The distribution of phosphorus in water, sediment and vegetation was influenced by many parameters. Previous studies have reported widely varying phosphorus removal efficiencies (from 14–97%) in constructed wetlands (Zhang et al. 2014; Almuktar et al. 2018; Cooper et al. 2020). In the present study, the removal efficiency of TP in the water was located in this range (Figs. 2 and 3). Phosphate is predominantly removed through assimilation into aquatic organisms and through co-precipitation with Fe, Al and Ca, a process commonly thought to decline in efficiency with wetland age as sediment sorption sites for phosphorus become saturated over time (Vymazal 2007; Almuktar et al. 2018). In the same wetland, the removal efficiencies of TP and TDP in the overlying water of reed and cattail community area were decreased from May in 2013 (Tian et al. 2017) to May in 2016 (this study). From 2013 to 2016, the removal efficiency of SDP in the overlying water of reed community area was increased while that in the cattail community area was decreased. According to the previous study, the high phosphate removal efficiencies recorded at the River Mun ICW after five years of operation indicated the wetland remained in good operational condition (Cooper et al. 2020). Thus, along with the age increase of wetland, the removal efficiency may be maintained by proper macrophyte selection and management.
The spatial distribution of TP in constructed wetlands was closely related to spatial variation of pH and ORP level, which were greatly influenced by oxygen transfer, pollutant distribution and microbial activity within the system (Yu et al. 2015). pH in water was suggested to affect the mineralization rate of sediment, and acidic or alkaline water can promote phosphorus exchange in water-sediment phases (Tang et al. 2018). In the present study, pH played an important in affecting the phosphorus concentrations and fractions, especially in the reed community area (Table 2). However, the release of phosphorus from sediment to water in the reed community area was not easier than the open water area cattail community area as indicated by the diffusion flux and ERI of reed community area (Figs. 10 and 11). Therefore, pH may not be the most important factors influencing the release of phosphorus from sediment to water. On the other hand, sedimentation of particulate OM, oxygen depletion on the sediment surface, and the activities of anaerobic microbes in sediments all resulted in the production of reducing substances that decrease the ORP of sediments (Li et al. 2007). The products of reduction reactions, such as organic anions, ammonium, and elemental sulfur, occur on the surface of sediments to hinder phosphorus from approaching the adsorption sites of the sediments. This causes phosphorus originally adsorbed to the sediments to be released to the overlying water column (Li et al. 2007). Besides, oxidation of sediments can decrease the level of phosphorus in overlying waters (Li et al. 2007). In a range of − 400 to 200 mV, raising the ORP of sediments was reported to decrease the TP and SRP in interstitial water (Li et al. 2007). This result was not in line with the present study. In this study, the ORP in the surface, overlying and interstitial water was not correlated to any phosphorus fractions (Table 2).
The TP in sediments of open water area at transect S7 (857.80 mg/kg) were lower than the concentrations at transect S1 (1234.74 mg/kg), while opposite in the macrophytes covered area. This result was in line with the variation of TP in the surface water and overlying water. It manifested that the increase of the TP at the effluent in the open water were caused by the release of phosphorus in sediments while the decrease of TP in the macrophytes covered area were attributed to the storage of phosphorus in sediments. Therefore, the presence of macrophytes contributed to the storage and removal of phosphorus from water.
The percentages of HCl-P and Org-P in TPsediments were relative stable along with the flow. Compare with the results of the same wetland in 2013 (Zhang et al., 2016), it was found that the average ratios of HCl-P/TP in sediments decreased while the ratios of Org-P/TP in sediments increased in the past three years. In 2013, the percentages of calcium-bound phosphorus in TP in sediments were 70.25 ± 5.10%, 74.14 ± 6.09% and 71.56 ± 3.00% for the open water area, reed community area and cattail community area, respectively. In 2016, the percentages decreased to 63.09 ± 4.22%, 66.51 ± 5.61% and 68.03 ± 3.16% for the open water area, reed community area and cattail community area, respectively. Moreover, from year 2013 to 2016, the percentages of Org-P in TP in sediments increased from 10.45 ± 2.25%, 15.16 ± 5.56% and 13.48 ± 1.77% to 16.93 ± 5.73%, 19.39 ± 5.41% and 15.83 ± 2.45% for the open water area, reed community area and cattail community area, respectively. It was because that a new sediment layer was continuously formed, and new detritus often contained a great amount of organic matter, which led to high organic matter in the surface layer of sediment, especially in the plant area (Huang et al. 2015). On the other hand, the increase range of Org-P/TP ratios was highest in the open water area (6.48%), followed by the reed community area (4.23%) and then the cattail community area (2.36%). The possible reason for this result was that the continuous accumulation of detritus resulted in the higher aboveground biomass as shown in Fig. 8. Besides the detritus accumulation, the near-bed flow velocity was reduced in dense patches of plants; thus, fine suspended sediment, often contained high concentrations of OM, nitrogen and phosphorus, was easily deposited (Huang et al. 2015).
The ratios of NaOH-P/TP in sediments decreased along with the flow (Fig. S4). According to the Pearson correlation analysis, the ratio of NaOH-P/TP in reed community area was positively correlated to the fresh weight and stem diameters at significance level of 5%. In cattail community area, NaOH-P was positively correlated to the stem length of cattail. For other phosphorus fractions or morphological characteristics of reed and cattail, there was no significant relationships. Therefore, the decrease of NaOH-P, i.e. iron/aluminum-bound phosphorus in TPsediments was accompanied with the reduction of reed biomass and the stem length of cattail. Previous research indicated that iron/aluminum-bound phosphorus supported growth of phytoplankton, and could also be used for the evaluation of algal available phosphorus (Zhou et al. 2001). It was exchangeable with OH − and other inorganic phosphorus compounds, which were soluble in bases (Kozerski and Kleeberg 1998; Tuszynska et al. 2013; Chen et al. 2015). Iron-bound phosphorus was particularly sensitive to the ORP, pH, DO and salinity (Jin et al. 2006b, a; Yao et al. 2006). This fraction of phosphorus can be either consumed in bio-utilization or combined with other ions in water (such as Ca2+ and Al3+). In the present study, for both the overlying water and the interstitial water, there was no significant difference in ORP, pH, DO or C between the macrophytes free area and macrophytes covered area. Therefore, the reason caused the variation of NaOH-P in the sediments need further research.
OM played an important role in sediments, serving as electron donors (Verdouw and Dekkers, 1980), and their mineralization resulted in changes in oxidation-reduction potential and pH, which would then lead to the variation of PFs. The OM percentages in this study were relative lower than a pervious study in lakes of Beijing (Liu et al. 2009) which should be due to the type of the wetland and the characteristics of inflow wastewater. In the present study, the studied wetland was constructed from a collapsed mine. The inflow water was originated from a river. However, the lakes in their study were encompassed or partly adjacent to landscape lawns fed with manure and chemical P fertilizers (Liu et al. 2009). Therefore, the OM contents were relative lower than those lakes. It was previously studied that phosphorus compounds were important component of OM in lake sediments (Li et al. 2010), and OM was closely associated with biogeochemical cycles of phosphorus (especially Org-P) in sediments (Ingall and Van Cappellen, 1990; Ruiz-Fernández et al., 2002; Zhang et al., 2008). In the previous study, OM percentages correlated with the sum of NH4Cl-P and BD-P, which would be readily released under anoxia (Liu et al. 2009). As shown in Table 3, OM contents were positively related to the TP in sediments in the absence or the presence of macrophytes. This result was different from an early experiment about adding OM into lake sediments (Wang et al., 2012). They suggested that there was also a decrease in all the different phosphorus fractions when OM was added to the sediment. That was because addition of OM resulted in increased rate of mineralization and phosphorus release (Azzoni et al., 2001), causing decreased TP in sediment (Wang et al., 2012). However, Org-P, NaOH-P or HCl-P was not related to OM in all the three studied area in this study. In the sediment of Tuojiang River (pH > 8), the OM was positively correlated with all the phosphorus fractions (except for Fe-P and De-P) (Liu et al. 2022).
Alkaline phosphatase plays an important role in phosphorus supply, and has often been studied as one of the important factors for releasing phosphate from organic phosphorus compounds in aqueous systems (Zhang et al., 2007; Wang et al., 2012). It can hydrolyze a variety of organic phosphorus compounds into orthophosphates under phosphate depletion conditions (Pettersson, 1980; Jin et al., 2006; Wang et al., 2012). With the activity of bacteria, in particular organic phosphorus mineralizing bacteria (OPB) (Yang et al. 2021), the APA increased in the sediment and the redox conditions in the sediment may turn from the oxic to anoxic simultaneously, which may further influence the Fe/Al-P fractions (Zhang et al., 2007). APA was found to be negatively correlated to the Org-P in in the sediments of the open water area, TP and Org-P in the sediments of reed community area (Table 3). The results indicated that APA was related to the mobile PFs, and the reduction of APA synthesis would be important in controlling phosphorus release.
Wetland plants impacted the phosphorus removal by uptake, root exudate, rhizosphere microorganisms and microenvironment etc. (Zhang et al. 2014; Mesquita et al. 2018). The role of plants on phosphorus removal was influenced by many factors such as species, nutrient loading rates, the season of the year and so on (Zhao et al. 2010). Plant uptake despite being a minor route for phosphorus removal also reached a maximum in spring-summer months (Mesquita et al. 2018). It was suggested that plant biomass had no significant correlation with removal efficiency (Yang et al. 2007). However, in this study, the biomass of reed was correlated to the TPsediment and TPoverlying (Table 4). Previous research indicated that plant biomass had a retention capacity of 25% of the phosphorus added in constructed wetlands (Silvan et al. 2004). Furthermore, the combination of the role of plants in oxygen transfer to the rhizosphere and water level fluctuations caused by higher evapotranspiration due to higher temperatures during spring-summer period could have contributed to enhance the potential redox which in turn seems to favor the increase of phosphorus removal by sorption and co-precipitation on elements minerals in the rhizosphere (Mesquita et al. 2018). In the previous study, reed was proved to be efficient in the prevention of contaminant release, including N and P (Dong 2013). The presence of reeds contributed mainly to increase the percentage of phosphorus retained in the Pi-Fe/Mn/Al fraction in the treatments with high nutrient level water (from ~ 35 to ~ 47%), which could be attributable to the presence of an aerenchyma system in reeds (Armstrong et al. 2000), that promotes oxygen and carbon dioxide fluxes throughout the rhizomes (Tercero et al. 2017). A more oxidant environment could have been favored in the rhizosphere of reed facilitating phosphorus accumulation in metal oxides. Huang et al. (2015) also found higher phosphorus concentrations bound onto Fe/Al compounds in soils of reed than in bare ground (Huang et al. 2015). On the other hand, reed was also reported to have a minor role in phosphorus removal (Tercero et al. 2017). The presence of reed probably had a more relevant role in phosphorus mobilization between the solid and soluble phases (Tercero et al. 2017). It is coincident with the present study as reed community area got lower diffusion flux and ERI than cattail community area indicating the lower risk of phosphorus.
In the present study, the phosphorus concentrations in the surface, overlying and interstitial water of cattail community area were averagely lower than that in the reed community area. However, the removal efficiencies of phosphorus in the surface and overlying water of cattail community area were lower than the reed community area. Meanwhile, there was no significant difference in the phosphorus concentrations of sediments and plants between the two macrophytes covered areas. It seemed that cattail induced lower phosphorus concentrations in its covered area while reed performed better in phosphorus removal from water along with the flow. The phosphorus distributions were greatly influenced by many environmental factors in reed community area while the relationships between PFs and environmental factors in cattail community were generally less significant. More research on other parameters should be taken into account in the future for revealing the mechanism of macrophytes on phosphorus removal and retention.