Introduction

P in sediments is one of the major sources of P in the overlying water of shallow lakes (Golterman 1995) and has a critical role in lake P cycle (Boström et al. 1989; Jensen et al. 1992). Although external pollution loading has been effectively controlled, the water quality of lakes has not been significantly improved. This lack of improvement is primarily attributable to the P release of sediments, which becomes the major source of P loading of lakes, and to internal P loading, which affects the ecological restoration of lakes over a long period of time (Pettersson 1998; Boström et al. 1988). Most previous studies indicate that the process of the P release of sediments is accompanied by such biochemical processes as desorption, ligand exchange, mineralization, and enzymatic hydrolysis (Boström and Pettersson 1982; Rydin 2000; Christophoridis and Fytianos 2006). These processes are jointly influenced by physical (resuspension of bottom sediments) (Søndergaard et al. 1992), chemical (temperature, redox potential, pH value, and Fe/P ratio) (Jensen et al. 1992; Kleeberg and Kozerski 1997; Andersen 1982; Wu et al. 2014), and biological factors (microbial activity) (Søndergaard et al. 2003). The amount of mobile P in sediments is also the intrinsic factor that impacts P release. Under the action of diffusion and sediment diagenesis, mobile P affects the P release and P retention effect of sediments (Rydin 2000). Loosely adsorbed P (NH4Cl-P), redox-sensitive P (BD-P), and labile organic P (OP) (NaOH-nrp) are the P fractions that generally exhibit strong activities in lake sediments. Under the condition of bioturbation, anaerobic, and biomineralization, these P fractions become potentially mobile and promote the P release of sediments (Boström and Pettersson 1982). Blue algal booms in summer may deplete the dissolved oxygen (DO) of the overlying water and cause anaerobic P release at the sediment–water interface of the lakes. Even a small amount of P release, which accounts for only a small proportion of P in sediments, has a significant impact on the P concentration in the overlying water of the lakes (Boström and Pettersson 1982). Boström et al. (1985) showed that internal P loading in summer is due to a substantial depletion of the OP of sediments. Dissolved OP (DOP) is the essential form of dissolved P in pore-water. Under certain conditions (i.e., decomposition of organic matter, metabolism of microbial cells, and activity of benthic organisms), the depletion of OP becomes the major source of soluble reactive P (SRP) (Boström et al. 1989). The influences of DOP on the composition of P in overlying waters cannot be ignored. Hence, changes in the OP fractions of sediments are predictive of the evolution of the water quality of lakes. However, only scanty attention has been given so far to the influence of the OP release of sediments on the water quality of lakes. Its contribution to the P load of lakes and its influence mechanism are yet to be clarified.

Erhai Lake (EH) is the second largest freshwater lake at the Yunnan–Guizhou Plateau, with a drainage area of 2565 km2. Although EH generally has good water quality, the agricultural non-point source pollution and the livestock and poultry breeding pollution to its north are severe. Its aquatic vegetation is undergoing large-scale degeneration, and its water quality tends to be eutrophic (Meng 2013). EH is currently in the early stage of eutrophication and the sensitive stage of transformation; protecting and controlling the lake are crucial. At present, the P release of sediments throughout EH is relatively lower than the external P loading, and its influence on the P concentration in the water body is insignificant. However, as the intensity of the control on external loading increases, the influence of internal loading release in the northern part of the lake, which is already seriously polluted, also increases.

In this study, the sediment cores of EH were collected and subjected to an incubation experiment. The P release flux and P fractionation were estimated at different sediment depths. The variances in the P fractionation of sediments before and after the incubation experiment were compared to determine the impact of the P release of sediments on the P concentration in the overlying water of EH. The results provide references for the prevention and control of water pollution in EH.

Materials and methods

Study area

EH is located at Dali Bai Autonomous Prefecture, Yunnan Province. As the second largest freshwater lake in Yunnan Province, EH belongs to the Lancang River system (100°05′–100°17′E, 25°36′–25°58′N). The height of the water above sea level is 1974 m. The catchment area of the drainage basin is 2565 km2, and the lake is 42.0 km long and 8.4 km wide. At the normal water level of 1974.0 m (height for coastal defense), the water area of the lake is 249.8 km2, with an average water depth of 10.5 m and a maximum depth of 20.9 m. The storage capacity of the lake is approximately 2.56 × 109 m3. The retention time of the lake water is approximately 2.75 years (Peng et al. 2005). In the drainage basin of EH, 117 rivers of different sizes exist, including the Miju, Luoshi, Yong’an, Xi’er, Boluo, and Cangshan Shiba Rivers (Pan et al. 1999). Since 1970s, with the development of human activities in EH, the water quality continues to deteriorate (Han et al. 2005). The eutrophication status of EH is mainly caused by agricultural non-point source (Zhang et al. 2014). The algal density increased from 1.0 million cells/L in 1992 to 13.0 million cells/L in 2013. In the progress of eutrophication, the phytoplankton dominate species changed from Anabaena in 1996 to Microcystis in 2013 (Zhu et al. 2015).

Sample collection

A gravity core sampler was used (organic glass tube with a diameter of 90 mm) for the samples. Four representative sites were chosen in four regions of EH in August 2010 (north EH1, middle EH2, plateau at the center of the lake EH3, and south EH4). The depth of the collected sediments was approximately 30–40 cm. The sediment–water interface at the sampling site was distinct, with low artificial disturbance. The positions and descriptions of the sampling sites are shown in Fig. 1 and Table 1, respectively.

Fig. 1
figure 1

Map of the studied area with sampling sites

Full size image
Table 1 Main characteristics of the sampling sites in EH
Full size table

Experimental method

In situ incubation experiment

Four of the two replicate sediment cores were collected and transported to a laboratory on ice. The overlying water of the sediment cores was removed and placed into a sampling bottle for analysis. The depth of the core sediments was kept basically consistent. A rubber plug was inserted at the bottom. A total of 2 L of lake water, from which algae and suspended particles had been filtered, was slowly added along the tube wall. The artificial disturbance at the sediment–water interface was minimized. After the addition of lake water, all the core samples were vertically placed into a glass incubator containing lake water. The samples were incubated away from light and without a lid.

The experiment was performed in a cycle of 3 days (Fig. 2). In each cycle, 200 mL of the overlying water in the middle section was collected. After sampling, an equal amount of lake water filtered by 0.45 µm Whatman GF/F membrane was slowly replenished along the tube wall. The water samples in each cycle were collected at 6, 12, 24, 48, and 72 h after the experiment; a total of five samples were therefore collected in each cycle. After one cycle of incubation experiment, the overlying water was removed through a siphon. The sediment samples at a surface layer of 2 cm (0–2 cm) were collected, and the next cycle of the experiment started at an interface of 2 cm. The above steps were repeated until an interface of 20 cm was reached. A total of eight continuous cycles of static incubation experiment were conducted. For the first 10 cm, sediment cores were collected at every 2 cm. For the next 10 cm, sediment cores were collected at every 5 cm. The collected sediment samples were placed in plastic bags, which were sealed and preserved at −4 °C. After freeze drying, the sediments were ground and sieved.

Fig. 2
figure 2

Schematic diagram of in situ incubation experiment

Full size image

Fractionation test of P in sediments

The P fractionation in sediments was conducted using Psenner’s sequential extraction method (Psenner et al. 1988). This method is among the most commonly used for the sequential extraction of P in lake sediments and has many research applications (Kisand 2005; Liu et al. 2009; Ribeiro et al. 2008; Zhu et al. 2013; Dong et al. 2011). The detailed extraction steps are shown in Fig. 3. The P in the sediments was separated as follows: (a) NH4Cl-P (pore-water P, loosely adsorbed P) containing NH4Cl-rp [(loosely adsorbed inorganic P (IP)] and NH4Cl-nrp (loosely adsorbed OP); (b) BD-P (P forms sensitive to low redox potential, such as P adsorbed to iron and manganese); (c) NaOH-rp (P exchangeable with OH−, mainly Al) and NaOH-nrp (OP including bacteria-incorporated P), which was detected after digestion; (d) HCl-P (P forms sensitive to low pH, assumed to consist mainly of apatite); and (e) Res-P, which consists mainly of refractory OP and the inert IP fraction.

Fig. 3
figure 3

P fractionation scheme used in the experiment, following Psenner et al. (1988)

Full size image

Analytical method

The analyzed water samples included the in situ overlying water of the core sediments, lake water filtered with membrane, water samples collected in the experiment, and replenished lake water. Total P (TP), Dissolved Total P (DTP), and SRP were determined through the Mo–Sb spectrophotometric method. The DOP of the water sample was calculated using the following formula:

$${\text{DOP}} = {\text{DTP}} - {\text{SRP}}$$
(1)

The P release rate of sediments (flux) F i was calculated using the following formula (Steinman et al. 2004):

$$F_{i} = {{\left[ {V(c_{n} - c_{0} ) + \sum\limits_{j = 1}^{n} {V_{j - 1} (c_{j - 1} - c_{a} )} } \right]} \mathord{\left/ {\vphantom {{\left[ {V(c_{n} - c_{0} ) + \sum\limits_{j = 1}^{n} {V_{j - 1} (c_{j - 1} - c_{a} )} } \right]} {(S \cdot t)}}} \right. \kern-0pt} {(S \cdot t)}}$$
(2)

where F i is the release rate [mg/(m2·d)]; V is the volume of the overlying water of the core sediments (L); c n , c 0 , and c j−1 are the contents of a certain substance at n, 0 (initial), and j−1 sampling (mg/L), respectively; c a is the content of a substance in the added water sample (mg/L); V j−1 is the volume of water at j−1 sampling (L); S is the contact area of the sediment–water interface of the core sample (m2); and t is the experiment time (d).

Result

Characteristics of P release flux of surface sediments and the spatial distribution in EH

The P release characteristics at the sediment–water interface in different lake regions are shown in Fig. 4. The average release flux of SRP at the surface sediments is 0.18 ± 0.11 mg/(m2·d) and varies within the range of 0.09–0.33 mg/(m2·d). The spatial order of P release is as follows: the highest is found in the north of the lake region (EH1), followed by the middle (EH2), the south (EH4), and the plateau at the center of the lake (EH3). The release flux of SRP in the north is about 1.7, 2.9, and 3.7 times that of the middle, south, and plateau of the lake, respectively. The release flux of DOP varies in the range of 0.58–1.66 mg/(m2·d), with an average of 1.09 ± 0.46 mg/(m2·d), which is about 5.9 times that of SRP. The spatial distribution throughout the entire lake shows a pattern similar to the SRP distribution. The release flux of DOP in the northern lake region is the highest, with a value of 1.4, 1.8, and 2.8 times that of the south, the middle, and the plateau of the lake, respectively.

Fig. 4
figure 4

P release flux at the sediment–water interface of EH

Full size image

The variations in the cumulative release of P at the sediment–water interface at different lake regions are obvious throughout the entire test period (Fig. 5). Figure 5a shows that, in each time period, the cumulative release of SRP from the sediments in the northern lake region (EH1) is always higher than that in the other lake regions. Within 0–24 h of the incubation experiment, the release of SRP from sediments in different regions of EH increases constantly, and the difference is obvious after 24 h. In the northern lake region (EH1), the SRP of the sediments is continuously released into the overlying water. Other lake regions are basically in a state of exchange equilibrium. At the end of the experiment after 72 h, the cumulative release of SRP from the sediments in the northern lake region (EH1) reaches up to 0.90 mg/m2. The cumulative release of SRP in other lake regions is close, varying in the range of 0.35–0.43 mg/m2.

Fig. 5
figure 5

Temporal variation in the cumulative SRP release at the sediment–water interface of EH

Full size image

Compared with SRP, the cumulative release of DOP at the sediment–water interface constantly rises over time (Fig. 5). Therefore, within a short period of time, the release potential of DOP from the sediments is higher than that of SRP. At the end of the experiment (72 h), the cumulative release of DOP from the sediments of EH is as high as 1.94–5.69 mg/m2, with an average of 2.67 ± 0.86 mg/m2. The cumulative release of DOP in the northern lake region (EH1) is the highest, whereas that at the plateau of the center of the lake (EH3) is the lowest. At the beginning and end of the experiment, the cumulative release of DOP from the sediments shows large differences. At the end of the experiment, the cumulative release in the four lake regions increases by 0.8–2.8 times that at the beginning. The cumulative release of DOP in the north and south shows the largest increase.

The P release characteristics of the sediments in EH are further analyzed. The release flux of SRP at the sediment–water interface in the following lakes is compared: the EH, the Dianchi Lake at Yunnan–Guizhou Plateau (DC), and the lakes at the middle and lower reaches of the Yangtze River, including the Dongting Hu (DTH), Donghu (DH), Poyang Hu (PYH), Liangzihu (LZH), Honghu (HH), and Taihu (TH) Lakes (Fig. 6). The results show that the release flux of SRP from the sediments of EH is significantly lower than that of the other lakes. The average release flux of SRP from the sediments of the lakes at the Yunnan-Guizhou Plateau [1.39 mg/(m2·d)] is approximately 1.6 times higher than that of the lakes in the middle and lower reaches of the Yangtze River [0.86 mg/(m2·d)]. For the lakes in the same region, a large difference exists in the release flux of SRP between lakes with different eutrophication levels. The release flux of SRP in DC is approximately 14.5 times that of EH. In the middle and lower reaches of the Yangtze River, the release flux of SRP varies less significantly among different lakes. The release flux of SRP from sediments is influenced not only by the environmental conditions of the sediment–water interface but also by the water body itself and the general situation of sediment pollution. Our study shows that the TP of sediments of the lakes at the Yunnan–Guizhou Plateau is significantly higher than that of the lakes in the middle and lower reaches of the Yangtze River. The TP of the surface sediments of DC is as high as 2800 mg/kg (Mao et al. 2005). In TH, the average TP is only approximately 560 mg/kg (Yuan et al. 2010). Although the average TP of the sediments in EH is higher than that in TH, the stable forms of P are dominant. The overall pollution situation of the water body and sediments is less severe than that of the lakes in the middle and lower reaches of the Yangtze River. Therefore, the release flux of P from the sediments is also lower.

Fig. 6
figure 6

Comparison of the SRP release flux in the Yunnan Plateau lakes and that in the Lower Yangtze Region lakes (DC: Mao et al. 2005; TH: Fan et al. 2006; other data: Zhang et al. 2008)

Full size image

P release flux of core sediments in EH and the changes

The average release flux of SRP in the core sediments of EH is 0.22 ± 0.66 mg/(m2·d). The highest is found in the north (EH1) and south (EH4) [0.25–0.27 mg/(m2·d)], followed by the middle (EH2) [0.20 mg/(m2∙d)]. The average release flux of SRP is the lowest at the plateau of the center of the lake (EH3) [0.15 mg/(m2·d)]. The results are shown in Fig. 7. The release flux of SRP from the sediments at different depths first increases and then decreases with depth. For the same burial depth, the release flux of SRP differs slightly. The release flux of SRP at the sediment–water interface in the north (EH1) and middle (EH2) is higher than that in the south and at the plateau of the center of the lake.

Fig. 7
figure 7

Vertical distribution of the SRP release flux in the sediments of EH

Full size image

The release flux of SRP increases slightly with the burial depth and reaches a maximum of 0.42 and 0.27 mg/(m2·d) at 4–6 cm. The increase is 25 and 38 % in comparison with that at the sediment–water interface, respectively. The release flux of SRP significantly decreases by 64.5 and 28.4 % in the north and middle at depths of 6–8 cm, respectively. Below the depth of 8 cm, the release flux of SRP from the sediments in the north and middle varies at a small rate, with a variation coefficient of 12.3 and 13.9 %, respectively. The release flux of SRP at the sediment–water interface at the plateau of the center of the lake (EH3) and in the south (EH4) decreases. A significant increase occurs at depths of 2–4 cm by 127 and 303 %, respectively, in comparison with that at the sediment–water interface. The release flux of SRP reaches 0.20 and 0.46 mg/(m2·d), respectively. At depths of 4–6 cm, the release flux of SRP at the plateau of the center of the lake and in the south decreases slightly by 0.16 and 0.30 mg/(m2·d), respectively. In the south, the release flux of SRP gradually decreases with the burial depth (below 6 cm). The release flux of SRP at depths of 6–10 cm varies insignificantly at the plateau of the center of the lake. Below 10 cm, the release flux of SRP declines continuously to 0.07 mg/(m2·d) at a sediment depth of 20 cm. This value is slightly lower than the release flux at the sediment–water interface.

The distribution of the release flux of DOP at different depths in EH shows a larger difference than that of SRP. The release flux is generally high at the surface layer and low at the bottom (Fig. 8). The average release flux of DOP of core sediments is 0.89 ± 0.38 mg/(m2·d), which is approximately 4.1 times that of SRP. The highest is found in the north, which reaches 1.41 ± 0.24 mg/(m2·d). The average release flux of DOP in the other three lake regions is lower, which varies in the range of 0.50–0.87 mg/(m2·d). The release flux of DOP in the north and south is high, with a distinct vertical variation; that is, the release flux decreases with depth in a fluctuating manner until it stabilizes. The release flux of DOP from the sediments in the north reaches a maximum of 1.79 mg/(m2·d) at the burial depths of 6–8 cm. The highest value in the south occurs at the sediment–water interface. The variation rate of the release flux of DOP in the south along the vertical direction is higher than that in the north. The distribution of the release flux of DOP in the middle and at the plateau of the center of the lake shows a similar pattern. The release flux varies greatly with depth. At surface layers of 0–10 cm, the release flux increases with burial depth in a fluctuating manner. At burial depths of 10 and 15 cm, the release flux of DOP declines rapidly and reaches the lowest value of 0.40 and 0.22 mg/(m2·d), respectively. These values are only 43.6 and 38.3 % that of the sediment–water interface. Below the depth of 15 cm, the release flux of DOP increases to a certain extent. Below the depth of 20 cm, the release flux of DOP is 2.2 and 3.5 times that of the depths of 10 and 15 cm, respectively.

Fig. 8
figure 8

Vertical distribution of the DOP release flux in the sediments of EH

Full size image

Discussion

P form as an important factor that influences the P release of sediments at different burial depths in EH

The direction of the diffusion of P in pore-water is influenced by the P concentration gradient at the sediment–water interface (Baudo 1990). When the P concentration in the overlying water is higher than that in the pore-water, the P concentration gradient forms at the sediment–water interface. As a result, the P in the overlying water diffuses into the pore-water and finally enters the sediments. When the P concentration in the overlying water is lower than that in the pore-water, P is released from the pore-water to the overlying water. The release rate is affected by the concentration gradient (Liao et al. 1994; Yu et al. 1995). The physical and chemical properties of the sediments (i.e., pH, Eh, and DO) can also influence the P release flux of sediments. The content of mobile P of surface sediments is an indirect predictive factor of internal P release (Rydin 2000; Ramm and Scheps 1997).

The results show that the P release of sediments of various burial depths is affected by the transport and transformation of mobile P. In this study, the variations in the P fractionation in sediments before and after incubation were compared to discuss the effect of burial depth on P release. A small difference in IP content was found in the sediments before and after incubation, and the average variation rate is only 2.6 %. The difference is larger only at the sediment–water interface, with a variation rate of approximately 10.3 % (Fig. 9a). The reason for the small variation of IP content before and after incubation may be that HCl-P and Res-P, which are stable forms of P, are the dominant components of the IP in the sediments of EH. The stable forms of P account for 62 % of IP. HCl-P mainly consists of Ca-P, including authigenic apatite, carbonate-P, and a trace amount of hydrolyzed OP (Kaiserli et al. 2002). Given that HCl-P is highly stable in sediments, it is only released at a low pH value. During the incubation experiment, the pH value of the water is approximately 8; hence, the transport and transformation of HCl-P are insignificantly affected. BD-P is redox-sensitive P and usually refers to phosphates enveloped by the film of hydroxides and oxides of metal, such as Fe and Mn (Hupfer et al. 1995). BD-P is transferred and transformed through the breaking and synthesis of the chemical bond. When Eh is low at the sediment–water interface, Fe3+ is reduced to Fe2+, which causes the dissolution and release of BD-P into the pore-water. Although BD-P has a high activity, the water body was maintained at an aerobic level in this study (i.e., DO was approximately 6–9 mg/L). Therefore, the BD-P in the sediments hardly transfers and transforms. NaOH-rp refers to P bound to metal oxides (Fe, Al) and mainly includes the P adsorbed by the hydrous oxides of Fe and Al in particulates. NaOH-rp refers to a part of the phosphates of Fe and Al and is greatly influenced by pH. When the pH value at the sediment–water interface is high, OH− in the water replaces or competes with PO4 3− for the adsorption sites in hydroxides. Therefore, Fe(OH)3 or Al(OH)3, which is more stable, is formed, which in turn facilitates the release of some P. During the incubation experiment with core sediments from EH, the pH value of the water body was approximately 8. This pH value is conducive to the transformation of NaOH-rp (Wu et al. 2014).

Fig. 9
figure 9

Contents of IP and OP in the sediments before and after the incubation experiment

Full size image

The OP content of the sediments in EH varies greatly before and after the incubation experiment, with an average variation rate of 32.0 % (Fig. 9b). The variation amplitude is generally larger in the surface sediments than in the bottom sediments. This finding is attributable to the difference in the degradation rate of OP in sediments at different depths. OP consists of NaOH-nrp and a trace amount of NH4Cl-nrp (Psenner et al. 1988). The major sources of OP include polyphosphates, monophosphates, and inositol in the cells of microbes (Hupfer et al. 1995). During transport and transformation, OP is mineralized by microbes into highly active IP. High-activity OP can be directly released into the overlying water, which increases the concentration of OP in the overlying water (Hupfer et al. 2007).

Rydin (2000) studied the potential transport capability of sediment P in Lake Erken. The results showed that NH4Cl-P, BD-P, and NaOH-nrp in sediments have a high transport capability. These forms of P can be released from the sediments within a short period of time and enter the overlying water. The present study generated a similar result. The form of P has an important impact on the P release of sediments at different burial depths in EH (Fig. 10). The content of NH4Cl-P in sediments at 0–4 cm varies slightly. At depths of 4–10 cm, the content declines significantly; below 10 cm, the variation decreases to a certain extent (Fig. 10a). NH4Cl-rp has the same characteristics of vertical variation (Fig. 10c).

Fig. 10
figure 10

Vertical variation of P form in sediments before and after the incubation experiment

Full size image

NH4Cl-P is the most active P fraction in the sediments. It mainly includes soluble IP in pore-water and phosphate ions weakly adsorbed on the surface of sediments (Syers et al. 1973). Under the action of bioturbation, this P fraction is highly likely to be released, directly impacting the water quality of lakes. Existing studies have also proven that the NH4Cl-P content of sediments is closely related to the P concentration of the overlying water and the eutrophication of lakes (Boström 1984).

In our study, the NaOH-nrp and NH4Cl-nrp contents of the sediments at the sediment–water interface more significantly decrease at the end of the experiment than at the beginning (Fig. 10b, d). At the end of the experiment, the NaOH-nrp content decreases by 30.48 mg/kg. The NaOH-nrp of surface sediments (10 cm) changes more significantly than that at the bottom. This phenomenon is mainly caused by the stronger microbial activity of surface sediments than those at the bottom layer. The NaOH-nrp in sediments provides the main P source for microbes (Zhu et al. 2013). As a result, the degradation rate of NaOH-nrp in surface sediments is obviously faster than that at the bottom layer (Ahlgren et al. 2006). Liang et al. (2004) also indicated that the microbial activity of sediments at a surface layer of 9 cm in Hongfeng Lake is high. The content of organic matter decreases rapidly as a result of mineralization. Below 11 cm, the microbial activity weakens; hence, the mineralization rate of the organic matter slows down. After the incubation experiment in the present study, the NH4Cl-nrp content of sediments at different depths is basically lower than that before the incubation. The reason is that this form of P has a high activity and easily enters the overlying water in a dissolved form under bioturbation. Therefore, the release flux of DOP is directly influenced.

The influence of the form of P in sediments on P release was further analyzed. The relationships between the contents of various forms of P and the release flux of SRP and DOP were investigated. The results are shown in Table 2. The release flux of SRP of the sediments in EH is positively significantly correlated with the contents of NH4Cl-rp, NH4Cl-P, and HCl-P. A type of NH4Cl-P, NH4Cl-rp is an important factor that influences SRP release. The release flux of DOP is positively significantly correlated with NH4Cl-P, BD-P, and HCl-P. This result indicates that the release of NH4Cl-nrp is the driving force for the release of DOP and has no significant correlation with NaOH-nrp. The reason is probably that the most NaOH-nrp are indirectly released into the overlying water. The major component of NaOH-nrp in lake sediments is polyphosphate, which accounts for over 50 % of NaOH-nrp (Hupfer et al. 1995). Part of NaOH-nrp is transformed into orthophosphates because of the biomineralization and enzymatic hydrolysis of surface sediments under an anaerobic environment. Consequently, P is likely to bind to metal ions to form highly stable IP. For example, phytic acid can be strongly adsorbed by Fe(OOH) (De Groot and Van Wijck 1993; Hupfer et al. 2007; 2008).

Table 2 Correlation between P flux and the content of P fractions (n = 16)
Full size table

Influence of the P release of sediments on the water quality of EH and the eutrophication risk

According to the experimental results, the average release flux of TP at the sediment–water interface throughout EH is 1.13 ± 0.55 mg/(m2·d). On this basis, we estimate the release amount of TP from the sediments in EH to be 114.2 t/a. According to the data provided by the People’s Government of Dali Prefecture in China, the total external P loading of EH in 2010 was approximately 131 t, with the main forms of agricultural non-point source and inflow rivers. The internal P release of sediments may accordingly account for 87.2 % of the total external P loading entering EH. The P balance of EH was estimated to determine the influence of the P release of sediments on the water quality of EH. The results are shown in Fig. 11, which indicates that the TP remaining in EH every year is approximately 27.2 t. If this part of P enters the overlying water and affects the P concentration in the water body, then the contribution of the remaining P to TP concentration in the overlying water can be up to 9.4 µg/L. If the average TP concentration in EH in 2010–2013 was 0.03 mg/L, then the contribution of the remaining P to TP concentration in the overlying water was 31 %.

Fig. 11
figure 11

P balance of EH

Full size image

For different lake regions, a large variation in the release flux of TP exists at the sediment–water interface. A great difference also exists in its influence on the water quality for various lake regions. The TP release flux of sediments in various lake regions was estimated based on the release flux of SRP and DOP (Fig. 12a). The release flux of TP in the northern region is the largest, with a value of approximately 0.73 mg/(m2·a), followed by that in the southern region [0.49 mg/(m2·a)]. The release flux at the plateau of the center of the lake is the smallest, with a value of only 0.25 mg/(m2∙a). The northern region is the key region for the control of the external pollution of EH, the area of which accounts for only 23.3 % of the total lake area. However, the TP release amount of the sediments in the area accounts for 35.4 % of that in the entire lake. The northern and southern regions are much disturbed by human activities, which cause a large amount of P input. The major rivers that flow into EH, including the Miju, Yong’an, and Luoshi Rivers, are found in the north. The drainage basin of these rivers accounts for 72 % of the total drainage basin of EH. A large amount of domestic pollution and agricultural non-point source pollution enters EH in this region. The previous research study showed that the P load entering the lake in the northern region accounts for 56 % of the total load (Jiao et al. 2013). Located in the south of the drainage basin, Dali Prefecture is seriously affected by industrial and urban domestic pollution. Although non-point source pollution is effectively controlled by the wastewater interception project, such nutrients as N and P accumulate constantly in EH. Therefore, the influence of P release from the sediments in the north and south on the water quality of the overlying water is significantly greater than that in the middle and at the plateau of the center of the lake.

Fig. 12
figure 12

Spatial and temporal variations of the TP release flux in EH sediments

Full size image

Spatial and temporal variations of the P release of sediment

Given that the degree of pollution of a drainage basin varies with historical period, a great difference exists in the P load into a river and the general situation of P pollution. The TP release of sediments varies with time (Fig. 12b). At the temporal scale, the year 1970 was considered as the dividing point. The P release of sediments in EH can be divided into two stages: low-value stage and high-value stage.

The first stage refers to the time before 1970, when the release amount of P was maintained at a low level, with an average of approximately 79.0 ± 14.1 t/a. The P release in the northern and middle lake regions was obviously higher than that in the south and at the plateau of the center of the lake. Before 1970, the water quality of EH was good, with abundant aquatic resources (Pan et al. 1999). At this stage, the drainage basin of EH underwent a slow economic and social development. The pollution load into the lake could be basically self-purified by the lake. Moreover, the sediments adsorbed a large amount of nutrients. The nutrients entering the lake accumulated in the sediments. With a small P release at the sediment–water interface, the water quality of EH was minimally influenced.

The second stage spanned from 1970 to the present day, when the P release of sediments continues to be at a high level. The P release ranges from 105.2 to 128.3 t/a, with an average of 114.2.0 ± 8.9 t/a. In the past 40 years, the release amount of TP from the sediments of EH showed a slowly increasing trend. After 1970, the social and economic development accelerated. Land use and human life pattern have changed obviously. Traditional ways of farming have given way to the massive application of fertilizers and pesticides. A large amount of N and P flowed into EH along with the surface runoff. West Erhai River Hydropower Station was constructed and put into operation in 1977. Since then, EH has maintained a low water level of 1972.3 m (Du 1994). The large-area fish cage culture since the 1990s has damaged the ecological environment of the riparian zone. This phenomenon resulted in the reduction of the diversity of aquatic creatures (Zheng et al. 2004), impairing the self-purification capacity of the lake ecosystem. Since 1970, the total N and TP of EH have been increasing, which finally led to the first massive blue algal boom in September 1996 (Du 1997). The TP concentration has declined from 0.01 mg/L to class 0.026 mg/L at present. The accumulation of N and P in sediments has accelerated.

The P release behavior at the sediment–water interface is affected by the alternation of seasons. The P release amount from sediments in summer and autumn accounts for most of the annual TP release of EH. Therefore, summer and autumn are the peak seasons of P release from sediments. They are the key time for the control of the risk of P release from sediments. Given the monsoon climate in the Yunnan–Guizhou Plateau, the average water temperature of EH in summer and autumn is relatively higher than in the other seasons (i.e., the surface water temperature is approximately 23.7 °C in August), with a long sunshine duration. The overlying water is weakly alkaline (i.e., with a pH value ranging from 8.3 to 8.8), which is conducive to the growth and propagation of algae. The growth of aerobic algae consumes a large amount of DO in the water, thereby changing the microenvironment of the sediment–water interface. As the aerobic diffusion boundary layer at the sediment–water interface is destroyed, the P transport and transformation of the surface sediments accelerate. Consequently, the P release flux at the sediment–water interface increases.

Compared with other eutrophicated lakes in China, the TP release amount of EH is not among the highest (Lu et al. 2013). OP is the main form of endogenous P release. In existing studies, the influence of the OP release of sediments on water quality is usually ignored. If the pollution of EH is further aggravated, the pH value of the water body will continue to increase, and DO will decrease. The release of OP and IP will add to the risk of P release from sediments. The internal P loading from sediments will produce a great impact on the water quality of EH.

Conclusion

The release flux of DOP at the sediment–water interface of EH is 1.09 ± 0.46 mg/(m2·d). The release flux of SRP is generally low, with a value of 0.18 ± 0.11 mg/(m2·d). DOP is the main form of P release at the sediment–water interface in EH, with an amount of approximately 6 times that of SRP. DOP and SRP present similar release characteristics in the various regions of the lake, with the highest concentration in the north (EH1), followed by the south (EH4), the middle (EH2), and the plateau of the lake (EH3). At different burial depths, the release flux of DOP is high in the surface sediments and low in the bottom sediments, which greatly differs from SRP. Along the vertical direction, the release flux of DOP first increases and then decreases. A slight difference occurs in the characteristics of the release flux of SRP in different lake regions at the same burial depth. The peak values are found at depths of 4–6 cm in the northern and middle lake regions and at depths of 2–4 cm in the south and at the plateau of the lake. P form is an important factor that influences the P release of sediments at different burial depths in EH. NH4Cl-rp is the main source of SRP released from the sediments of EH. The transport and transformation of NH4Cl-nrp and NaOH-nrp are closely associated with the DOP release of the sediments of EH.

In this study, the release flux of TP at the sediment–water interface of EH is estimated at 114.2 t/a, which accounts for 87.2 % of the external P loading into the lake. A P of 27.2 t remains in the overlying water of EH every year. If this part of P directly influences the P concentration in the overlying water, then the P remaining in the overlying water will contribute 9.4 µg/L to the TP concentration of EH (31 %). The key area for the control of internal P pollution in EH is the northern region, and the key seasons for the control are summer and autumn. Significant attention should be given to the influence of the OP release of sediments on the water quality of EH. If the water pollution of EH is aggravated, with an increase in pH value and a decrease in DO, then the exchange of OP and IP at the sediment–water interface will enhance the risk of P release. This phenomenon will lead to the further deterioration of the water quality of EH.