Biogeochemistry (2014) 119:361–370 DOI 10.1007/s10533-014-9972-3

Turnover of DNA-P and phospholipid-P in lake sediments Julia V. Paraskova • Per J. R. Sjo¨berg Emil Rydin

•

Received: 16 October 2013 / Accepted: 20 February 2014 / Published online: 19 March 2014 Ó Springer International Publishing Switzerland 2014

Abstract Identifying and quantifying the forms of phosphorus (P) in lake sediments is a prerequisite for understanding lake trophic status and possible exports of P downstream. Organic P is one of the most important P forms found in the sediment, where orthophosphate diesters, including DNA and phospholipids, represent a degradable P pool that can support primary production and eutrophication. In this study, sediment cores from the eutrophic Lake Erken ˚ nnsjo¨n, both in steady and the oligotrophic Lake A state regarding long-term P input revealed trends in the degradation of DNA-P and PL-P with sediment depth. Comparisons were performed based on the differentiation of essentially permanent or recalcitrant P and temporary, potentially mobile P for the respective fractions. The temporary P pool was defined as the part of the total P pool calculated for values higher than the

level at which the measured P concentration converged to a constant value and the recalcitrant pool was defined as the difference between the total and the temporary. The temporary diester-P pool comprised over 20 % of the total temporary P in Lake Erken and ˚ nnsjo¨n. The decrease in P around 4 % in Lake A concentrations with depth was more rapid for DNA-P compared to PL-P in both lakes, suggesting that DNAP has a more prominent role in internal loading. The study shows that P mobilization potential can be different for different P fractions, which is important when assessing their contribution to internal loading of P within an aquatic system. Keywords Phosphorus turnover  Sediment  ˚ nnsjo¨n DNA  Phospholipids  Erken  A

Introduction Responsible Editor: Jacques C Finlay

Electronic supplementary material The online version of this article (doi:10.1007/s10533-014-9972-3) contains supplementary material, which is available to authorized users. J. V. Paraskova  P. J. R. Sjo¨berg Department of Chemistry - BMC, Uppsala University, P.O. Box 599, Husargatan 3, 751 24 Uppsala, Sweden E. Rydin (&) Erkenlaboratoriet, Department of Ecology and Genetics, EBC, Uppsala University, Norr Malma 4200, Norrta¨lje, 761 73 Uppsala, Sweden e-mail: [email protected]

Phosphorus (P) is an essential element for primary production and the important role it plays in the eutrophication of aquatic ecosystems gives it unprecedented environmental significance (Foy 2005). The dynamics of P is complex, because it can be readily degraded, solubilized or mineralized, but also because P bioavailability and turnover are affected by many factors (Pierzynski et al. 2005; Benitez-Nelson and Buesseler 1999; Bostro¨m et al. 1988). Water bodies retain matter from the catchment and sediment accumulation areas represent sites where matter is

123

362

buried and transferred from the biosphere to the geosphere. On a catchment scale, lake sediments are often viewed as a sink for P. However, a fraction of the accumulated P originating from both organic and inorganic sources may be released back into the water column, via internal loading, thereby making sediments a source of P (Bostro¨m et al. 1988). Recycled sediment P can thus be a major factor contributing to maintaining eutrophic conditions in lakes (Carey and Rydin 2011). Some allochthonous and autochthonous forms of P are deposited and buried in the sediment, while others are mobilized and enable the diffusion of dissolved phosphate to the water column (Holdren and Armstrong 1980; Drake and Heaney 1987; Hupfer and Lewandowski 2005). Quantitatively, organic P represents one of the most important P forms found in the sediment (Edlund and Carman 2001), where degradable organic forms of P represent the main primary source of recyclable P in the sediment (Ahlgren et al. 2005; Reitzel et al. 2006; Rydin 2000). Since P cycling is dependent on a great number of physical, chemical and biological processes, including the type of P being recycled, the prerequisite for accurate assessment of internal P loading is thorough quantification of the different P forms present in the sediments (Dittrich et al. 2013). Under steady-state conditions the vertical distribution of P in the sediment profile enables the differentiation of temporary P and recalcitrant P and several studies have shown that P concentrations decline with depth until reaching convergence (Rydin 2000; Hupfer and Lewandowski 2005). For example, Hupfer and Lewandowski (2005) defined two pools of P, temporary and permanent, after observing converging P concentrations in the older layers of the sediments for various inorganic and organic P. Given that no further decline in concentration is observed below the convergence point the amount and forms of P found in the underlying layers can be considered more or less permanently buried within the sediment. Consequently, the P found above the convergence concentration level can be considered temporary (Rydin 2000). The relative size of the temporary P pool and the estimate of the time it takes for it to degrade can be used to compare systems with different nutrient status and the contribution of the temporary organic P forms to the total P (TP) pool can be used as a measure for internal loading. The main difficulty in studying organic P in environmental samples is that there are many different

123

Biogeochemistry (2014) 119:361–370

forms of organic P, all of which have potentially different degradation and turnover rates, and that compound-specific methods for extraction and quantification are limited. In addition to traditional sequential extraction schemes, developed for the separation of inorganic forms of P but which indirectly estimate the total organic P in a sample (Chang and Jackson 1957; Psenner and Pucsko 1988), there are procedures that fractionate P into pools based on their relative solubility or biological relevance (Bowman and Cole 1978; Hedley et al. 1982; Pant et al. 1994). Solution P-31 nuclear magnetic resonance spectroscopy (31P NMR) on sodium hydroxide (NaOH) extracts has given insight into the distribution of organic P groups based on their binding structure. NMR spectra allows for the detection and monitoring of phosphate monoesters, phosphate diesters, phosphonates, polyphosphates and pyrophosphates (Newman and Tate 1980; Reitzel et al. 2006; Wang and Pant 2010). However, the resolution, especially in the orthophosphate diester chemical shift region, is somewhat limited and the alkaline environment presents a risk of hydrolysis of the labile organic P forms (Cade-Menun 2005). The rationale behind this project is driven by both the fact that orthophosphate diesters have been shown to be important in P cycling in aquatic systems, and the need to overcome the inability of other techniques, such as NMR, to resolve this region, via new methods. In the present work, P derived from two diester-P sources, DNA and phospholipids (PLs), is examined with a recently developed method (Paraskova et al. 2013), which offers selective extraction and quantification of DNA-P and PL-P in particulate matter. Sediment depth distribution of DNA-P and PL-P is studied in two lakes, one oligotrophic and one eutrophic. The sediment cores provide a unique opportunity for studying the transformation rates of biogenic P species, because both lakes are considered to be in steady state regarding P load. The purpose of examining two distinct organic P fractions and their distribution within the sediment is to show that P mobilization potential can be different for different P fractions. This fact is of importance when assessing their contribution to internal loading of P within an aquatic system. The main goal of this study is to show that this methodology is an effective tool for studying the abundance, degradability and turnover of DNA and PLs in sediments of lakes with different nutrient status.

Biogeochemistry (2014) 119:361–370

363

Fig. 1 Positioning of the sampling sites (crosses) in the studied lakes and their location in Sweden (circles)

Materials and methods Study sites ˚ nnsjo¨n (63°160 N 12°330 E) is an oligotrophic Lake A ˚ re in mountain lake, situated in the municipality of A northern Sweden. The lake is part of the Indalsa¨lven river system in the province of Ja¨mtland. The lake area is 65 km2 with a mean depth of 15 m. The catchment area is unpopulated and the lake is unregulated, making it a prime example of a ‘‘system in its natural state’’ (Ahlgren et al. 2006). There are no external P load ˚ nnsjo¨n. However, as the lake is measurements for Lake A situated in pure wilderness, it can be considered to have had a constant external P load over the last century, albeit variations between years caused by differences in factors such as precipitation can be expected. The only inhabited place within the catchment, the small town of ˚ nn, has had a declining number of inhabitants, from A above one hundred in the mid-1900s to about 50 today. Lake Erken (59°510 N, 18°350 E) is a eutrophic lake north of Norrta¨lje in the province of Uppland in central Sweden. The lake area is 24 km2 with a mean depth of 9 m. The catchment is a good study object of a system in a steady state, because the external P load has been constant over the course of a century due to a more or less constant population and unchanged agricultural

practices (Rydin 2000). The steady-state rationale is additionally supported by the fact that the total P (TP) in the water column has remained constant since measurements were first made in 1930. Samples Sediment cores were collected with a gravity core sampler at 17 m depth in Lake Erken and 21 m depth ˚ nnsjo¨n in September 2012 (Fig. 1). The at Lake A study areas have been chosen to represent sediment accumulation areas (Weyhenmeyer et al. 1997). The cores were sliced into 1 cm segments directly after sampling and refrigerated at 4 °C until analysis. Sediment dating Reitzel et al. (2007) established a sedimentation rate for Lake Erken based on cesium-137 (137Cs) activity. The same methodology was used to assess the ˚ nnsjo¨n (Table S1; Fig. sedimentation rate for Lake A S1 in Online Resource). Calculations Calculations of the annual average sedimentation rates were based on measured amounts of accumulated

123

364

matter in the sediment core per reference area of 1 m2 in each 1 cm layer under the assumption of a steady state, i.e. constant yearly input of matter. Total DNA-P and total PL-P contributions up to each depth in 1 cm increments were calculated. The convergence level for all P fractions was set at the concentration where measurements reached a constant value, assessed as a difference of less than a few percent between adjacent sediment layers. The temporary P pool was set as the fraction of the total P pool calculated for values higher than the convergence level and the recalcitrant pool was defined as the difference between the total and the temporary (Hupfer and Lewandowski 2005). To highlight the differences in the degradation rates of DNA-P and PL-P in the two lakes the relative contribution of the respective fraction towards the temporary TP pool in each studied core was calculated. Additionally, the level at which 50 % of the temporary P pool of the respective fraction was lost was marked (Tables S2–S4 in Online Resource). Chemicals and instrumentation All chemicals were of analytical grade (Sigma-Aldrich, Germany), unless otherwise specified. Ultrafiltration was achieved by centrifugation with Nanosep Centrifugal Devices 30 K (PALL Life Sciences). A vortex mixer (Vortex-GenieÒ 2, Scientific Industries, Inc.) with an attachment (MoBio Laboratories, Inc.) was used for mechanical cell disruption. Purification of DNA was done with a mixture of phenol, chloroform and isoamyl alcohol (v:v:v 25:24:1) and a mixture of chloroform and isoamyl alcohol (v:v 24:1). P content was quantified by UV/Vis-spectrometry with UNICAM 5626 UV/Vis spectrometer (Unicam Limited, Cambridge, UK) with a 4 cm flow-through cuvette. Calibration was done with solutions prepared from singleelement P stock (Spectrascan, Teknolab AB, Kungsbacka, Sweden) by dilution with Milli-Q water (Millipore, Bedford, MA).

Analysis Water content was determined by gravimetry after oven-drying the samples at 100 °C to constant weight. Organic content was determined after dry ashing at 550 °C for 4 h of the oven-dried samples. Total P (TP) was determined after dry ashing and hydrochloric acid

123

Biogeochemistry (2014) 119:361–370

digestion by the molybdenum blue method (Murphy and Riley 1962) using ascorbic acid as a reducing agent and a detection wavelength of k = 882 nm. The extraction and quantification of DNA-P and PLP followed a method by Paraskova et al. (2013). A short description of the applied method follows. DNA-P was extracted with extraction buffer based on sodium chloride, ethylenediaminetetraacetic acid (EDTA) and Tris–HCl, glass beads, lysozyme, proteinase K and sodium dodecyl sulfate (SDS). Following purification, the DNA was separated by ultrafiltration. The PL-P was extracted with a single-phase mixture of chloroform, methanol and citrate buffer (pH 4.0). The mixture was split into two phases, the organic phase was collected and the solvent was evaporated under a stream of nitrogen. Phosphate in the DNA material was released through digestion with potassium persulfate. Phosphate in the PL material was released through dry ashing at 550 °C for 4 h and subsequent dissolution of the sample in 1 mL 1 M hydrochloric acid. All samples and standard solutions were autoclaved for 1 h at 120 °C. After cooling, acidic samples were neutralized with an aliquot of 1 M sodium hydroxide. All experiments were carried out in duplicate or triplicate and mean values were used in the calculations. Precision was evaluated by calculating relative standard deviation (RSD) as a percentage from the ratio of the standard deviation and the average in mg P/g DW sample.

Results Sediment characteristics and results of the experiments are presented in Table 1 (Lake Erken) and Table 2 ˚ nnsjo¨n). Both lakes showed a decline in water (Lake A content, organic content and P content with sediment depth and, consequently, age. All P fractions measured showed the highest P contents in the uppermost sediment layers with a convergence at some deeper layer. A common convergence level between 0.7 and 0.8 mg P/g DW is anticipated (Fig. 2), however, in Lake Erken the decline in TP continues throughout the ˚ nnsjo¨n the decline 30 cm core studied, while in Lake A in TP levels out at about 6–7 cm. The relative changes in TP between adjacent measurements are less than 1 % below this point. Subsequently, the convergence level was set at 7 cm, where sediment is approximately 30 years old. The temporary and recalcitrant P pools

Biogeochemistry (2014) 119:361–370

365

Table 1 Sediment characteristics and extraction results for Lake Erken (mean ± SD, n = 3) Depth (cm)

Agea (years)

DWb (%)

Organic contentc (%)

Water content (%)

Density (g/cm3)

# of years per cm layer

TPd (mg P/g DW)

DNA-Pe (mg P/g DW)

PL-Pf (mg P/g DW)

0-1

1

6.3 ± 0.2

21.2 ± 2.8

94

1.03

1.0

2.12 ± 0.07

0.284 ± 0.004

0.083 ± 0.005

1-2

2

7.4 ± 0.0

21.0 ± 0.3

93

1.04

1.2

2.14 ± 0.02

0.269 ± 0.019

0.083 ± 0.001

2-3

3

8.1 ± 0.2

19.2 ± 0.6

92

1.04

1.3

1.93 ± 0.11

0.255 ± 0.014

0.081 ± 0.004

3-4

5

8.6 ± 0.0

19.8 ± 0.3

91

1.04

1.4

1.77 ± 0.06

0.215 ± 0.030

0.075 ± 0.004

4-5

6

9.2 ± 0.1

19.3 ± 0.4

91

1.05

1.5

1.54 ± 0.01

0.212 ± 0.012

0.077 ± 0.004

5-6

8

9.7 ± 0.0

19.0 ± 0.2

90

1.05

1.6

1.40 ± 0.02

0.209 ± 0.003

0.070 ± 0.001

6-7

9

10.1 ± 0.1

19.2 ± 0.4

90

1.05

1.6

1.31 ± 0.07

0.162 ± 0.005

0.070 ± 0.004

9-10

15

11.0 ± 0.1

18.3 ± 0.4

89

1.06

1.8

1.31 ± 0.06

0.146 ± 0.000

0.066 ± 0.001

14-15

26

15.2 ± 0.4

16.2 ± 0.1

86

1.09

2.5

1.07 ± 0.02

0.035 ± 0.007

0.046 ± 0.006

19-20

39

16.9 ± 0.3

15.3 ± 0.1

83

1.10

2.8

1.01 ± 0.03

0.043 ± 0.001

0.038 ± 0.002

24-25 29-30

54 70

17.9 ± 0.1 19.4 ± 0.3

14.3 ± 0.3 14.3 ± 0.5

82 81

1.10 1.11

3.0 3.3

0.97 ± 0.09 0.83 ± 0.12

0.026 ± 0.004 0.018 ± 0.002

0.021 ± 0.002 0.017 ± 0.001

a

Sediment age determined by Reitzel et al. (2007)

b

Dry weight (DW) determined after drying until constant weight at 100 °C

c

Organic content measured by ashing the dried samples for 4 h at 550 °C

d

Total phosphorus (TP) determined after acid digestion of ashed samples, using the molybdenum blue method (Murphy and Riley 1962)

e, f

DNA-P and phospholipid-P (PL-P) determined with method of Paraskova et al. (2013)

˚ nnsjo¨n (mean ± SD, n = 3) Table 2 Sediment characteristics and extraction results for Lake A Depth (cm)

Agea (years)

DWb (%)

Organic contentc (%)

Water content (%)

Density (g/cm3)

# of years per cm layer

TPd (mg P/g DW)

DNA-Pe (mg P/g DW)

PL-Pf (mg P/g DW)

0-1

2

13.6 ± 0.1

10.10 ± 0.04

86

1.08

1.8

0.84 ± 0.03

0.054 ± 0.018

0.024 ± 0.007

1-2

5

20.8 ± 0.1

8.19 ± 0.15

79

1.13

2.8

0.81 ± 0.06

0.029 ± 0.014

0.018 ± 0.001

2-3

8

24.5 ± 0.2

7.49 ± 0.08

76

1.16

3.4

0.87 ± 0.04

0.047 ± 0.010

0.015 ± 0.001

3-4

13

31.6 ± 0.1

5.95 ± 0.05

68

1.22

4.6

0.74 ± 0.03

0.026 ± 0.005

0.012 ± 0.000

4-5

17

32.0 ± 0.2

5.88 ± 0.03

68

1.23

4.7

0.79 ± 0.04

0.028 ± 0.003

0.014 ± 0.002

5-6

23

37.5 ± 0.1

5.31 ± 0.04

62

1.28

5.8

0.75 ± 0.04

0.023 ± 0.001

0.008 ± 0.001

6-7

29

37.9 ± 0.5

5.77 ± 0.02

62

1.28

5.8

0.73 ± 0.01

0.026 ± 0.004

0.009 ± 0.001

9-10

47

40.0 ± 0.3

5.36 ± 0.07

60

1.30

6.2

0.74 ± 0.01

0.026 ± 0.000

0.009 ± 0.001

14-15

79

41.6 ± 0.3

5.49 ± 0.12

58

1.32

6.6

0.76 ± 0.05

0.029 ± 0.000

0.007 ± 0.001

19-20

113

43.7 ± 0.1

5.19 ± 0.04

56

1.34

7.0

0.69 ± 0.01

0.024 ± 0.000

0.008 ± 0.001

a

Sediment age determined after (Reitzel et al. 2007)

b

Dry weight (DW) determined after drying until constant weight at 100 °C

c

Organic content measured by ashing the dried samples for 4 h at 550 °C

d

Total phosphorus (TP) determined after acid digestion of ashed samples, using the molybdenum blue method (Murphy and Riley 1962)

e, f

DNA-P and phospholipid-P (PL-P) determined with method of Paraskova et al. (2013)

for Lake Erken were approximated for the entire profile because a clear convergence level has not been observed. In Lake Erken the temporary TP pool

comprises 26 % of the TP in the chosen part of the ˚ nnsjo¨n, the corresponding sediment core. In Lake A share is less than 10 % (Table 3).

123

366

Biogeochemistry (2014) 119:361–370

Fig. 2 Total phosphorus (TP) distribution with age for Lake ˚ nnsjo¨n (solid markers) and Lake Erken (open markers). A Results show mean values of triplicate measurements

˚ nnsjo¨n (solid Fig. 3 DNA-P distribution with age for Lake A markers) and Lake Erken (open markers). Results show mean values of triplicate measurements

˚ nnsjo¨n Table 3 Total P (TP) pools in Lake Erken and Lake A

Table 4 DNA-P pools in Lake Erken calculated above respective convergence level

TP (g/m2 core)

(%)

Lake ˚ nnsjo¨n A pool

TP (g/m2 core)

(%)

Pool

DNA-P (g/m2 core)

Permanent

38

74

Permanent

65

92

Permanent

0.6

Temporary

13

26

Temporary

6

8

Temporary

1.8a

Total

2.4

Lake Erken pool

DNA-P In the surface sediment of Lake Erken the DNA-P content is close to an order of magnitude higher than in ˚ nnsjo¨n (Fig. 3; the surface sediment of Lake A Tables 1, 2). The DNA-P decline is extensive in the eutrophic lake, where only a few percent of the surface layer DNA-P remains present in the deepest layers. In comparison, approximately half of the surface DNA-P ˚ nnsjo¨n. is detected in the deepest layers of Lake A From a depth of 15 cm in Lake Erken and 4 cm in ˚ nnsjo¨n the DNA-P concentration is the same, Lake A around 0.03 mg P/g DW, and remains relatively constant further down in the sediment profiles. Based on this convergence point, the relative sizes of the temporary and recalcitrant P pools for each lake were calculated (Tables 4, 5 and S3 in Online Resource).

PL-P Similarly to the DNA-P, the highest concentrations of PL-P are found in the surface sediments of both lakes, although these are lower than DNA-P concentrations and their decline is also slower (Fig. 4; Tables 1, 2). A

123

Temporary TP (g/m2 core)

% of Total DNA-P

% of Temporary TP

24 13.2b

76

14

100

a

convergence level at 26 years

b

convergence level at [70 years

˚ nnsjo¨n calculated above Table 5 DNA-P pools in Lake A respective convergence level Pool

DNA-P (g/m2 core)

Permanent

0.3

Temporary

0.1a

Total

0.4

Temporary TP (g/m2 core)

% of Total DNA-P

% of Temporary TP

72 5.6b

a

convergence level at 13 years

b

convergence level at 29 years

28

1.9

100

common convergence concentration level around 0.01 mg P/g DW is anticipated (Tables S2A, S2B in Online Resource), however, in Lake Erken the decline in PL-P continues throughout the 30 cm core studied, ˚ nnsjo¨n the decline in PL-P levels out at while in Lake A about 5 cm. The temporary and recalcitrant P pools for Lake Erken were approximated for the entire profile. ˚ nnsjo¨n the size of the pools were calculated For Lake A based on the convergence level (Tables 6, 7 and S4 in Online Resource).

Biogeochemistry (2014) 119:361–370

367

˚ nnsjo¨n (solid Fig. 4 PL-P distribution with age for Lake A markers) and Lake Erken (open markers). Results show mean values of triplicate measurements

Table 6 Phospholipid P (PL-P) pools in Lake Erken calculated above respective convergence level Pool

PL-P (g/m2 core)

Permanent

0.8

Temporary

1.1a

Total

1.9

Temporary TP (g/m2 core)

% of Total PL-P

% of Temporary TP

40 13.2b

60

9

100

a

convergence level at [70 years

b

convergence level at [70 years

˚ nnsjo¨n calTable 7 Phospholipid P (PL-P) pools in Lake A culated above respective convergence level Pool

PL-P (g/m2 core)

Permanent

0.16

Temporary

0.10a

Total

0.26

Temporary TP (g/m2 core)

% of Total PL-P

% of Temporary TP

40 5.6b

a

convergence level at 23 years

b

convergence level at 29 years

60

1.8

100

Discussion ˚ nnsjo¨n are considered to be in Lake Erken and Lake A steady state in regards to external nutrient load and productivity over the time period represented by the studied sediment cores (Rydin 2000; Ahlgren et al. 2006). The different layers in the sediment represent different ages and the accumulated sediment can be viewed as a P repository in which diagenetic processes can be studied. Both P mobilization and retention

within the sediment are dependent on many factors, and hypothetically, different chemical forms of P contribute differently to internal loading. Under the assumption of steady state, the yearly contributions to the surface sediment concentrations of DNA-P and PL-P are considered constant and the decline in the concentration of each of these fractions with depth represents mineralization. Consequently, the mineralization of P leads either to the release of phosphate into the water column or its entrapment within the sediment in some other form. In Lake Erken, mobilized phosphate is expected to be released from the sediment since eutrophic lake sediments appear to have limited ability for phosphate binding, while in ˚ nnsjo¨n the phosphate is expected to be buried Lake A in the sediment, because oligotrophic lake sediments generally have better capability to immobilize P (Carey and Rydin 2011). Reitzel et al. (2007) established a sedimentation rate for Lake Erken based on 137Cs activity to be 656 g DW m-2 year-1. Subsequent calculations based on the assumption of a constant accumulation, determined the age of the 30 cm sediment layer to be approximately 70 years old, which agrees well with previously established age. Similar dating was under˚ nnsjo¨n. Under the taken for the sediment of Lake A assumption of constant sedimentation of matter, the sedimentation rate was calculated to be 834 g DW m-2 year-1. The age of the 20 cm sediment layer in ˚ nnsjo¨n can thus be assumed to be about Lake A 115 years. The difference in age and accumulation rates between the sediment profiles of the two lakes is due to the higher share of inorganic material in the ˚ nnsjo¨n, which leads to the sediment of Lake A formation of a denser sediment core.

DNA-P The convergence concentration level for DNA-P is ˚ nnsjo¨n and reached at the depth of 5 cm in Lake A 15 cm in Lake Erken. The changes in concentration in the deeper levels are very small, implying that the remainder of this fraction is more or less permanently buried in the sediment. The decrease of DNA-P concentration with depth is considerably more rapid in the upper part of the sediment, with a 50 % decrease observed in the upper 2 cm of the sediment in Lake ˚ nnsjo¨n and in the upper 5 cm of the sediment in Lake A

123

368

Erken. The size of the total DNA-P pool above the convergence point and consequently, the relative contribution of the temporary DNA-P to the temporary TP pool, is six times larger in the eutrophic lake. The most striking difference, however, is the ratio between the temporary and recalcitrant DNA-P pools in the two lakes. In Lake Erken more than 75 % of the DNA-P is ˚ nnsjo¨n almost 75 % is temporary, while in Lake A recalcitrant. These findings provide strong evidence that DNA-P in the eutrophic lake plays an important role in internal loading. A central issue for understanding P turnover in aquatic systems is ascertaining the source of organic P in the sediment profile. P originating from DNA and PLs may represent settling of dead organic matter, e.g. phytoplankton, or P incorporated into the microbial community in response to the input of degradable carbon. However, it does not appear to be of consequence whether the DNA is of intracellular or extracellular origin. The fate and ecological relevance of extracellular DNA (eDNA) has been reviewed before. Pietramellara et al. (2009) suggests a way of estimating the amount of eDNA by plotting the amount of microbial DNA versus microbial biomass. The extrapolation of a positive intercept corresponds to the amount of eDNA. Haglund et al. (2003) studied the metabolic state of bacteria in the sediment of Lake Erken separating nucleoid-containing cells capable of metabolic activity, defined as live cells, from cells lacking nucleoids and thus incapable of growth, deemed as dead cells. Their results showed that bacterial abundance of both dead and live cells was highest at the surface and decreased with depth for both dead and live cells. The nucleoid-containing cell abundance decreased from 4.3 9 1010 cells (g DW-1) at the surface sediment to 0.5 9 1010 cells (g DW-1) at the depth of 25 cm, the lowest measured sediment. This corresponds to a decrease of circa 90 %, assuming 100 % abundance at the surface sediment, which is in agreement with the DNA-P degradation results in our study. DNA-P content in the different sediment layers correlates well (R2 ¼ 0:94) with the concentration of live cells in corresponding sediment depth (Fig. 5). The intercept close to zero suggests that virtually all DNA-P found in the sediment of Lake Erken originates from bacteria. The assumption made here is that DNA, whether associated or not with bacteria remains in the sediment layer with which it

123

Biogeochemistry (2014) 119:361–370

Fig. 5 Correlation plot of DNA-P and total bacterial count in Lake Erken sediment, under the assumption that 100 % abundance is found at the highest concentration (top layer). Results of present study plotted versus results of Haglund et al. (2003). Each point represents relative abundance at a given layer

first comes in contact. In other words, it does not migrate vertically in the sediment profile.

PL-P The degradation of PLs in the two studied lakes differs ˚ nnsjo¨n the degradation occurs significantly. In Lake A in the top 6 cm of the sediment, similarly to DNA-P. The total PL-P pool above the convergence point is half the size of the DNA-P pool, and accounts for less than 2 % of the TP pool, however, based on estimates of sediment age, it takes twice as long for the pool’s turnover. The ratio between the temporary and recalcitrant PL-P pool is similar to that of DNA-P, suggesting that the PL-P contribution to mineralized ˚ nnsjo¨n is also very small. The sediment P in Lake A degradation of PLs in Lake Erken is constant in the whole 30 cm studied. Hypothetically the concentration at convergence for PL-P is similar to that of Lake ˚ nnsjo¨n, which indicates that further degradation A takes place in the older sediment. The PL-P pools have nevertheless been calculated for the profile studied. The temporary PL-P pool is somewhat larger than the recalcitrant PL-P pool, suggesting that PL-P can also be important for internal loading. The relative contribution of temporary PL-P to temporary TP is 9 %, slightly smaller than the contribution of temporary DNA-P. However, since degradation is much slower, the impact on internal loading ought to be smaller. It is recognized that one pathway for P cycling in aquatic systems is through microbial turnover.

Biogeochemistry (2014) 119:361–370

Fig. 6 Correlation plot of PL-P and total bacterial count in Lake Erken sediment, under the assumption that 100 % abundance is found at the highest concentration (top layer). Results of present study plotted versus results of Haglund et al. (2003). Each point represents relative abundance at a given layer

Bacteria re-mineralize dissolved and particulate matter and are an important component in aquatic food webs (Heath 2005). Total bacteria count in the work of Haglund et al. was correlated with the PL-P concentration in the presented study (R2 ¼ 0:85). The assumption made was that bacteria cells, both those with and without a nucleus, have PL containing membranes. The position of the intercept indicates a surplus of PLs, which suggests that approximately one-third of the PLs in the sediment, is not associated with bacteria cells (Fig. 6). The origin of the remaining fraction PL-P could be attributed to cell membranes from e.g. phytoplankton. The role of organic P in lake sediments is important and its total amount, relative proportion to the TP content and the rate of mobilization all have to be considered when the relative contributions of individual P fractions to internal loading are studied. It is essential to investigate water bodies with different nutrient status in order to assess the role of organic P in internal loading. It has been previously suggested that if diagenesis is fast, sediment does not play an important role in temporary P storage (Hupfer and Lewandowski 2005). However, the speed of diagenesis can be an important factor when considering which forms of P contribute to internal loading (Dittrich et al. 2013). The substantial decline of DNA-P and PL-P concentrations with sediment depth in Lake Erken supports the hypothesis that diester-P in sediment is an importance source of P. The relative contribution of the temporary pool of diester-P to the temporary TP pool is over 20 % and can be considered to be a driving force behind internal loading. It

369

represents a degradable, and therefore potentially bioavailable, form of P. DNA-P is degraded faster than PL-P, which further shows that different P forms can contribute differently to internal loading under the same conditions. The practical implications include the need to address organic P separately when control measures for P mediation are being proposed. The degradation of organic P compounds in Lake Erken has been examined previously with 31P NMR (Ahlgren et al. 2005; Reitzel et al. 2007) and provide comparable results to the presented study. It is however important to note that the present method is both faster and cheaper. It has been acknowledged that quantification of organic P compounds with NMR has to be performed with care due to difficulties associated with peak integration. The advantage of the current method is that it selectively targets the two diester P compounds. It is undeniable that there are multiple challenges associated with organic P determination in environmental samples and there are currently no methods that accurately map the entire organic P pool. Therefore it is of importance to develop and test new methods for the quantification of the organic fractions that may be small, but not insignificant. The presented results provide evidence for the recycling of DNA-P and PL-P in the sediment and suggest that these two fractions are an important source of bioavailable organic P in aquatic systems. The orthophosphate diester fraction can comprise a significant part of the P turnover in the sediment. Because the relative size of the P pools is significantly larger in the eutrophic lake, diester-P can be considered a major contributor to the internal recycling of P and the maintaining of the lake’s trophic status. Acknowledgments This project was funded by Stiftelsen Lantbruksforskning, SLF (Swedish Farmers’ Foundation for Agricultural Research) and Olsson-Borghs foundation.

References Ahlgren J, Tranvik L, Gogoll A, Waldeba¨ck M, Markides K, Rydin E (2005) Sediment depth attenuation of biogenic phosphorus compounds measured by P-31 NMR. Environ Sci Technol 39(3):867–872. doi:10.1021/es049590h Ahlgren J, Reitzel K, Danielsson R, Gogoll A, Rydin E (2006) Biogenic phosphorus in oligotrophic mountain lake sediments: differences in composition measured with NMR spectroscopy. Water Res 40(20):3705–3712

123

370 Benitez-Nelson CR, Buesseler KO (1999) Variability of inorganic and organic phosphorus turnover rates in the coastal ocean. Nature 398(6727):502–505. http://www.nature. com/nature/journal/v398/n6727/suppinfo/398502a0_S1. html Bostro¨m B, Andersen JM, Fleischer S, Jansson M (1988) Exchange of phosphorus across the sediment—water interface. Hydrobiologia 170:229–244. doi:10.1007/ bf00024907 Bowman RA, Cole CV (1978) An Exploratory Method for Fractionation of Organic Phosphorus From Grassland Soils. Soil Sci 125(2):95–101 Cade-Menun BJ (2005) Characterizing phosphorus in environmental and agricultural samples by P-31 nuclear magnetic resonance spectroscopy. Talanta 66(2):359–371 Carey C, Rydin E (2011) Lake trophic status can be determined by the depth distribution of sediment phosphorus. Limnol Oceanogr 56(6):13 Chang SC, Jackson ML (1957) Fractionation of Soil Phosphorus. Soil Sci 84(2):133–144 Dittrich M, Chesnyuk A, Gudimov A, McCulloch J, Quazi S, Young J, Winter J, Stainsby E, Arhonditsis G (2013) Phosphorus retention in a mesotrophic lake under transient loading conditions: Insights from a sediment phosphorus binding form study. Water Res 47(3):1433–1447. doi:10. 1016/j.watres.2012.12.006 Drake JC, Heaney SI (1987) Occurence of phosphorus and its potential remobilization in the littoral sediments of a productive English lake. Freshw Biol 17(3):513–523. doi:10. 1111/j.1365-2427.1987.tb01072.x Edlund G, Carman R (2001) Distribution and diagenesis of organic and inorganic phosphorus in sediments of the Baltic proper. Chemosphere 45(6–7):1053–1061. doi:10. 1016/S0045-6535(01)00155-2 Foy RH (2005) The return of the phosphorus paradigm: agricultural phosphorus and eutrophication. In: Sims JT, Sharpley AN (eds) Phosphorus: agriculture and the environment american society of agronomy. Madison, Wisconsin, pp 911–939 Haglund AL, Lantz P, To¨rnblom E, Tranvik L (2003) Depth distribution of active bacteria and bacterial activity in lake sediment. FEMS Microbiol Ecol 46:31–38 Heath RT (2005) Microbial turnover of organic phosphorus in aquatic systems. In: Turner BL, Frossard E, Baldwin DS (eds) Organic phosphorus in the environment. CABI, Wallingford, pp 185–203 Hedley MJ, Stewart JWB, Chauhan BS (1982) Changes in inorganic and organic soil-phosphorus fractions induced by cultivation practices and by laboratory incubations. Soil Sci Soc Am J 46(5):970–976

123

Biogeochemistry (2014) 119:361–370 Holdren GC, Armstrong DE (1980) Factors affecting phosphorus release from intact lake sediment cores. Environ Sci Technol 14(1):79–87. doi:10.1021/Es60161a014 Hupfer M, Lewandowski J (2005) Retention and early diagenetic transformation of phosphorus in Lake Arendsee (Germany)—consequences for management strategies. Arch Hydrobiol 164(2):143–167. doi:10.1127/0003-9136/ 2005/0164-0143 Murphy J, Riley JP (1962) A modified single solution method for the determination of phosphate in natural waters. Anal Chim Acta 27:31–36 Newman RH, Tate KR (1980) Soil phosphorus characterisation by 31P nuclear magnetic resonance. Commun Soil Sci Plan 11(9):835–842. doi:10.1080/00103628009367083 Pant HK, Edwards AC, Vaughan D (1994) Extraction, molecular fractionation and enzyme degradation of organically associated phosphorus in soil solutions. Biol Fertil Soils 17(3):196–200. doi:10.1007/bf00336322 Paraskova JV, Rydin E, Sjo¨berg PJR (2013) Extraction and quantification of phosphorus derived from DNA and lipids in environmental samples. Talanta 115:336–341. doi:10. 1016/j.talanta.2013.05.042 Pierzynski GM, McDowell RW, Sims JT (2005) Chemistry, cycling, and potential movement of inorganic phosphorus in soils. Phosphorus: agriculture and the environment. American Society of Agronomy, Madison, pp 53–86 Psenner R, Pucsko R (1988) Phosphorus fractionation advantages and limits of the method for the study of sediment P origins and interactions. Ergebnisse der Limnologie 30: 43–60 Reitzel K, Ahlgren J, Gogoll A, Jensen HS, Rydin E (2006) Characterization of phosphorus in sequential extracts from lake sediments using 31P nuclear magnetic resonance spectroscopy. Can J Fish Aquat Sci 63(8):1686–1699. doi:10.1139/f06-070 Reitzel K, Ahlgren J, DeBrabandere H, Waldeba¨ck M, Gogoll A, Tranvik L, Rydin E (2007) Degradation rates of organic phosphorus in lake sediment. Biogeochemistry 82(1):15–28 Rydin E (2000) Potentially mobile phosphorus in Lake Erken sediment. Water Res 34(7):2037–2042. doi:10.1016/ S0043-1354(99)00375-9 Wang J, Pant H (2010) Identification of organic phosphorus compounds in the Bronx River bed sediments by phosphorus-31 nuclear magnetic resonance spectroscopy. Environ Monit Assess 171(1–4):309–319. doi:10.1007/ s10661-009-1280-3 Weyhenmeyer GA, Ha˚kanson L, Meili M (1997) A validated model for daily variations in the flux, origin, and distribution of settling particles within lakes. Limnol Oceanogr 42(7):1517–1529