w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/watres

In situ, high-resolution imaging of labile phosphorus in sediments of a large eutrophic lake Shiming Ding a,*, Chao Han a, Yanping Wang b, Lei Yao a, Yan Wang a, Di Xu a, Qin Sun c, Paul N. Williams d, Chaosheng Zhang e a

State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China b College of Forest Resources and Environment, Nanjing Forestry University, Nanjing 210037, China c Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes, Ministry of Education, College of Environment, Hohai University, Nanjing 210098, China d Institute for Global Food Security, School of Biological Sciences, Queen's University Belfast, Belfast BT9 7BL, United Kingdom e GIS Centre, Ryan Institute and School of Geography and Archaeology, National University of Ireland, Galway, Ireland

article info

abstract

Article history:

Understanding the labile status of phosphorus (P) in sediments is crucial for managing a

Received 18 November 2014

eutrophic lake, but it is hindered by lacking in situ data particularly on a catchment scale. In

Received in revised form

this study, we for the first time characterized in situ labile P in sediments with the Zr-oxide

28 January 2015

diffusive gradients in thin films (Zr-oxide DGT) technique at a two-dimensional (2D),

Accepted 3 February 2015

submillimeter resolution in a large eutrophic lake (Lake Taihu, China, with an area of

Available online 14 February 2015

2338 km2). The concentration of DGT-labile P in the sediment profiles showed strong variation mostly ranging from 0.01 to 0.35 mg L
1 with a considerable number of hotspots.

Keywords:

The horizontal heterogeneity index of labile P varied from 0.04 to 4.5. High values appeared

Eutrophication

at the depths of 0e30 mm, likely reflecting an active layer of labile P under the sediment

High resolution

ewater interface (SWI). Concentration gradients of labile P were observed from the high-

Phosphorus

resolution 1D DGT profiles in both the sediment and overlying water layers close to the

Sediment-water interface

SWI. The apparent diffusion flux of P across the SWI was calculated between
21 and

Zr-oxide DGT

65 ng cm
2 d
1, which showed that the sediments tended to be a source and sink of overlying water P in the algal- and macrophyte-dominated regions, respectively. The DGTlabile P in the 0e30 mm active layer showed a better correlation with overlying water P than the labile P measured by ex situ chemical extraction methods. It implies that in situ, high-resolution profiling of labile P with DGT is a more reliable approach and will significantly extend our ability in in situ monitoring of the labile status of P in sediments in the field. © 2015 Elsevier Ltd. All rights reserved.

* Corresponding author. Tel./fax: þ86 25 86882207. E-mail address: [email protected] (S. Ding). http://dx.doi.org/10.1016/j.watres.2015.02.008 0043-1354/© 2015 Elsevier Ltd. All rights reserved.

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

1.

Introduction

Eutrophication is a major environmental issue for freshwater ecosystems in the world. This phenomenon is defined by over-enrichment of a water body with one or more limiting growth factors needed for photosynthesis, resulting in excess growth of phytoplankton and algal and the formation of hypoxia (Schindler, 2006). Nutrient enrichment accelerated by anthropogenic activities has been identified as the major cause of eutrophication (Smith and Schindler, 2009). This is in particular linked to point and nonpoint discharges of phosphorus, which is the nutrient that often limits the biological productivity in aquatic ecosystems (Lu¨rling and Oosterhout, 2013). Sediment is the major sink for P in lakes. It plays an important role in daily and seasonal cycling of P through entrapment and mobilization of P primarily by iron redox cycling (Jiang et al., 2008; Smith et al., 2011). This process has been recognized as an important factor in regulating the trophic state of shallow lakes (Søndergaard et al., 1998). The release of P from sediments can be high and sustained for a long period, thereby contributing a significant amount of P to the water column (Fan et al., 2004; Welch and Cooke, 2005). The internal loading of P is regarded as a major factor responsible for the delay of lake recovery following a reduction in external P input (Søndergaard et al., 2007). Accordingly, understanding the labile status of P in sediments is crucial for managing or recovering a eutrophic lake. Two basic processes are involved in the dynamic release of sediment P to the overlying water. One is the diffusion process of pore water P depending on the concentration gradients in the vicinity of the sedimentewater interface (SWI) (Chowdhury and Al Bakri, 2006; Xu et al., 2012; Wu et al., 2001). The other one is the resupply process of solid P to pore water P through the release of P from the binding sites of sediment solids (Ding et al., 2010; Monbet et al., 2008). The available solid P fraction is in dynamical equilibrium with pore water P and varies under different biogeochemical environments and conditions, where the easily changeable or mobile P fractions are termed as labile P. Accordingly, an accurate measurement of labile P should simultaneously include the two P fractions by considering the dynamics between pore water and sediment solids. Until now, measurement of sediment labile P (or mobile P) is mostly performed using ex situ chemical extraction methods (Condron and Newman, 2011). Sequential extraction procedures are often employed to extract several labile chemical forms operationally defined, including those P loosely adsorbed to sediment solids and associated with Fe oxyhydroxides (Zhu et al., 2013). A direct and ease-of-use method is to employ a single-step extraction procedure, such as the use of ascorbic acid solution to remove/dissolve the reactive solid phase P pool (Rozan et al., 2002). However, it has been recognized that the common ex situ techniques can cause considerable analytical errors due to changes of the sediment samples in chemical properties when exposing them to the air, such as oxygenation of Fe(II) and the subsequent adsorption of soluble P by Fe(III) oxyhydroxides (Condron and Newman, 2011). The labile P information is

101

typically collected at a cm level, while the sediment is characterized by distinct heterogeneity in chemical distributions even on a submillimer spatial scale (Stockdale et al., 2009). Meanwhile, the dynamic characteristics between pore waters and sediment solids are not incorporated. Diffusive gradients in thin films (DGT) is an in situ, dynamical, and high-resolution technique which has the potential to satisfy the requirement for the accurate measurement of labile P in sediments (Zhang et al., 2014; Huo et al., 2014). This technique was established based on Fick's First Law of Diffusion. It relies on a diffusion flux of a solute to the DGT device, in which a binding layer was held and used for immobilization of the solute. The DGT-measured fractions come from pore water and the further release of the solute from the sediment solids to resupply the pore water (Zhang et al., 1995). The uptake process of DGT fits the kinetic solidesolution interaction of P regarding its lability in sediments. Moreover, DGT has been developed for two-dimensional (2D), submillimeter measurement of labile P in sediments in combination with analysis using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) or a routine procedure using 2D slicing (Ding et al., 2011; Santner et al., 2010). Recently, the high-capacity Zr-oxide DGT, a new DGT technique which used amorphous zirconium hydroxide as the binding agent in its binding gel, has been combined with the computer-imaging densitometry (CID) technique for the submillimeter-scale imaging of labile P in sediments (Ding et al., 2013). A similar development had been achieved by
s et al. (2011) on another gel technique, diffusive equiliPage bration in thin films (DET). As the CID operation process is quite easy in batch analysis of samples (Widerlund and Davison, 2007), the Zr-oxide DGT provides a potentially powerful tool in in situ measurement of labile P in a large number of sediment profiles. The high-resolution measurement with DGT enables imaging of the heterogeneous distribution of labile P and especially its concentration gradient in the vicinity of the SWI, which is the major force to drive the exchange of P between the sediment and overlying water (Xu et al., 2012). Despite these advantages of DGT, its application is still largely limited to laboratory trials, while it is seldom tested in situ on a large spatial scale, such as on a catchment scale. This has impeded the overall understanding of the lability of P in sediments and its cycling in lakes especially with large spatial variation in ecological type and water quality. Lake Taihu is the 3rd largest freshwater lake in China. It has a surface area of 2338 km2 and mean depth of 1.9 m. This lake has different water quality status and diverse ecological types across the whole lake region (Duan et al., 2009; Ma et al., 2008; Yu et al., 2013). Algal blooms have occurred with an increasing frequency and intensity in the northern bays which have spread to the central and south parts of the lake (Duan et al., 2009). In contrast, aquatic vegetation (mainly submerged and floating-leaved species) are mainly distributed in several bays in the east and southeast, where the water is clear and algal blooms rarely occur (Ma et al., 2008). Large difference in total P content was found in the sediments (Bai et al., 2009). The internal loading of P is thus suspected to be an important factor influencing the regional trophic state of the lake, which has been partly verified using

102

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

conventional ex situ methods (Fan et al., 2004; Zhu et al., 2013). In this study, the spatial changes of labile P were investigated in sediments of Lake Taihu using the Zr-oxide DGT. The flat DGT probe for sediment deployment was designed to satisfy the requirement for field sampling. It was then applied to in situ collect the distribution information of labile P at a submillimeter level from a total number of 30 sites representative of the whole status of the lake. The heterogeneous distribution of labile P in sediment profiles was analyzed, and the apparent fluxes of P across the sedimentewater interface were estimated. Finally, the relationship between DGT-labile P and overlying water P were investigated and compared to ex situ chemical extraction methods.

2.

Materials and methods

2.1.

Description of sampling sites

A total number of 30 sites were selected in this study (Fig. 1). These sites were generally evenly distributed in the lake. Among these sites, Sites 1 to 8 were located in the algaldominated region in the north part of the lake. Sites 22 to 30 were located in the macrophyte-dominated regions in the east and southeast of the lake. Other sites have no visible macrophyte coverage, but many of them occasionally suffer from floating blooms transported by wind. It should be noted that Site 29 was in an aquaculture region, where an amount of fresh aquatic vegetation and small aquatic animals are annually added in the water column to feed the crabs. Aquatic

Fig. 1 e Location of sampling sites in Lake Taihu. The circle and square legends show sites located in the algaldominated (Sites 1e8) and macrophyte-dominated (Sites 22e30) regions, respectively. The triangle legend shows sites in the other regions which occasionally suffer from floating blooms (Sites 9e21).

vegetation also dominated this site over the years under the human intervention.

2.2.

DGT preparation and field deployment

The principle of DGT was demonstrated in Supplementary Information (Fig. S1). The diffusive gel was prepared with 15% acrylamide and a 0.3% agarose-derived cross-linker according to published procedures (Zhang and Davison, 1995). The Zr-oxide gel capable of a high-resolution measurement of P was prepared according to Ding et al. (2011). When assembling the DGT probe, the Zr-oxide gel (with the Zr-oxide settled surface facing up) was covered sequentially by a 0.40-mmthick diffusion gel and a 0.13-mm-thick cellulose nitrate filter membrane (Whatman, 0.45-mm pore size). A new design for the DGT probe was developed to satisfy the requirements of field sampling, especially in the accurate identification of the SWI position when retrieving the probe from the sediment. The details are presented in the Supplementary Information (Fig. S2, S3). The developed Zroxide DGT probes were transported to Lake Taihu in October, 2013, and deoxygenated with nitrogen for at least 16 h. They were transported to each sampling site by placing them in a container. The probes were inserted into the sediment using a releasing device (Ding et al., 2013). They were retrieved after 48 h and brought to the laboratory. Two speedboats were used to perform the fieldwork from the north and the south of the lake, respectively. Deployment as well as retrieval of the DGT probes was finished within a day in order to obtain a synchronous sampling at all the sites.

2.3.

Analyses of DGT and other samples

After retrieval of the probes, the accumulated masses of P in the Zr-oxide binding gels were measured using a recently developed coloration technique (Ding et al., 2013). Each binding gel was first heated in hot water (85  C) for 5 d. The gel was then immersed in the mixed molybdenum reagent, with the volume of the added reagent 200 times of the volume of the gel. The vessel was kept in an incubator at 35  C for 45 min. After retrieval, the gel was immediately rinsed using cool water and then immersed in cool water for at least 5 min to stop the color development. The surfaces of the Zr-oxide settled sides were scanned using a flat-bed scanner (Canon 5600F) at a resolution of 600 dpi, corresponding to a pixel size of 42 mm  42 mm. The grayscale intensity of the scanned images corresponding to the open window of the flat DGT probe was analyzed with ImageJ 1.46 (downloaded from http://rsb.info.nih.gov/ij). It should be noted that the actual spatial resolution measured for labile P was lower than defined due to the lateral diffusion of P in the diffusive layer during DGT deployment (Warnken et al., 2006). The overlying water and surface sediment (5 cm) were sampled at each site, with the sediment collected using a gravity sampler. They were transported to the laboratory under low temperature. Total P, total dissolved P (TDP) and dissolved reactive P (DRP) in water samples were measured using the molybdenum blue method following standardized treatments (Jin and Tu, 1990). The sediments from each site were freeze-dried at
80  C. After that, total P in sediments

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

was determined by extraction of P with 1.0 M HCl after treatment of the sediments at 450  C (Ruban et al., 2001). Labile P in sediment samples was determined using both single-step and sequential extraction procedures. An ascorbic acid solution was used in the single-step procedure to extract reactive solid phase P pool bound to amorphous iron (Rozan et al., 2002). The sequential extraction was performed according to the procedure by Rydin (2000). The sediment was extracted in order by 1 M NH4Cl, 0.11 M Na2S2O4/NaHCO3 (BD), and 0.1 M NaOH solutions. Concentration of labile (potential mobile) P was the sum of the extracted total P from the first two steps and nonreactive P (NaOH-nrp) from the last step (Reitzel et al., 2005).

2.4.

where Dg is the thickness of the diffusive layer, D is the diffusion coefficient of the phosphate in the diffusive layer, t is the deployment time, A is the exposed area of the gel, and M is the accumulated mass of P calibrated from the grayscale intensity on the Zr-oxide gel surface using the equation established earlier (Ding et al., 2013). The distribution of 1D DGT-labile P in the vicinity of the SWI was used to estimate the apparent diffusion flux of P across the SWI. Taking different mechanisms in resupplying the DGT uptake in the sediment and overlying water into consideration, the apparent flux was calculated as the sum of those from the overlying water and the sediments respectively.

Data processing J ¼ Jw þ Js ¼
Dw

The concentration of labile P measured by DGT was calculated using the equation: CDGT ¼

MDg DAt

Table 1 e Concentration of total P and P fractions in sediment and overlying water samples collected in Lake Taihu. Sediment (mg g
1)

Samples

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 a

Overlying water (mg L
1)

TP

Labile Pa

Labile Pb

TP

DTP

DRP

436.5 533.9 941.2 467.4 1552.8 455.8 513.9 916.8 490.8 424.8 434.3 491.7 537.1 764.2 477.8 544.8 367.7 517.9 474.5 435.4 475.8 508.3 536.4 457.1 304.2 420.1 428.6 461.6 1209.5 420.9

190.0 182.9 184.7 116.6 352.4 231.2 90.1 235.9 120.7 112.8 151.5 109.5 113.2 123.4 146.4 79.9 145.4 117.2 135.4 194.5 125.3 82.3 159.1 156.9 83.5 113.8 150.4 111.6 377.4 153.1

257.8 295.2 258.4 315.4 718.2 318.9 281.9 321.4 157.3 274.3 263.2 261.1 300.4 286.6 278.6 277.7 263.5 251.1 250.3 280.8 259.3 265.7 300.2 288.6 234.9 252.8 245.4 255.5 487.2 279.0

0.134 0.117 0.108 0.084 0.200 0.138 0.081 0.037 0.114 0.097 0.111 0.064 0.101 0.057 0.082 0.041 0.041 0.075 0.070 0.038 0.106 0.023 0.074 0.032 0.038 0.059 0.038 0.017 0.051 0.065

0.078 0.126 0.079 0.064 0.143 0.086 0.068 0.057 0.059 0.062 0.065 0.030 0.112 0.023 0.029 0.027 0.041 0.009 0.026 0.006 0.053 0.024 0.014 0.084 0.023 0.017 0.051 0.018 0.014 0.036

0.045 0.073 0.029 0.038 0.073 0.066 0.038 0.028 0.030 0.026 0.029 0.022 0.048 0.015 0.019 0.011 0.024 0.004 0.010 0.003 0.028 0.006 0.006 0.026 0.006 0.013 0.018 0.005 0.004 0.028

Extracted by a single-step extraction with ascorbic acid solution (Rozan et al., 2002). b Extracted by a multiple-step extraction (Rydin, 2000). TP, DTP and DRP point to total P, dissolved total P and dissolved reactive P, respectively.

103

    dCDGT dCDGT
4Ds dxw ðx¼0Þ dxs ðx¼0Þ

where Jw and Js is the flux of P from the overlying water to the SWI and from the sediment to the SWI, respectively. J is the flux across the SWI. (dCDGT/dxw)(x ¼ 0) and (dCDGT/dxs)(x ¼ 0) are the DGT-labile concentration gradients (i.e., the slopes) in the overlying water and sediment, respectively. In order to achieve an accurate estimation of the flux, the distance of ~5 mm from the SWI to the overlying water as well as to sediment was used to fit the gradient. 4 is porosity in sediment and was estimated at 0.9 in the top 5 mm layer (Gao et al., 2009). Dw and Ds are the diffusion coefficients of H2PO
4 in water and sediment, respectively. Ds was calculated from the diffusion co3 efficient of H2PO
4 in water using 4 for 4  0.7 (Li and Gregory, 1974; Ullman and Aller, 1982). Horizontal heterogeneity index of labile P at each depth of 2D profiles was calculated as the standard deviation divided by the mean concentration of labile P according to Pischedda et al. (2011). Correlation analyses between P fractions were performed using SPSS ver. 13.0.

3.

Results and discussion

3.1. Concentrations of P in the overlying waters and sediments The concentrations of various P species in the overlying waters and sediments are listed in Table 1. Based on the 30 sites investigated, concentrations of total P (TP), total dissolved P (TDP), dissolved reactive P (DRP) in the overlying water samples varied between 0.017 and 0.20 mg L
1, 0.006e0.143 mg L
1, and 0.003e0.073 mg L
1, respectively. Concentration of TP in sediments ranged from 304 to 1552 mg kg
1. Concentrations of labile P in sediments, determined by a single-step and a sequential extraction procedure respectively, ranged from 80 to 377 mg kg
1 and 157e718 mg kg
1, respectively. They accounted for 15%e51% (single-step procedure) and 27%e77% (sequential extraction procedure) of the total P in sediments. Higher values of P, whether in the overlying waters or sediments, were found in the north part of Taihu, which were consistent with previous reports (Bai et al., 2009; Yu et al., 2013; Zhu et al., 2013). Site 29 from an aquaculture region also showed a high value in sediment P, but the concentration

104

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

of P in the overlying waters remained at a low level compared to the northern region.

3.2.

Two-dimensional distribution of DGT-labile P

The 2D profiles of labile P are shown in Fig. 2. The concentration of labile P mostly varied from 0.01 to 0.35 mg L
1. Generally, higher concentrations of labile P in the deeper depths of the profiles were observed in the northern algaldominated bays (Sits 1e5) and in the southeastern macrophyte-dominated regions (Sites 24, 29 and 30). Relatively high concentrations simultaneously appeared in the upper depths of the profiles in the north part of the lake (Sites

2e4 and 6e8). In contrast, lower concentrations of labile P were observed at other sites which were mainly distributed in the central and western parts of the lake, with the values in the deeper depths of the profiles being generally higher than those in the upper depths. A special feature was found for Site 27, which had an abrupt increase of labile P from the sediment to the overlying water. This feature may reflect a different environment controlling P lability between the sediment and overlying water. A similar phenomenon occurred in sediments of Sites 2, 4 and 7. In Site 22, there was a small dot (~5 mm diameter) with strongly enhanced concentrations of labile P (up to 0.48 mg L
1) near the bottom of the profile. There were quite a

Fig. 2 e Two-dimensional distribution of DGT-labile P in sediment-overlying water profiles. The dashed lines show the location of the sedimentewater interface.

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

105

few small concentrated dots (~1 mm diameter) at other sites (e.g., 12, 15, 16, and 18). The hotspot feature has been reported previously for labile P and sulfide in sediments, which were likely caused by strong decomposition of active organic matter (Ding et al., 2013; Widerlund and Davison, 2007).

above reactions, causing a relatively even distribution of labile P as observed. As a consequence, the partition in horizontal variability of labile P with depth may reflect a corresponding change in P lability originated from sedimentation history or mechanistic processes.

3.3.

3.4.

Horizontal heterogeneity of DGT-labile P

Changes of the horizontal heterogeneity index of labile P with depth in the 2D profiles are shown in Fig. 3. It is evident that four sites (including 15, 21, 23 and 26) showed high values up to 4.5. These sites are mainly distributed in the macrophytedominated regions (Sites 23 and 26) or the regions close to the lake bank (Sites 15 and 21). These regions generally have a high sedimentation rate, and the settling particles have not been mixed enough prior to sedimentation. Moreover, the sediments are easily disturbed by root growth, animal bioturation and human activities (Lewandowski et al., 2005;
s et al., 2011; Teal et al., 2013). Page Irrespective of a strong variation of the heterogeneity index with the depth among different sites, higher values were concentrated at the depths between 0 and 30 mm and then between 30 and 80 mm, followed by a low and stable state in the deeper depths below 80 mm. The great heterogeneity at the 0e30 mm depths may reflect a high activity of labile P in the young sediment layer less than 15 years old based on an average sedimentation rate of 2.1 mm yr
1 in Taihu (Qin et al., 2007). Furthermore, this layer is close to the SWI, where early diagenesis reactions take place at a high strength along with the sharp decrease in oxygen concentration and strong degradation of organic matter (Anschutz et al., 2007). Xu et al. (2012) found that iron(III) oxyhydroxides was reduced in this subsurface layer in Lake Taihu, which may exert a dominant effect on P remobilization and cause a large variability of labile P (Xu et al., 2012). At the 30e80 mm depths, the reduction of sulfate is strengthened in this lake (Yin and Fan, 2011). This process can affect the lability of P through the formation of insoluble iron sulfide compounds (e.g., pyrite) and the associated release of P bound on iron oxides (Rozan et al., 2002). At the deeper layers below the depth of 80 mm, P in sediments should become more inert and stable after undergoing the

Fig. 3 e Changes of horizontal heterogeneity index of labile P with depth in sediments of Lake Taihu. The numbers inside the plot are the serial numbers of sites.

One-dimensional distribution of DGT-labile P

The 1D distribution of DGT-labile P based on average horizontal concentrations at each depth is shown in Fig. 4. Similar to the 2D profiles, labile P showed different changes with depth among the different sites. A total of 16 sites had an increasing trend (Sites 1, 3, 5, 9, 11, 12, 14, 15, 17, 21, 23, 24, 25, 27, 29 and 30) with depth. A total of 4 sites showed small changes along the profile (Sites 8, 13, 22 and 28). Other sites showed a peak or valley in the middle zone (e.g., Sites 2, 4, 6, 7, 10, 16, 18 and 26), or had an irregular variation (e.g., Site 19). A strong peak appeared in Site 22, which corresponded to the hotspot observed in its 2D profile (Fig. 2). It should be noted that there were concentration gradients of labile P in the vicinity of the SWI in many sites. Such a phenomenon has been reported in sediments on P, sulfide and metal ions (Amato et al., 2014; Gao et al., 2009; Monbet et al., 2008). There were also steep concentration gradients in the overlying water in at least 11 sites (Sites 3e4, 6e8, 19, 23, 25, 26, 28, 29), which has been rarely reported by others. It is evident that high-resolution measurement greatly facilitated the discovery of this phenomenon. Among the 11 sites, 8 sites showed upward increases of labile P, while 3 sites (Sites 4 and 29) had opposite changes. It reflected that there was P diffusion from the overlying water to the SWI in the 8 sites and an inverse diffusion for the other 3 sites.

3.5. Estimation of apparent diffusion flux of P across the sedimentewater interface The apparent fluxes of P across the SWI for all the sites are shown in Fig. 5. The value ranged from
21 to 65 ng cm
2 d
1 (the negative and positive values represent the flux to the sediment and water, respectively). The values were among the SWI fluxes of P reported previously in Lake Taihu calculated from sediment core incubation or pore water P profiles (Fan et al., 2006; Zhang et al., 2006). It demonstrated that the sediment acted as a sink or a source of P to the overlying water in these sites. Among the 6 sites (Sites 1e6) located in the northern algal-dominated bay of the lake, 5 sites showed net P flux to the overlying water. In contrast, among the 9 sites (Sites 22e30) located in the southeastern macrophytedominated regions, 6 sites showed net P flux to the sediments. It reflected that the sediments tended to be a source of water P in the algal-dominated (or contaminated) region and to be a sink of water P in the macrophyte-dominated (or clear water) regions. The result was consistent with previous reports that the internal loading process of P from the sediment was stronger in contaminated or algal-dominated regions than in the clear-water or macrophyte-dominated regions (Kisand and Noges, 2003; Søndergaard et al., 1998; Wu et al., 2001; Zhu et al., 2013). This study thus provided an in situ and direct evidence to support this hypothesis.

106

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

Fig. 4 e One-dimensional distribution of DGT-labile P in sediment-overlying water profiles. The dotted lines showed the positions of the sedimentewater interface.

An exception is for Site 29 in the aquaculture region, which had the flux far exceeding the other sites. The greatest flux of P was comparable to the high amounts of labile P in sediment determined by DGT and extraction methods as mentioned earlier (Fig. 4, Table 1). The fluxed P from sediment in this site should be taken up rapidly by macrophyte, resulting in a low concentration of P in the overlying water as observed (Table 1). Most of the other sites showed small values in P flux, reflecting

that there was nearly an equilibrium in P diffusion between sediment and overlying water.

3.6. Relationship between DGT-labile P and overlying water P In order to test whether the DGT-labile P is a useful indicator in reflection of sediment P status, the correlation between

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

107

Fig. 5 e Apparent diffusion flux of P across the sedimentewater interface calculated based on the 1D DGT-labile P profiles.

DGT-labile P as well as P flux across the SWI with overlying water P was investigated. Total P and extracted labile P were also investigated for comparison. The data for Site 29 were excluded from this investigation due to its abnormal values. The results showed that total P in sediment had significant correlations (p < 0.05) with DRP, DTP and TP in the overlying water (Table 2). Better correlations were found for labile P from both the single- and multi-step extractions, reflected by a very significant level (p < 0.01) with the three overlying water P species. The best correlations were found for the DGT-labile P at 0e30 mm depths, especially with DRP. This is expected because the P measured by DGT was DRP in pore water and from the release of sediment solids. The correlations of the DGT-labile P at 0e80 mm depths with water DRP and DTP were also superior to those of total P and extracted labile P. It demonstrated that the DGT-labile P, especially those in the subsurface layer (0e30 mm) identified from labile P partition in horizontal heterogeneity, was a better indicator in reflecting the lability of sediment P compared to the traditional indicators. The DGT-derived P flux across the SWI was also

Table 2 e Correlation coefficients between sediment P and overlying water P.a Sediment P

Total P Labile P from single-step extraction Labile P from sequential extraction Average labile P from DGT measurement at 0e30 mm depths Average labile P from DGT measurement at 0e80 mm depths Apparent diffusion flux of P across the SWI a

Overlying water P DRP

DTP

TP

0.407* 0.596** 0.528** 0.813**

0.483* 0.564** 0.543** 0.790**

0.488* 0.585** 0.533** 0.611**

found to be correlated with DRP and DTP, further supporting this result.

4.

Conclusions

Imaging labile P distributions with the Zr-oxide DGT in a totally 30 sites showed a considerable variation at a 2D, submillimeter resolution. An active layer of labile P at the depths of 0e30 mm was identified according to the changes of the horizontal heterogeneity index. The apparent fluxes of P across the SWI were calculated using 1D DGT profiles, and their values ranged from
21 to 65 ng cm
2 d
1, from which it could be distinguished that the sediments tended to be a source and sink of overlying water P in the algal- and macrophyte-dominated regions, respectively. A better correlation was further found for the DGT-labile P in the 0e30 mm active layer with overlying water P in comparison to ex situ chemical extraction methods, demonstrating that DGT was a more reliable approach in monitoring of P status in sediments.

Acknowledgments The authors thank Prof. Weiping Hu for providing speed boats in field sampling. This study was jointly sponsored by the National Scientific Foundation of China (21177134, 41322011, 41471402) and the Nanjing Institute of Geography and Limnology, CAS (NIGLAS2010KXJ01).

0.680** 0.723** 0.386* 0.408* 0.477** 0.295

The data from Site 29 were excluded from statistic analysis; the one and two asterisks show the significant levels at p < 0.05 and 0.01, respectively.

Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.watres.2015.02.008.

108

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

references

Amato, E.D., Simpson, S.L., Jarolimek, C.V., Jolley, D.F., 2014. Diffusive gradients in thin films technique provide robust prediction of metal bioavailability and toxicity in estuarine sediments. Environ. Sci. Technol. 48 (8), 4485e4494. Anschutz, P., Chaillou, G., Lecroart, P., 2007. Phosphorus diagenesis in sediment of the Thau Lagoon. Estuar. Coast. Shelf Sci. 72 (3), 447e456. Bai, X.L., Ding, S.M., Fan, C.X., Liu, T., Shi, D., Zhang, L., 2009. Organic phosphorus species in surface sediments of a large, shallow, eutrophic lake, Lake Taihu, China. Environ. Pollut. 157 (8e9), 2507e2513. Chowdhury, M., Al Bakri, D., 2006. Diffusive nutrient flux at the sediment-water interface in Suma Park Reservoir, Australia. Hydrol. Sci. J. 51 (1), 144e156. Condron, L.M., Newman, S., 2011. Revisiting the fundamentals of phosphorus fractionation of sediments and soils. J. Soil Sediments 11 (5), 830e840. Ding, S.M., Jia, F., Xu, D., Sun, Q., Zhang, L., Fan, C.X., Zhang, C.S., 2011. High-resolution, two-dimensional measurement of dissolved reactive phosphorus in sediments using the diffusive gradients in thin films technique in combination with a routine procedure. Environ. Sci. Technol. 45 (22), 9680e9686. Ding, S.M., Wang, Y., Xu, D., Zhu, C.G., Zhang, C.S., 2013. Gelbased coloration technique for the submillimeter-scale imaging of labile phosphorus in sediments and soils with diffusive gradients in thin films. Environ. Sci. Technol. 47 (14), 7821e7829. Ding, S.M., Xu, D., Sun, Q., Yin, H.B., Zhang, C.S., 2010. Measurement of dissolved reactive phosphorus using the diffusive gradients in thin films technique with a highcapacity binding phase. Environ. Sci. Technol. 44 (21), 8169e8174. Duan, H.T., Ma, R.H., Xu, X.F., Kong, F.X., Zhang, S.X., Kong, W.J., Hao, J.Y., Shang, L.L., 2009. Two-decade reconstruction of algal blooms in China's Lake Taihu. Environ. Sci. Technol. 43 (10), 3522e3528. Fan, C.X., Zhang, L., Bao, X.M., You, B.S., Zhong, J.C., Wang, J.J., Ding, S.M., 2006. Migration mechanism of biogenic elements and their quantification on the sediment-water interface of Lake Taihu: II. chemical thermodynamic mechanism of phosphorus release and its source-sink transition. J. Lake Sci. 18 (3), 207e217 (in Chinese). Fan, C.X., Zhang, L., Qin, B.Q., Wang, S.M., Hu, W.P., Zhang, C., 2004. Estimation on dynamic release of phosphorus from wind-induced suspended particulate matter in Lake Taihu. Sci. China Ser. D-Earth Sci. 47 (8), 710e719. Gao, Y., Lesven, L., Gillan, D., Sabbe, K., Billon, G., De Galan, S., Elskens, M., Baeyens, W., Leermakers, M., 2009. Geochemical behavior of trace elements in sub-tidal marine sediments of the Belgian coast. Mar. Chem. 117 (1e4), 88e96. Huo, S., Zhang, J., Yeager, K., Xi, B., Wang, J., He, Z., Wu, F., 2014. High-resolution profiles of dissolved reactive phosphorus in overlying water and porewater of Lake Taihu, China. Environ. Sci. Pollut. Res. 21 (22), 12989e12999. Jiang, X., Jin, X., Yao, Y., Li, L., Wu, F., 2008. Effects of biological activity, light, temperature and oxygen on phosphorus release processes at the sediment and water interface of Taihu Lake, China. Water Res. 42 (8e9), 2251e2259. Jin, X.C., Tu, Q.Y., 1990. Lake Eutrophication Survey Specification, second ed. China Environmental Science Press, Beijing. Kisand, A., Noges, P., 2003. Sediment phosphorus release in phytoplankton dominated versus macrophyte dominated shallow lakes: importance of oxygen conditions. Hydrobiologia 506 (1e3), 129e133.

Lewandowski, J., Schadach, M., Hupfer, M., 2005. Impact of macrozoobenthos on two-dimensional small-scale heterogeneity of pore water phosphorus concentrations in lake sediments: a laboratory study. Limnol. Oceanogr. 50, 1106e1118. Li, Y.H., Gregory, S., 1974. Diffusion of ions in seawater and in deepsea sediments. Geochim. Cosmochim. Acta 38, 703e714. Lu¨rling, M., Oosterhout, F.v, 2013. Controlling eutrophication by combined bloom precipitation and sediment phosphorus inactivation. Water Res. 47 (17), 6527e6537. Ma, R.H., Duan, H.T., Gu, X.H., Zhang, S.X., 2008. Detecting aquatic vegetation changes in Taihu Lake, China using multi-temporal satellite imagery. Sensors 8 (6), 3988e4005. Monbet, P., McKelvie, I.D., Worsfold, P.J., 2008. Combined gel probes for the in situ determination of dissolved reactive phosphorus in porewaters and characterization of sediment reactivity. Environ. Sci. Technol. 42 (14), 5112e5117.
s, A., Teasdale, P.R., Robertson, D., Bennett, W.W., Scha € fer, J., Page Welsh, D.T., 2011. Representative measurement of twodimensional reactive phosphate distributions and codistributed iron(II) and sulfide in seagrass sediment porewaters. Chemosphere 85 (8), 1256e1261. Pischedda, L., Poggiale, J.C., Cuny, P., Gilbert, F., 2011. Oxygen distribution heterogeneity related to bioturbation quantified by planar optode imaging. In: Global Change: Mankind-Marine Environ. Interac, pp. 277e282. Qin, B.Q., Xu, P.Z., Wu, Q.L., Luo, L.C., Zhang, Y.L., 2007. Environmental issues of Lake Taihu, China. Hydrobiologia 581, 3e14. Reitzel, K., Hansen, J., Andersen, F.O., Hansen, K.S., Jensen, H.S., 2005. Lake restoration by dosing aluminum relative to mobile phosphorus in the sediment. Environ. Sci. Technol. 39 (11), 4134e4140. Rozan, T.F., Taillefert, M., Trouwborst, R.E., Glazer, B.T., Ma, S.F., Herszage, J., Valdes, L.M., Price, K.S., Luther, G.W., 2002. Ironsulfur-phosphorus cycling in the sediments of a shallow coastal bay: implications for sediment nutrient release and benthic macroalgal blooms. Limnol. Oceanogr. 47 (5), 1346e1354. Ruban, V., Lopez-Sanchez, J.F., Pardo, P., Rauret, G., Muntau, H., Quevauviller, P., 2001. Development of a harmonised phosphorus extraction procedure and certification of a sediment reference material. J. Environ. Monit. 3 (1), 121e125. Rydin, E., 2000. Potentially mobile phosphorus in Lake Erken sediment. Water Res. 34 (7), 2037e2042. Santner, J., Prohaska, T., Luo, J., Zhang, H., 2010. Ferrihydrite containing gel for chemical imaging of labile phosphate species in sediments and soils using diffusive gradients in thin films. Anal. Chem. 82, 7668e7674. Schindler, D.W., 2006. Recent advances in the understanding and management of eutrophication. Limnol. Oceanogr. 51 (1), 356e363. Smith, L., Watzin, M.C., Druschel, G., 2011. Relating sediment phosphorus mobility to seasonal and diel redox fluctuations at the sediment-water interface in a eutrophic freshwater lake. Limnol. Oceanogr. 56 (6), 2251e2264. Smith, V.H., Schindler, D.W., 2009. Eutrophication science: where do we go from here? Trends Ecol. Evol. 24 (4), 201e207. Søndergaard, M., Jeppesen, E., Lauridsen, T.L., Skov, C., Van Nes, E.H., Roijackers, R., Lammens, E., Portielje, R., 2007. Lake restoration: successes, failures and long-term effects. J. Appl. Ecol. 44 (6), 1095e1105. Søndergaard, M., Jensen, J.P., Jeppesen, E., 1998. Internal phosphorus loading in shallow Danish lakes. In: International Conference on Trophic Interactions in Shallow Freshwater and Brackish Waterbodies, Blossin, Germany, pp. 145e152. Stockdale, A., Davison, W., Zhang, H., 2009. Micro-scale biogeochemical heterogeneity in sediments: a review of

w a t e r r e s e a r c h 7 4 ( 2 0 1 5 ) 1 0 0 e1 0 9

available technology and observed evidence. Earth-Sci. Rev. 92 (1e2), 81e97. Teal, L.R., Parker, E.R., Solan, M., 2013. Coupling bioturbation activity to metal (Fe and Mn) profiles in situ. Biogeosciences 10 (4), 2365e2378. Ullman, W.J., Aller, R.C., 1982. Diffusion coefficients in nearshore marine sediments. Limnol. Oceanogr. 27, 552e556. Warnken, K.W., Zhang, H., Davison, W., 2006. Accuracy of the diffusive gradients in thin-films technique: diffusive boundary layer and effective sampling area considerations. Anal. Chem. 78 (11), 3780e3787. Welch, E.B., Cooke, G.D., 2005. Internal phosphorus loading in shallow lakes: importance and control. Lake Reserv. Manag. 21 (2), 209e217. Widerlund, A., Davison, W., 2007. Size and density distribution of sulfide-producing microniches in lake sediments. Environ. Sci. Technol. 41 (23), 8044e8049. Wu, F., Qing, H., Wan, G., 2001. Regeneration of N, P and Si near the sediment/water interface of lakes from Southwestern China Plateau. Water Res. 35 (5), 1334e1337. Xu, D., Wu, W., Ding, S., Sun, Q., Zhang, C., 2012. A highresolution dialysis technique for rapid determination of dissolved reactive phosphate and ferrous iron in pore water of sediments. Sci. Total Environ. 421e422 (0), 245e252. Yin, H., Fan, C., 2011. Dynamics of reactive sulfide and its control on metal bioavailability and toxicity in metal-polluted

109

sediments from lake Taihu, China. Arch. Environ. Contam. Toxicol. 60 (4), 565e575. Yu, T., Zhang, Y., Wu, F.C., Meng, W., 2013. Six-decade change in water chemistry of large freshwater lake Taihu, China. Environ. Sci. Technol. 47 (16), 9093e9101. Zhang, C.S., Ding, S.M., Xu, D., Tang, Y., Wong, M.H., 2014. Bioavailability assessment of phosphorus and metals in soils and sediments: a review of diffusive gradients in thin films (DGT). Environ. Monit. Assess. 186, 7367e7378. Zhang, H., Davison, W., 1995. Performance characteristics of diffusion gradients in thin films for the in situ measurement of trace metals in aqueous solution. Anal. Chem. 67 (19), 3391e3400. Zhang, H., Davison, W., Miller, S., Tych, W., 1995. In-situ high resolution measurements of fluxes of Ni, Cu, Fe, and Mn and concentrations of Zn and Cd in porewaters by DGT. Geochim. Cosmochim. Acta 59 (20), 4181e4192. Zhang, L., Fan, C.X., Wang, J.J., Zheng, C.H., 2006. Space-time dependent variances of ammonia and phosphorus flux on sediment-water interface in Lake Taihu. Environ. Sci. 27 (8), 1537e1543 (in Chinese). Zhu, M., Zhu, G., Li, W., Zhang, Y., Zhao, L., Gu, Z., 2013. Estimation of the algal-available phosphorus pool in sediments of a large, shallow eutrophic lake (Taihu, China) using profiled SMT fractional analysis. Environ. Pollut. 173 (0), 216e223.