4.1. Decomposition Process of Floating-Leaved Macrophytes
In this study, we have used two ponds to conduct our in situ monitoring. As for the difference between the two ponds, Pond A was an old one fully covered with floating-leaved macrophytes and thick silt, while Pond B was newly dug, in which initial succession of aquatic macrophytes and silt accumulation had not started. Hence, we could hypothesize that the fluctuations of N and P levels in the waterbody of Ponds A during our monitoring period were mainly caused by the decomposition of biomass. One would argue that the sediment might exert strong effects on the waterbody through physicochemical deposition and redissolution of N and P on the water–sediment interface. We thought that it was also influenced by environmental factors such as DO, which were eventually triggered by the decomposition process.
Our study revealed two obvious phases in the decomposition of floating-leaved macrophytes, with the first phase being faster and the second slower, which is in agreement with former reports of wetland plants decomposition with litterbags [25
], indicating our initial hypothesis that treating the ponds in drying season without much water exchange as relatively enclosed ecosystems was practical, and the litterbags could be skipped in similar research. In the first phase, the leaching of water-soluble substances was the primary mechanism for the loss of litter mass [27
], which released phenols, sugars, and elements such as N, P, and potassium into water. In the second phase, microbial decomposition and transformation of hard-to-decompose substances (such as cellulose) were mainly carried out in water and sediment [26
]. The fragmentation of aquatic macrophytes residues was also affected by micro-, meio- and macroinvertebrates [28
]. Abiotic factors such as temperature, DO, and pH may further affect the decomposition of aquatic macrophytes [29
]. Consistent with the reports of Enriquez et al. [31
] and Chimney et al. [32
], decomposition rates were negatively correlated with litter C:N and C:P molar ratios and positively correlated with N and P contents. The decomposition rate in our study was fast in Patch a dominated by T. bispinosa
, while relatively slow in Patch c dominated by A. imbircata
. This was due to the higher initial TN, TP, and lower C:N and C:P molar ratios in T. bispinosa
compared to those in A
4.2. Effect of Floating-Leaved Plant Decomposition on Nitrogen and Phosphorus
It is well known that the decomposition of plant biomass increases the amount of organic matter in the environment [33
], and it is also considered as the carbon source of constructed wetlands [34
]. The promotion of microbial activity along with the decomposition of aquatic macrophytes was also reported [35
]. Therefore, it was reasonable to deduce that the decomposition of floating-leaved macrophytes released a large amount of dissolved organic matter, among which were organic N and P compounds, and this constituted the direct influence of biomass deposition on water N and P levels.
However, from our long-term monitoring results, no linear correlation between N levels (TNW or TNS) or P levels (TPW or TPS) and biomass was observed at our research site. Reduction in biomass alone could not explain the variation in N and P levels. Therefore, we built the structural equation model to better understand the relationships between the parameters and the driving force(s) of the N and P dynamics during plant decomposition. The model showed that besides the direct effect of aquatic macrophytes decomposition, regulation of DO and pH could indirectly influence N and P cycles in Pond A.
In our study, significant changes in DO and ORP were detected in relation to biomass decomposition, which we thought, through regulation of the redox state of the water body, indirectly affected the TPW
. P can bind to metal ions at the water–sediment interface of aquatic ecosystems, forming Al-bound P (Al-P), Fe-bound P (Fe-P), and Ca-bound P (Ca-P) [36
]. Among those inorganic phosphorus compounds, Fe-P is the most unstable. When ORP is low, the water–sediment interface is in the reduced state [37
], and reductive dissolution of phosphorus–bearing iron (oxyhydr)oxides is regarded as a primary mechanism responsible for the mobilization of phosphorus in sediments [39
], as Fe3+
is transformed into Fe2+
, and P is released to the water body, which increases the TPW
. Therefore, with low DO and ORP in the first deposition phase, we assumed part of the P from sediment in Pond A was also redissolved into the water body, which further increased TPW
. With the recovery of DO and ORP in the second phase, the water–sediment system turned into the oxidize state, and P in the water was adsorbed by colloidal Fe(OH)3
and precipitated or combined with Fe3+
to form FePO4
, then deposited in the silt along the bottom [41
], thus TPW
decreased and TPS
The water pH levels in Ponds A and B were both between 7 and 9, and the overall pH level in Pond A was generally lower than that in Pond B. We assumed that the plant decomposition and the release of organic acids decreased the pH level in Pond A. However, with the data we detected on-site, it was a little difficult to explain the influence of pH on P levels. The SEM model indicated a positive relationship between water pH and TPW
but a negative relationship between water pH and TPS
, whereas reversed relationships were observed by RDA and Spearman correlation analysis. With the results from Jin et al. [42
], P release was the lowest in neutral conditions, and an increase in pH could free P from its binding to ferric complexes due to the competition between hydroxyl ions and the bound P ions [42
]. With our understanding, we also suggested that the SEM results could be more reasonable, as increased pH in water promoted the release of P from sediment into the water. The opposite results from RDA and Spearman correlation analysis might be reflections of other environment factors on P. As was mentioned by Chen et al. [44
], besides DO and pH, temperature, nitrate, organic matter in the environment as well as the activity of microorganism, etc., could all influence P migration and transformation across the water–sediment interface. In addition, we only monitored the water pH level, and that of the sediment might be different. Further monitoring of pH level in sediment was recommended in future study. Nevertheless, the contradictory results obtained in this study serve as a reminder of the importance of method selection when analyzing field data.
The N levels in water and sediment of Pond A were both consistently higher than those in Pond B. Unlike P, N in Pond A remained relatively stable during the plant decomposition process. We suggested that the release of N from decayed biomass continued, whereas N had experienced fast nitrogen mineralization, nitrification, denitrification and anammox [45
]. As has been suggested, dissolved organic carbon released by plant decomposition enhanced the activity of denitrifying microorganisms [46
]. Volatilization of N as N2
O and NH3
in Pond A might balance the N newly released from biomass.
From our research of the N and P dynamics during aquatic macrophyte decomposition, it was difficult or improper to simply conclude if the decomposition process had positive or negative effects on water quality. As for the conflicting opinions raised by different researchers, we thought these were mainly due to the parameters chosen to represent the water quality. Nevertheless, the joint use of multiple quantitative methods of RDA, Spearman correlation analysis, and SEM offers a better explanation of the complex relationships among the parameters in the aquatic ecosystem.
Generally, the decomposition of biomass better explained P levels than N levels in the water body of the ponds we studied, indicating that P was the decisive factor that caused the endogenous eutrophication of ponds during plant decomposition. In Schindler’s [8
] research, no evidence was found that proved eutrophication can be managed by controlling N inputs, while there was much evidence that controlling P inputs allows management at the whole–lake scale. Therefore, in order to control the outbreak of water eutrophication, it is necessary to remove the floating-leaved macrophytes from the ecosystem before their decomposition. For those eutrophic ponds with thick layers of silt in the sediment, dredging was also recommended.