Effects of pH on phosphorus form transformation in lake sediments Open Access

Phosphorus is an important factor that affects lake water function, as well as water ecosystem composition and structure, and it is one of the key nutrient elements in the eutrophication of lake water (Bao et al. 2017; Daneshgar et al. 2018; Yang et al. 2018; Ja et al. 2020). Internal P loading is the important factor in eutrophication, and the eutrophication state can be maintained for a long time (Yang et al. 2020). Phosphorus release from sediments is a key process that affects the effectiveness of eutrophication mitigation (Ma et al. 2021). The primary reason for the change in phosphorus content in sediments is the adsorption and release of phosphorus in sediments. In essence, a change in the external environmental conditions causes a change in the phosphorus form in sediments. It is speculated that a change in the phosphorus form is the primary reason for a change in the phosphorus content in water and sediments. In addition, a change in the phosphorus form can affect the migration and transformation of phosphorus through the sediment–water interface. Different forms of phosphorus have different activities and cause different effects in water (Ja et al. 2020). One form of phosphorus in sediments can be converted into another, affecting the bioavailability of phosphorus, as well as phosphorus release to the overlying water, which further influences the eutrophication degree of lake water.

Influences of the phosphorus exchange between sediments and the water column can be physical, chemical, and biological (Wu et al. 2014). pH is one of these factors, and it has been the subject of several studies (Bostrom 1988; Wu et al. 2014; He et al. 2017; Temporetti et al. 2019). Jin and colleagues found that, in alkaline conditions, P release increased with rising pH (Jin et al. 2006). Variations in water column pH had a significant impact on sediment mobile P, including OP (Kong et al. 2021). These studies were performed in laboratories, where a pH gradient was simulated, and the release amount was determined. However, the change of each form of phosphorus was not determined. The question of how a change in pH changes the form of phosphorus in sediments remains undetermined. Therefore, it is necessary to study the forms of phosphorus in sediments under different pH gradients. Wuliangsuhai Lake is located in a cold and arid region with obvious seasonal characteristics, resulting in significant changes in its external environmental conditions. These changes would inevitably change the forms of phosphorus, affect the transfer and conversion of phosphorus at the sediment–water interface, and thereby affect the phosphorus content in the lake.

The aim of this study is to evaluate the effect of pH change on the different phosphorus form contents in the sediments of the lake. Therefore, in this study, the phosphorus forms and the effects of pH on their fraction transformation are determined based on laboratory simulation experiments. The focus of this study is to explore the part of the phosphorus form that is affected by pH change, which results in a change in the phosphorus content in the sediments.

Wuliangsuhai Lake (40°36′ to 41°03′N, 108°43′ to 108°57′E) is located in Urad Qianqi, Bayandao City, in Inner Mongolia. The water area is 293 km², with an average lake elevation of 1,018.5 m, and the water depth is 1.1–2.8 m, with an average depth of 1.8 m. The annual average precipitation in Wuliangsuhai Lake is 224 mm, and the annual evaporation is 2,456 mm. The annual sunshine duration is 3,185.5 h, and the daily illumination rate can reach 72%. The annual average temperature is 5.8 °C, the extreme maximum temperature is 37.7 °C, and the extreme minimum temperature is −30.8 °C. The average frost-free period over the years is approximately 152 days. The lake freezes in early November and melts in early April of the following year. The frozen period typically lasts 5–6 months, and the thickness of the ice is approximately 0.3–0.6 m (Liu et al. 2017; Tian et al. 2020).

In this study, in total 10 sampling sites were established in Wuliangsuhai Lake (Figure 1). The site locations were determined using a geographic positioning system (GPS). Surface sediment samples were collected on June 27, 2016 using a column sampler.

Water samples were collected from each site with a plastic bottle (10 L). The pH, dissolved oxygen, total dissolved solids (TDS), water temperature, and conductivity were measured using a YSI Professional Plus instrument in situ.

All of the surface sediment samples were placed into clean sealed bags, and the air in the bags was drained. The samples were stored in the dark at low temperature and then brought back to the laboratory for analysis.

The sediment samples from each same sampling point were divided into fresh and air-dried soil samples and then treated separately. The air-dried soil samples were naturally dried in a cool and ventilated place at room temperature. They were then crushed with wooden sticks to remove impurities such as large and small stones, animal residues, and plant roots. After grinding, the substances were put through 100 mesh sieves, mixed well, and bagged for the index determination.

For the determination of pH of the sediments, 10 g of sediment samples that were air-dried and ground through a 100-mesh sieve (accurate to 0.0001 g) were placed in a 50-ml beaker, and 25 ml deionized water was added (ratio of soil to water was 1:2.5). This was mixed and stirred for 1 min using a glass rod until the soil particles were well dispersed, and left to sit for 30 min. An electrode with a pH meter was then inserted into the sample suspension (at the mud–water interface) to measure the pH value. The method was carried out according to the China national standard NY-T 1377–2007.

Several fresh sediment samples were collected and dried at 108 °C, and the moisture content was calculated using the mass difference method.

A total of 0.3 g of the air-dried soil sample was run through a 100-mesh sieve, and 5 ml nitric acid, 2 ml sulfuric acid, and 1 ml hydrogen peroxide were added for digestion. The supernatant was absorbed and placed in a volumetric flask into which one drop of dinitrophenol indicator was added, a 4 mol/L NaOH solution, as drops until the solution turned yellow. Then, one drop of a 2 mol/L sulfuric acid solution was added to make the yellow color of the solution disappear. Then, a molybdenum antimony sulfate reagent was added, and ultra-pure water was used to obtain a constant volume of 50 ml. After 30 min, a wavelength of 660 nm was used for colorimetry on a UV-Vis light spectrophotometer, and the absorbance value of the measured solution was recorded. According to the standard curve, the phosphorus content of the solution tested was obtained, and then the total phosphorus content of the sediment was calculated.

The different phosphorus form contents were determined using the sequential extraction scheme according to Psenner (1988). The extraction process separated the forms of phosphorus in the sediment into weakly adsorbed phosphorus (NH₄Cl-P), iron-manganese-bound phosphorus (Fe/Mn-P), iron-aluminum-bound phosphorus (Fe/Al-P), calcium-bound phosphorus (Ca-P), and residual phosphorus (Res-P). The continuous extraction steps were as follows:

Step 1. NH₄Cl-P was extracted. A dried and ground soil sample weighing 0.25 g was placed into a 50-ml centrifuge tube, and 25 ml of NH₄Cl (l-mol/L) was added. This was oscillated for 1 h (220 rpm, 25 °C) at a constant temperature and then centrifuged for 10 min (10,000 rpm). The supernatant was removed and extracted again using the same procedure, and this was combined with the supernatant. The content of the dissolved total phosphorus in the extract was determined after being filtered using a 0.45-μm filter membrane.

Step 2. The residue was cleaned, and 25 ml of deionized water was added. This was oscillated for 10 min (220 rpm, 25 °C), centrifuged for 5 min (10,000 rpm), and the liquid was removed.

Step 3. Fe/Mn-P was extracted. A 25 ml solution of 0.11 mol/L NAHCO₃ and 0.11 mol/L safety powder–sodium disulfite was added. This was oscillated for 0.5 h (220 rpm, 40 °C), centrifuged for 10 min (10,000 rpm), and then the supernatant was filtered using a 0.45-μm filtration membrane to obtain the extract. Step 2 was repeated to clean the residue. In order to eliminate the interference of the insurance powder, 5–10 ml of the extract at a constant volume was collected, and 1.6 ml of 5% potassium persulfate was added. The test liquid was then cooled to room temperature. The content of the dissolved total phosphorus in the extract was determined after being filtered using a 0.45-μm filter membrane.

Step 4. Cleaning of the residue was performed by adding 25 ml deionized water, oscillating for 10 min (220 rpm, 25 °C), and centrifuging for 5 min (10,000 rpm); the liquid was then removed.

Step 5. Fe/Al-P was extracted. First, 25 mL of NaOH (1 mol/L) was added, and this was oscillated for 16 h (220 rpm, 25 °C), centrifuged for 10 min (10,000 rpm), and the supernatant was collected. The content of the dissolved total phosphorus in the extract was determined after being filtered using a 0.45-μm filter membrane.

Step 6. Cleaning of the residue was performed by adding 25 ml of deionized water, oscillating for 10 min (220 rpm, 25 °C), and centrifuging for 5 min (10,000 rpm); the liquid was then removed.

Step 7. Ca-P was extracted by adding 25 ml of HCl (0.5 mol/L), oscillating for 16 h (220 rpm, 25 °C), and centrifuging for 10 min (10,000 rpm); then, the supernatant was collected. The content of the dissolved total phosphorus in the extract was determined after being filtered using a 0.45-μm filter membrane.

Step 8. Cleaning of the residue was performed by adding 25 ml of deionized water, oscillating for 10 min (220 rpm, 25 °C), and centrifuging for 5 min (10,000 rpm); then, the liquid was removed.

Step 9. Res-P was extracted by adding 25 mL of NaOH (1 mol/L), oscillating for 24 h (220 rpm, 85 °C), and centrifuging for 10 min (10,000 rpm); then, the supernatant was collected. The content of the dissolved total phosphorus in the extract was determined after being filtered using a 0.45-μm filter membrane.

Point J11, located in the lake center area, was found to accumulate a high phosphorus content in the sediments after long-term monitoring and was chosen as the research object (TP: 1,790.07 mg/kg). The effects of different pH conditions on the phosphorus release in the surface sediments of Wuliangsuhai Lake were studied. The water utilized for the experiment was the overlying water at the origin point after filtering through a 0.45-μm filter (TP: 0.0466 mg/L; pH: 7.65). The specific experiment was as follows.

In total, 20 g of dried surface sediment samples were weighed and placed at the bottom of three 500-ml beakers. The beakers were then slowly injected with 400 ml of experimental water with a modulated pH. The pH value of the beakers was regulated using l mol/L HCl and 1 mol/L NaOH (Wang et al. 2009).

According to the pH monitoring data of Wuliangsuhai Lake water from 2002 to 2016, the water pH was between 4.0 and 10.8, and the mean values of pH during winter, spring, summer, and autumn were 8.7, 9.0, 8.2, and 8.4, respectively (Figure 2). The pH value indicated that the water environment of Wuliangsuhai Lake was neutral–alkaline. Hence, the static simulation experiments of phosphorus release in sediments were determined under the pH conditions of pH = 4.0, pH = 7.0, and pH = 10.0. pH reconditioning was conducted daily. The liquid level scale was recorded, and 50 mL of each water sample were collected with a syringe at intervals of 6, 12, 24, 48, 72, 96, 120, 168, 216, 264, 312, and 360 h. The experiment lasted for 15 days, and in total 12 samples were collected. Three parallel experiments were performed under each pH condition for a total of 108 samples. In addition, a filtered sample water from the original sampling point was added to the original scale and cultured in a 20 °C incubator. The water samples were then collected, centrifuged, and passed through a 0.45-μm filter to determine the TP (Jin et al. 2004). After the experiment, the phosphorus forms in the sediments were determined following the same procedures described above. The control is the original sample without pH adjustment.

The potential bioavailable phosphorus can contribute significantly to local primary production when this fraction enters the water column. As such, an assessment of a water ecological ecosystem should recognize not only the phosphorus content but also the phosphorus forms.

The TP content in the surface sediments of Wuliangsuhai Lake ranged from 840.1 to 2,709.0 mg/kg, with an average of 1,656.7 mg/kg and a coefficient of variation of 0.2 as shown in Figure 3. Phosphorus in sediments is most closely related to benthic organisms and has an important impact on their habitats. If the content exceeds a certain level, it will exert toxic effects on benthic organisms. At present, experts and scholars refer to ‘the Ministry of Environment and Energy of Ontario, Canada’ for their research. The Canadian baseline values focus on the safety of sediments that have no toxic effects on benthic organisms. In China, there is one standard, which is ‘the thresholds of TN and TP reference values in sediments of typical lakes in eastern China’, the standard is mainly aimed at typical lakes in eastern China. So, according to the guidelines for a sediment quality assessment based on the ecotoxicity effect issued by the Ministry of Environment and Energy of Ontario, Canada, the TP content that generates the lowest ecotoxicity in sediments is 600 mg/kg, and the content that generates severe ecotoxicity is 2,000 mg/kg (Mudroch & Azcue 1995). According to this standard, the TP content of all the sampling points in the surface sediments of Wuliangsuhai Lake was higher than 600 mg/kg, and the TP content value of 20% of the sampling points was higher than 2,000 mg/kg. This indicated that the TP in most of the surface sediments of Wuliangsuhai Lake might have exerted certain ecotoxicity effects on benthic organisms, and some areas might had caused serious ecotoxicity effects. This poses a great threat to the benthic communities and the ecological environment of the Wuliangsuhai Lake area.

The content of NH₄Cl-P in the surface sediments of Wuliangsuhai Lake was very low, and it ranged from 15.7 to 31.1 mg/kg (Table 1 and Figure 4), with an average of 23.9 mg/kg and a coefficient of variation of 0.4 (Table 1). This accounted for 0.7%–3.6% of the TP and 1.7% of the average percentage of TP. The content of Fe/Mn-P in the surface sediments of Wuliangsuhai Lake ranged from 125.3 to 193.9 mg/kg, with an average of 161.2 mg/kg and a coefficient of variation of 0.1 (Table 1 and Figure 4). Reduction–oxidation reaction (redox)-sensitive phosphorus is very sensitive to the redox potential (Ruban et al. 2001). The content of Fe/Al-P in the surface sediments of Wuliangsuhai Lake ranged from 30.1 to 88.1 mg/kg, with an average of 64.0 mg/kg and a coefficient of variation of 0.2 (Table 1 and Figure 4). When the sediment–water interface is anaerobic, Fe/Al-P dissolves and is released into the overlying water. It is then absorbed and utilized by algae and other aquatic plants (Hao et al. 2019). The content of Ca-P in the surface sediments of Wuliangsuhai Lake ranged from 249.1 to 1,498.3 mg/kg, with an average of 944.8 mg/kg and a coefficient of variation of 0.2 (Table 1 and Figure 4). The content of Res-P in the surface sediments of Wuliangsuhai Lake ranged from 283.1 to 940.0 mg/kg, with an average of 462.7 mg/kg and a coefficient of variation of 0.2 (Table 1 and Figure 4). Res-P was the most stable among all the forms of phosphorus in the sediments (Kaiserli et al. 2002; Lane & Autrey 2016; Parsons et al. 2017).

There were large differences in the contents of inorganic phosphorus components in the surface sediments of Wuliangsuhai Lake. The component distribution order was: Ca-P (54.2%) > Res-P (28.7%) > Fe/Mn-P (11.0%) > Fe/Al-P (4.2%) > NH₄Cl-P (1.6%) (Figure 4). Therefore, the primary form of phosphorus in the sediments of Wuliangsuhai Lake was Ca-P.

It was more appropriate to estimate the phosphorus contents derived from the loading of the sediments by the contents of the BAP rather than the TP in the sediments. The single sample non-parametric K-S test showed that the data listed in Table 2 were normal. The BAP content in the sediments of Wuliangsuhai Lake ranged from 179.1 to 307.8 mg/kg, with an average of 247.9 mg/kg, accounting for 9.9%–36.6% of the TP (Figure 5). It can be seen that the phosphorus in the surface sediments of Wuliangsuhai Lake had a certain transfer ability. This was primarily composed of Fe/Mn-P, accounting for 60.2%–70.9% of the BAP (Table 2). Fe/Mn-P is most closely related to the process of phosphorus transfer and conversion in lake water environments. This potentially active phosphorus has an important relationship with loading in lakes.

Under the neutral condition of pH = 7, the concentration of TP in the overlying water was the lowest. After 24 h of the release experiment, the TP concentration in the overlying water reached a maximum, with a value of 0.17 mg/L (Figure 6(a)). Under the acidic condition of pH = 4, the TP concentration in the overlying water was slightly higher than in the neutral condition. After 72 h of the release experiment, the TP concentration in the overlying water reached a maximum of 0.45 mg/L (Figure 6(a)). Under the alkaline condition of pH = 10, the TP concentration in the overlying water was the highest. After 48 h of the release experiment, the TP concentration in the overlying water reached a maximum of 0.78 mg/L, and the maximum value appeared one day earlier than that under acidic conditions (Figure 6(a)). The values in Figure 6 are the average of the three samples.

An increase in the pH of the overlying water is beneficial for the release of phosphorus in sediments, and this is primarily because the pH of the overlying water directly affects the existence modes of phosphate. When the overlying water is weakly acidic, the form of phosphorus is primarily HPO₄²⁻, and when the former is weakly alkaline, the latter is primarily H₂PO₄⁻. The adsorption was the largest in the second condition, and the magnesium salt, silicate, alum inosilicate, ferric hydroxide colloids, and other molecules in the sediments were all involved in the adsorption process. An increase in pH could promote the desorption of phosphate radical ions from the ferric hydroxide colloid and release more phosphate into water, which will enhance the release of phosphorus from sediments. This would result in an increase in the phosphorus content in the overlying water. Therefore, a pH increase favors the release of phosphorus in sediments.

The release content is equal to the total phosphorus content at each sampling time point minus the total phosphorus content in the initial water sample. These observations suggest that phosphorus release from the sediments occurred in both acidic and alkaline conditions, and the alkaline condition was more favorable. However, the least amount of phosphorus was released under the neutral condition (Figure 6(b)).

The proportion of NH₄Cl-P in the TP increased with an increase in the pH. NH₄Cl-P accounted for 0.4%, 0.6% and 0.5% of the TP when the pH value was 4.0, 7.0 and 10.0 respectively. (Figure 7). The content of NH₄Cl-P in the original sediment sample was 21.6 mg/kg. After the release experiment, its content in the sediment sample was 3.2, 6.7, and 5.3 mg/kg under the conditions of pH = 4, pH = 7, and pH = 10, respectively (Figure 8). The NH₄Cl-P in the sediments was released and converted into a portion of the soluble phosphorus in the overlying water. The weakly adsorbed phosphorus (NH₄Cl-P) primarily included loose adsorbed phosphorus on the surface of the sediment minerals, dissolved active phosphorus in the interstitial water, and phosphorus released from dead biological cell residues containing carbon (Zhou et al. 2005).

The content changes of Fe/Mn-P and Fe/Al-P were similar under the different pH conditions. The proportion of Fe/Mn-P and Fe/Al-P in the TP decreased with an increase in pH. When the pH value was 4.0, Fe/Mn-P and Fe/Al-P accounted for 25.5% and 7.69% of the TP, respectively. When the pH value was 7.0, it accounted for 14.8% and 4.8% of the TP, respectively. When the pH value was 10.0, it accounted for 9.8% and 3.1% of the TP, respectively (Figure 7). The Fe/Mn-P in the original sediment samples was 152.7 mg/kg, and the Fe/Al-P was 54.6 mg/kg. After the release experiment, under the condition of pH = 4, the Fe/Mn-P and Fe/Al-P increased significantly and were 206.4 and 62.1 mg/kg, respectively (Figure 9). Due to an increase in H⁺, Fe³⁺, and A1³⁺ combined with PO₄³⁻, HPO₄²⁻, and H₂PO₄⁻ in the water, they were re-adsorbed into the sediments. Under the condition of pH = 7, the Fe/Mn-P and Fe/Al-P were slightly reduced. This was because in a neutral environment, the primary forms of phosphate are HPO₄²⁻ and H₂PO₄⁻, which are easily combined with some metal elements and then adsorbed in sediments, and thus the amount of release is minimal. Under the condition of pH = 10, Fe/Mn-P and Fe/Al-P decreased significantly by 95.6 and 30.1 mg/kg, respectively, before and after release. Under alkaline conditions, phosphate is primarily ion exchanged, that is, OH⁻ can be exchanged with a phosphate radical in Fe³⁺ and Al³⁺ cements, resulting in a large solution of iron- and aluminum-bound phosphorus in sediments. Phosphorus combined with Fe³⁺ and Al³⁺ was re-released into the water due to ion exchange. Fe/Mn-P plays an important role in the fixation and release of phosphorus at the sediment–water interface of the lake and is a significant component of potentially active phosphorus in sediments. Fe/Mn-P and Fe/Al-P are sensitive to pH. Under high pH conditions, the adsorbed phosphorus combined with amorphous or weakly crystalline iron and aluminum salts, such as ferrohydrate (FeOOH), blue iron ore (Fe₃(PO₄)₂·(H₂O)₈), diaspore (Al₂O₃·H₂O), and other molecules begin to be released to the overlying water continually. Under alkaline conditions, the surface of the variable charge body in the sediment carries a negative charge, which reduces the absorbability of H₂PO₄⁻ in water. In addition, OH⁻ exists in large quantities in water and competes with the anions in the colloid in the sediment. Additionally, it exchanges with H₂PO₄⁻ in the Fe/Al-P and increases the phosphorus content in the overlying water. This is consistent with the research results of Huang et al. (2003) and Jin et al. (2004).

The proportion of Ca-P in the TP increased with an increase in pH. The Ca-P content was 215.5 mg/kg, and Ca-P accounted for 26.7% of the TP in the sediment when the pH value was 4.0. The Ca-P accounted for 47.7% of the TP in the sediment when the pH value was 7.0. The Ca-P accounted for 53.7% of the TP in the sediment when pH value was 10.0. Under the alkaline condition of pH = 10, the Ca-P in the sediment samples increased significantly after the release experiment (Figure 9). The ratio of Ca-P to TP in the sediment under a low pH condition was lower than that under a high pH condition, indicating that acidic conditions were more conducive to the release of Ca-P in the sediment than neutral and alkaline conditions. Ca-P is highly stable in sediments and difficult to be dissolved. It is susceptible to low pH (Xin et al. 2020). Under acidic conditions, H⁺ has a dissolution effect on Ca-P, which promotes the release of Ca-P in the sediment, resulting in an increase in H₂PO₄⁻ in the water. This is consistent with the research results of Jin et al. (2004, 2006). Under alkaline conditions, Ca-P would produce deposits with certain substances in the sediments, causing an increase in its content. The release of some Fe/Mn-P in sediments might precipitate again and be converted to Ca-P, thereby increasing the amount of Ca-P.

Res-P is the most stable form of phosphorus, and it is not sensitive to pH changes and has difficulty participating in a short-time phosphorus cycle. The residual phosphorus content remains basically the same under the different pH conditions before and after the release experiment (Figure 10). In this experiment, one-way analysis of variance (ANOVA) was used to analyze the significance of Res-P content at different pH. It is found that all the data satisfy the normal distribution by using the Shapiro-Wilk normality test. According to the homogeneity of variance test, the significance of Res-P was all greater than 0.05, and the variance was homogeneity. Therefore, multiple comparison was used to analyze the significance of each treatment, and there was no significant difference in Res-P content among all groups (P > 0.05). Before the release experiment, the Res-P is 318.1 mg/kg, and after the release experiment the Res-P is 320.4, 321.4 and 319.8 mg/kg. Relatively speaking, after the release experiment, the residual phosphorus content in the sediment samples slightly increased compared with the original.

In general, pH was a factor that affected the release of phosphorus in sediments. Under neutral conditions, the phosphorus in the sediments was stable and not easily released. Acidic or alkaline conditions were favorable for the release of phosphorus in the sediments, with a larger release amount under alkaline conditions. Acidic conditions were conducive to the dissolution and release of Ca-P, and alkaline conditions were more conducive to the release of Fe/Mn-P and Fe/Al-P. As an important bioavailable phosphorus source in the surface sediments of Wuliangsuhai Lake, Fe/Mn-P accounted for the largest proportion. During the ice-free period, the water in Wuliangsuhai Lake exhibited weak alkaline characteristics. This occurs particularly from June to September, when algae and other phytoplankton grow in large quantities through intense photosynthesis, causing an increase in the lake water pH (Quan et al. 2019). This is beneficial to the release of iron–manganese chelated phosphorus in sediments to the overlying body, leading to an increase in the phosphorus content of the water. In addition, the phosphorus form is also affected by water redox characteristics under alkaline conditions. Fe³⁺ is lower and dissolved under a reduced oxidation–reduction potential, and phosphorus is released into the interstitial water. When the sediment–water interface is under anaerobic conditions, Fe/Al-P dissolves and is released into the overlying water. It is then taken up by algae and other aquatic plants. Located in the Mongolian Plateau Lakes, Wuliangsuhai Lake has obvious seasonal characteristics and undergoes an alternating process of ice-free season and an icebound season. Winter is a long period of the annual cycle for many lakes in the northern hemisphere. Low irradiance, ice, and snow cover lead to poor light penetration into the water column of these lakes. Therefore, in northern lakes, respiration often exceeds primary production leading to low dissolved oxygen concentrations (Song et al. 2019). The content of Fe/Mn-P and Fe/Al-P in the sediments of Wuliangsuhai Lake was 9.91%–28.56%. When the redox potential of Wuliangsuhai Lake changed, that is, during the icebound season, these two types of phosphorus could dissolve and be released into the overlying water, resulting in increases in the TP content in the water.

The TP content in surface sediments of Wuliangsuhai Lake ranged from 840.16 to 2,709.03 mg/kg, with an average of 1,656.75 mg/kg. The order of the different inorganic phosphorus in the sediments was: Ca-P > Res-P > Fe/Mn-P > Fe/Al-P > NH₄Cl-P. The BAP in surface sediments accounted for 9.99–36.64%.

The pH alters the distribution of phosphorus types in sediments (NH₄Cl-P, Fe/Mn-P, Fe/Al-P, and Ca-P) and, hence, influences the content of phosphorus in lake water. Phosphorus release from the sediments occurred under both acidic and alkaline conditions, and the alkaline condition was more favorable. However, the least amount of phosphorus was released under the neutral condition. The effect of pH on phosphorus release was primarily reflected by the existing forms of phosphorus. A low pH promoted the release of Ca-P, and a high pH promoted the release of Fe/Mn-P and Fe/Al-P. Fe/Mn-P was an important bioavailable phosphorus in the surface sediments of Wuliangsuhai Lake, accounting for the largest proportion of easily transferable and converted phosphorus.

Wuliangsuhai Lake is characterized by weak alkaline characteristics throughout the year with high pH values. The iron–manganese chelated phosphorus in the sediments was found to be easily released into the overlying water. Therefore, water pH should be more accurately monitored and it is necessary to control the pH value of the water during the environmental treatment of Wuliangsuhai Lake. In addition, the changes of phosphorus form should be analyzed from the important factors affecting the release of phosphorus. Compared with subtropical and tropical lakes, lakes in cold regions have thicker ice sheets during the freezing period, which block the material exchange between the atmosphere and water. The existence of ice sheet changes the transmission of light to water and affects the light field environment of water. The organic matter content, salinity, eutrophication and other physical and chemical properties of the subglacial water changed greatly, especially the dissolved oxygen content of the bottom water. Due to the barrier of the ice sheet, the re-oxygenation process formed by natural aeration changes greatly, and the concentration of dissolved oxygen is at the lowest value. In particular, when snowfall occurs during the frozen period, the content of dissolved oxygen will be zero (extreme anaerobic conditions) (Song et al. 2019). The change of dissolved oxygen must have a great influence on the occurrence form of phosphorus, so we should analyze how the change of dissolved oxygen affects the occurrence form of phosphorus in the future.

The work was supported by the National Natural Science Foundation of China (52060022; 51779118; 51869020; 51909123). Ministry of Science and Technology of the Peoples’ Republic of China (2017YFE0114800); The Inner Mongolia Autonomous Region Science and Technology Department (2019MS05032); The National Key Research and Development Program of China (2019YFC0409201;2019YFC0409204).

All relevant data are included in the paper or its Supplementary Information.