887, yet phosphorus loading did not peak until the mid-twentieth century. A possible explanation is that

anoxic conditions, expected to have developed in bottom waters during high nutrient loading in the early twentieth century, released inorganic phosphate from complexation with iron compounds in the sediment, as shown in previous studies (e.g. Ander- son and Rippey 1994). This highlights the importance of identifying alternative measures by which to reconstruct historical changes in phosphorus status.

Phytate appears to offer one such measure, because it was preserved to depth in sediments at detectable concentrations that reflected closely the known changes in the trophic status of the ecosystem. In particular, phytate reflected trends in diatom assemblages (Fig. 1) and agreed well with the history of nutrient loading from urban wastewater (Tikkanen et al. 1997). Although phytate concentrations did not correspond to the trend in total sediment phosphorus, they did match changes in other sediment characteristics indicative of trophic status, including total carbon, loss on ignition, and heavy metal concentrations.

Phytate stability in sediments

Importantly, phytate did not appear to be influenced by anoxic conditions in the same way as total sediment phosphorus, being preserved at depth in the brackish and anoxic sediments of Toolonlahti for at least 200 years. This confirms the apparent refrac- tory nature of phytate in the environment (McKelvie 2007; Turner et al. 2002). Reduction of wryo-inositol hexakisphosphate sorbed onto iron oxides appears to form an insoluble Fe4-phytate that resists enzymatic hydrolysis (De Groot and Golterman 1993). This is not consistent with the apparent acceleration of phytate decomposition by anoxic conditions in marine sediments (Suzumura and Kamatani 1995a), but this might be due to the effect of salinity on phytate stability. Suzumura and Kamatani (1995b) reported that concentrations of phytate and a series of other stereoisomeric forms of inositol hexakisphosphate declined in sediments along a freshwater- marine (i.e. salinity) continuum, being relatively abundant in river sediments, but declining markedly in estuarine and coastal marine sediments, where they occurred in only trace concentrations and were restricted to the sediment surface. The significance of the change in salinity in destabilizing phytate was demonstrated in subsequent studies by Gardolinski

(8)

et al. (2004), in which a rapid release of organic phosphorus, expected to include phytate, was observed from freshwater sediment when salinity was increased to >10%o. Organic phosphorus was not released at salinity values <10%o, suggesting that the salinity in Toolonlahti (4.8%o) is too low to destabi- lize phytate. Further experiments are clearly required to determine the conditions under which phytate remains stable in aquatic sediments.

macrophytes have been reported to contain mainly the lower-order inositol phosphates (Weimer and Armstrong 1979). Phytate is abundant in some soils (Turner et al. 2002), but this cannot have made a major contribution to the concentrations reported here as there is no agricultural land in the catchment serving the embayment.

Behavior of other phosphate compounds Sources of phytate in the embayment

Given the history of nutrient loading from urban wastewater and the absence of agriculture in the catchment, it seems almost certain that urban wastewater was the source of phytate in the embayment, particularly the high concentrations in the first half of the twentieth century. Phytate constitutes the majority of the total phosphorus in seed grains and is therefore abundant in human diets (Harland and Oberleas 1987).

However, humans do not effectively digest phytate, so it can constitute a considerable fraction of the total phosphorus in human feces and sewage. For example, a recent study reported fecal phytate concentrations of 4—11 g P kg^-1 dry weight, which constituted between 24 and 54% of the total fecal phosphorus (Joung et al.

2007). The values were dependent on age and diet, although phytate remained abundant in feces even from low-phytate diets. Of the relatively large number of studies that have determined the phosphorus composition of sewage samples, only Cosgrove (1973) specifically measured phytate, although aerobically- digested sewage is rich in phosphate monoesters, which almost certainly include some phytate (e.g.

Hinedi et al. 1989; Smith et al. 2006).

Additional sources of phytate from within the embayment, expected to contribute to 'background' concentrations, could include seeds, pollen, macro- phyte tissue and phytoplankton. Phytoplankton become more abundant during eutrophic conditions, while the abundance of macrophytes tends to increase during the initial stages of nutrient enrichment, but then decline during eutrophic conditions due to increased turbidity and light limitation (Korhola and Blom 1996). However, in a study on Tokyo Bay, neither zooplankton nor phytoplankton (including a cultured sample of the diatom Skeletonema costaturri) contained a detectable concentration of phytate (Suzumura and Kamatani 1995b), while aquatic

Other phosphorus compounds detected in the sediment core, specifically DNA and pyrophosphate, declined with depth in a manner indicative of rapid biological turnover. This is likely to contribute to the maintenance of eutrophic conditions in the embayment and suggests that these compounds are unsuitable as paleo-indicators of phosphorus status.

This was expected for pyrophosphate, which decom- poses more rapidly than organic phosphates during diagenesis (Ahlgren et al. 2006; Hupfer et al. 1995, 2004), although we did not detect long-chain polyphosphate in any sample, despite its abundance in many freshwater lake sediments (Hupfer et al. 1995, 2004). Concentrations of DNA declined markedly in the upper sediment layers, but persisted in deeper layers to a much greater extent than pyrophosphate.

This indicates either partial stabilization or presence in living microbes and may explain why DNA has been reported to be both stable (Carman et al. 2000) and unstable (Ahlgren et al. 2006) in anoxic Baltic Sea sediments.

Phospholipids increased as a proportion of the organic phosphorus throughout the core, which is surprising given their apparent instability in the environment (Ahlgren et al. 2006; Watts et al.

2002). However, the measurement of phospholipids by alkaline extraction and solution ³¹P NMR spectroscopy is confounded by the sensitivity of some phospholipids to chemical hydrolysis in strong alkali (Makarov et al. 2002; Turner et al. 2003a), so values determined by this procedure should be regarded as underestimates.

The unidentified signal at 6.91 ppm is detected commonly in soil extracts (e.g. Turner et al. 2003c), but to our knowledge has been reported only once in aquatic sediment—an anoxic sediment from waters off the coast of Finland (Carman et al. 2000). The signal persists following hypobromite oxidation (Turner and Richardson 2004), so probably represents

(9)

an inositol phosphate. The unidentified signal at 2.96 ppm that occurred in all samples was also evident in spectra of anoxic Baltic Sea sediments (Ahlgren et al. 2006; Carman et al. 2000). It was previously assigned to teichoic acids (Ahlgren et al.

2006), although these compounds occur around 2.3 ppm in alkaline soil extracts (Makarov et al.

2002), which is ~0.7 ppm from the signal in our spectra. Glucose-1-phosphate occurs close to 3 ppm in alkaline extracts (Turner et al. 2003a), but would be expected to degrade rapidly in the environment.

As the 2.96 ppm signal detected here followed an almost identical pattern to phytate in the sediment profile, it may represent an inositol phosphate.

Potential use of inositol phosphate stereoisomers as paleo-indicators

Phytate (i.e. salts of the myo form of inositol hexakisphosphate) is the only stereoisomeric form of inositol phosphate found in grain seeds, raw sewage, and animal manures. In contrast, other stereoisomeric forms of inositol hexakisphosphate (i.e. scyllo, neo, D-chiro) occur only in significant quantities in soil (Turner et al. 2002). For example, scvWo-inositol hexakisphosphate is common in soils, but was not detected in sediments of the current study, confirming the negligible input of soil-derived sediment to the embayment. The ratio of wryo-inositol hexakisphosphate (phytate) to other stereoisomeric forms in sediment could therefore be used as a sensitive indicator of historical changes in pollution from sewage and/or animal manure. The myo and scyllo forms of inositol hexakisphosphate can be determined using the methodology described here, which can be refined for low concentrations of inositol phosphates by including a hypobromite oxidation step (Turner and Richardson 2004).

Determination of neo- and D-c/z/ro-inositol hexakisphosphate would require more sensitive methodology, such as the combination of ion exchange and gas chromatography used to detect inositol stereoisomers in Tokyo Bay sediments (Suzumura and Kamatani 1993).

Acknowledgments We thank Alex Blumenfeld (University of Idaho) and Paul Leeson (University of Helsinki) for analytical support, and S. Vaalgamaa and A. Korhola (University of Helsinki) for assistance in the field.

References

Ahlgren J, Tranvik L, Gogoll A, Waldeback M, Markides K, Rydin E (2005) Sediment depth attenuation of biogenic phosphorus compounds measured by P NMR. Environ Sci Technol 39:867-872. doi:10.1021/es049590h Ahlgren J, Reitzel K, Tranvik L, Gogoll A, Rydin E (2006)

Degradation of organic phosphorus compounds in anoxic Baltic Sea sediments: a P nuclear magnetic resonance study. Limnol Oceanogr 51:2341-2348

Anderson NJ (1997) Historical changes in epilimnetic phosphorus concentrations in six rural lakes in Northern Ireland. Freshw Biol 38:427-440. doi:10.1046/j.l365- 2427.1997.00249.x

Anderson NJ (1998) Variability of diatom-inferred phosphorus profiles in a small lake basin and its implications for histories of lake eutrophication. J Paleolimnol 20:47-55.

doi:10.1023/A:1007923813511

Anderson NJ, Rippey B (1994) Monitoring lake recovery from point-source eutrophication: the use of diatom-inferred epilimnetic total phosphorus and sediment chemistry.

Freshw Biol 32:625-639. doi: 10.11 ll/j.1365-2427.

1994.tb01153.x

Andren E, Shimmield G, Brand T (1999) Environmental changes of the last three centuries indicated by siliceous microfossil records from the southwestern Baltic Sea.

Holocene 9:25-38. doi:10.1191/095968399676523977 Bennion H, Fluin J, Simpson GL (2004) Assessing eutrophi-

cation and reference conditions for Scottish freshwater lochs using subfossil diatoms. J Appl Ecol 41:124-138.

doi: 10.1111/j. 1365-2664.2004.00874.X

Blomqvist S, Larsson U, Borg H (1992) Heavy metal decrease in the sediments of a Baltic Bay following tertiary sewage treatment. Mar Pollut Bull 24:258-266. doi:10.1016/

0025-326X(92)90564-M

Bradshaw EG, Anderson NJ (2001) Validation of a diatom- phosphorus calibration set for Sweden. Freshw Biol 46:1035-1048. doi:10.1046/j.l365-2427.2001.00732.x Cade-Menun BJ, Preston CM (1996) A comparison of soil

extraction procedures for P NMR spectroscopy. Soil Sci 161:770-785. doi: 10.1097/00010694-199611000-00006 Cade-Menun BJ, Benitez-Nelson CR, Pellechia P, Paytan A

(2005) Refining P nuclear magnetic resonance spectroscopy for marine particulate samples: storage conditions and extraction recovery. Mar Chem 97:293- 306. doi:10.1016/j.marchem.2005.05.005

Carman R, Edlund G, Damberg C (2000) Distribution of organic and inorganic phosphorus compounds in marine and lacustrine sediments: a P NMR study. Chem Geol 163:101-114. doi:10.1016/S0009-2541(99)00098-4 Cooper SR, Brush OS (1991) Long-term history of Chesapeake

Bay anoxia. Science 254:992-996. doi: 10.1126/science.

254.5034.992

Cosgrove DJ (1973) Inositol polyphosphates in activated sludge. J Environ Qual 2:483-485

De Groot CJ, Golterman HL (1993) On the presence of organic phosphate in some Camargue sediments: evidence for the importance of phytate. Hydrobiologia 252:117-126 EDDI (2001) European Diatom Database. Newcastle Univer-

sity, UK, http://craticula.ncl.ac.uk/Eddi/jsp/datasets.jsp

(10)

Engstrom DR, Wright HE Jr (1985) Chemical stratigraphy of lake sediments as a record of environmental change.

In: Haworth EY (ed) Lake sediments and environmental history. University of Leicester Press, Leicester, pp 11-67

Gardolinski PCFC, Worsfold PJ, McKelvie ID (2004) Seawater induced release and transformation of organic and inorganic phosphorus from river sediments. Water Res 38:688-692. doi:10.1016/j.watres.2003.10.048

Hall RI, Leavitt PR, Smol JP, Zirnhelt N (1997) Comparison of diatoms fossil pigments and historical records as measures of lake eutrophication. Freshw Biol 38:401^-17. doi:

10.1046/j.l365-2427.1997.00251.x

Harland BE, Oberleas D (1987) Phytate in foods. World Rev Nutr Diet 52:235-259

Hinedi ZR, Chang AC, Lee RWK (1989) Characterization of phosphorus in sludge extracts using phosphorus-31 nuclear magnetic resonance spectroscopy. J Environ Qual 18:323-329

Hupfer M, Gachter R, Riiegger H (1995) Polyphosphate in lake sediments: 31P NMR spectroscopy as a tool for its identification. Limnol Oceanogr 40:610-617

Hupfer M, Rube B, Schmieder P (2004) Origin and diagenesis of polyphosphate in lake sediments: a P-NMR study.

Limnol Oceanogr 49:1-10

Jensen HS, Mortensen PB, Andersen F0, Rasmussen E, Jensen A (1995) Phosphorus cycling in a coastal marine sediment, Aarhus Bay, Denmark. Limnol Oceanogr 5:908-917 Joung H, Juen BY, Li SJ, Kim J, Woodhouse LR, King JC et al

(2007) Fecal phytate excretion varies with dietary phytate and age in women. J Am Coll Nutr 26:295-302 Korhola A, Blom T (1996) Marked early 20th century pollution

and the subsequent recovery of Toolo Bay central Hel- sinki as indicated by subfossil diatom assemblage changes. Hydrobiologia 341:169-179. doi:10.1007/BF00 018120

Mackereth FJH (1969) A short core sampler for subaqueous deposits. Limnol Oceanogr 14:145-151

Makarov MI, Haumaier L, Zech W (2002) Nature of soil organic phosphorus: an assessment of peak assignments in the diester region of 31P NMR spectra. Soil Biol Biochem 34:1467-1477. doi:10.1016/S0038-0717(02)00091-3 McKelvie ID (2007) Inositol phosphates in aquatic systems. In:

Turner BL, Richardson AE, Mullaney EJ (eds) Inositol phosphates: linking agriculture and the environment. CAB International, Wallingford, pp 261-277

Smith MTE, Cade-Menun BJ, Tibbett M (2006) Soil phosphorus dynamics and phytoavailability from sewage sludge at different stages in a treatment stream. Biol Fertil Soils 42:186-197. doi: 10.1007/s00374-005-0014-0 Sommers LE, Harris RF, Williams JDH, Armstrong OF, Syers

JK (1972) Fractionation of organic phosphorus in lake sediments. Soil Sci Soc Am Proc 36:51-54

Suzumura M, Kamatani A (1993) Isolation and determination of inositol hexaphosphate in sediments from Tokyo Bay.

Geochim Cosmochim Acta 57:2197-2202. doi:10.1016/

0016-7037(93)90561-A

Suzumura M, Kamatani A (1995a) Mineralization of inositol hexaphosphate in aerobic and anaerobic marine-sediments—implications for the phosphorus cycle. Geochim

Cosmochim Acta 59:1021-1026. doi: 10.1016/0016-7037 (95)00006-2

Suzumura M, Kamatani A (1995b) Origin and distribution of inositol hexaphosphate in estuarine and coastal sediments.

Limnol Oceanogr 40:1254-1261

Tikkanen M, Korhola A, Seppa H, Virkanen J (1997) A long- term record of human impacts on an urban ecosystem in the sediments of Toolonlahti Bay in Helsinki Finland.

Environ Conserv 24:326-337. doi:10.1017/S03768929 9700043X

Turner BL, Newman S (2005) Phosphorus cycling in wetlands:

the importance of phosphate diesters. J Environ Qual 34:1921-1929. doi:10.2134/jeq2005.0060

Turner BL, Richardson AE (2004) Identification of scyllo- inositol phosphates in soils by solution phosphorus-31 nuclear magnetic resonance spectroscopy. Soil Sci Soc Am J 68:802-808

Turner BL, Paphazy MJ, Haygarth PM, McKelvie ID (2002) Inositol phosphates in the environment. Philos Trans R Soc Lond Ser B 357:449-469. doi:10.1098/rstb.2001.0837 Turner BL, Mahieu N, Condron LM (2003a) Phosphorus-31 nuclear magnetic resonance spectral assignments of phosphorus compounds in soil NaOH-EDTA extracts.

Soil Sci Soc Am J 67:497-510

Turner BL, Mahieu N, Condron LM (2003b) Quantification of nryo-inositol hexakisphosphate in alkaline soil extracts by solution 31P NMR spectroscopy and spectral deconvolution. Soil Sci 168:469-478. doi: 10.1097/00010694- 200307000-00002

Turner BL, Mahieu N, Condron LM (2003c) The phosphorus composition of temperate pasture soils determined by NaOH-EDTA extraction and solution 31P NMR spectroscopy. Org Geochem 34:1199-1210. doi:10.1016/

80146-6380(03)00061-5

Turner BL, Cade-Menun BJ, Condron LM, Newman S (2005) Extraction of soil organic phosphorus. Talanta 66:294- 306. doi:10.1016/j.talanta.2004.11.012

Turner BL, Newman S, Newman J (2006) Organic phosphorus sequestration in subtropical treatment wetlands. Environ Sci Technol 40:727-733. doi: 10.1021/es0516256 Turner BL, Richardson AE, Mullaney EJ (eds) (2007) Inositol

phosphates: linking agriculture and environment. CAB International, Wallingford

Vaalgamaa S (2004) The effect on urbanisation on Laajalahti Bay Helsinki City as reflected by sediment geochemistry.

Mar Pollut Bull 48:650-662. doi:10.1016/j.marpolbul.

2003.10.008

Watts EE, Dean PAW, Martin RR (2002) 31P nuclear magnetic resonance study of sediment microbial phospholipids. Can J Anal Sci Spectrosc 47:127-133

Weckstrom K (2006) Assessing recent eutrophication in coastal waters of the Gulf of Finland (Baltic Sea) using subfossil diatoms. J Paleolimnol 35:571-592. doi:

10.1007/sl0933-005-5264-l

Weimer WC, Armstrong DE (1979) Naturally occurring organic phosphorus compounds in aquatic plants. Environ Sci Technol 13:826-829. doi:10.1021/es60155a003 Wetzel RG (2001) Limnology. Lake and river ecosystems.

Academic Press, San Diego