3.1 Characteristic analysis of DOM spectral indexes
The salinity variations in pore water contributed to the DOM fluorescence property changes. The UV spectral indexes and fluorescence spectral indexes are shown in Fig. 1. The spectral slope ratio (SR) indicated the DOM molecular weight and aromaticity (Feng et al., 2022), which was inversely proportional to the DOM molecular weight. There were significant fluctuations in the SR values in each group. SR had a decreasing trend over time because bacterial colonies used organic matter with small molecular weights for their metabolism and substance synthesis (Zhang et al., 2020). The SR values of group S1 were significantly higher than those of groups S2 and S3 (P < 0.05). Higher salinity favored the retention of high-molecular-weight substances (Strehse et al., 2018), resulting in a relatively high proportion of aromatic macromolecular stable substances in the high-salinity group. Therefore, salinity significantly affected the DOM in pore water, and lower salinity caused lower DOM aromaticity in pore water. The biological index (BIX) was used to measure the proportion of organic matter of autochthonous origin, and BIX > 1.0 indicated that the DOM was predominantly of autochthonous origin (Rodríguez-Vidal et al., 2020; Feng et al., 2022). In general, the BIX values were greater than 1.0 among all the groups, indicating that DOM was mainly endogenous in pore water. The BIX values of group S1 were significantly lower than those of groups S2 and S3 (P < 0.05), implying that higher salinity inhibited the endogenous effects of DOM. Wetland plants and bacterial communities play a central role in the translocation and transformation of DOM components (Muscarella et al., 2019; Zhang et al., 2020). High salt stress may disrupt the balance of DOM uptake and release by reeds and affect bacterial biosynthesis as well as metabolic activity, thereby reducing the endogenous contribution to DOM in pore water. The HIX values in group S1 were much lower than those in groups S2 and S3 (P < 0.05), showing that higher salinity enhanced the degree of humification. Different species of bacteria tended to discriminate between high or low salinity xenobiotic metabolism and signal transduction, which affected DOM production or degradation (Xiao et al., 2022).
3.3 Influence of salinity variations in DOM fluorescent components
The DOM components in each group are shown in Fig. 2. The percentage of each DOM component significantly varied among the different groups. Protein-like components (C2, C3, C4) dominated and fulvic acid-like component (C1) was less predominant. The dominant component changed from C2 at the beginning of the experiment to C1 at the end of the experiment, especially for groups S2 and S3, suggesting that fulvic acid-like substances were more easily retained in high salinity pore water. The percentage of C1 in group S1 consistently increased from 2.48% on the 1st day to 39.45% at the end of the experiment (Fig. 2a), which reached 54.88% in group S2 (Fig. 2b) and 48.79% in S3 (Fig. 2c). However, component C2 in group S1 decreased from a high percentage (68.83%) on the 1st day to 10.27% at the end of the experiment (Fig. 2a), and the percentages in groups S2 and S3 on the last day were 3.65% (Fig. 2b) and 4.17% (Fig. 2c), respectively, explaining the higher HIX values of groups S2 and S3. Compared with the original samples, the Fmax of DOM in group S1 significantly increased within the first 4 d of the experiment, especially that of the protein-like components (C2 and C4), while the Fmax of DOM in groups S2 and S3 showed a decreasing trend. The DOM concentrations in group S1 were higher than those in groups S2 and S3. Hence, higher salinity inhibited the release of protein-like substances and was conducive to DOM precipitation and adsorption. It was reported that organic matter is continuously back-flocculated and reflocculated under natural conditions, transparent exopolymer p sedimentation articles usually exuded by plants can be used as precursors for flocculation, and Na+ destroys the bilayer structure of sediment colloids (Wang et al., 2018), which changes colloidal particles from phase repulsion to phase adsorption, destabilizing colloids and promoting DOM adsorption and precipitation. Furthermore, protein-like and fulvic acid-like components are mostly hydrophobic, and the solubility of the protein components in pore water decreases at high salinity (Butturini et al., 2022). In addition, the reed plant cells grew and multiplied rapidly in group S1, accelerating the cell life cycle, and then ruptured to release DOM. Tyrosine and tryptophan components are preferentially released by plants. Salinity reduces plant productivity and carbon input (Kou et al., 2020). In addition, an imbalance between DOM released from cells due to the senescence or lysis of the microbial cells of primary producers under salt osmotic stress and DOM mineralization due to heterotrophic depletion of bacteria may also be the reason for the reductions in the protein-like components and total DOM contents in the higher-salinity groups. In groups S2 and S3, the fulvic acid-like component (C1) varied differently from the protein-like components (C2, C3, and C4) in DOM, with an increasing trend in C1 from the 40th day to the 70th day. This implied that salinities of 3600 and 6000 mg/L promoted DOM transformation to some extent. Protein-like fluorescent substances and fulvic acid-like fluorescent substances are structurally linked (Feng et al., 2022). It has been suggested that fulvic acid-like substances are the degradation products of tryptophan or protein-like and soluble microbial by-products (Cheng et al., 2021), explaining the increasing tendency of fulvic acid-like substances (C1) at the end of the S2 and S3 experiments (Fig. 2e). The Fmax of components C1 and C3 in groups S2 and S3 were relatively higher than those in group S1; however, no significant difference was found among them (P > 0.05) (Fig. A2). The Fmax of C2 and C4 in group S1 was significantly higher than that in groups S2 and S3 (P < 0.05) (Fig. A2). Salinity acts as a major driver of bacterial community composition and diversity in water bodies. There are also differences in the utilization and release production of DOM components among different bacteria. The bacterial utilization of organic matter in the environment is sequential, and bacteria preferentially utilize protein-like substances (Zhang et al., 2020). This biodegradation process was accelerated under high salinity, especially that of tyrosine-like components and aromatic proteins, which may be the reason for the significant differences in Fmax of protein-like substances (C2 and C4) among the groups.
3.4 Response of bacterial communities to environmental factors and DOM
The relative abundances of the dominant genera of each group were investigated (Fig. 3). As shown in Fig. 3, autotrophic bacteria remained the dominant species because nutrients were consumed and bacterial metabolites accumulated. Thiobacillus (8.86%-21.63%), Massilia (8.85%-37.44%), and Flavobacterium (2.48%-14.4%) were the dominant genera during the first 40 d of the experimental groups, while on the 55th and 70th days, only Thiobacillus (10.87%-18.96%) dominated. The salinity gradients caused changes in key bacterial communities. During the first 40 d of the experiment, the relative abundance of Thiobacillus in group S3 was higher than that in groups S1 and S2. At the end of the experimental period, the relative abundance of Thiobacillus in group S2 was the highest (13.28%). The relative abundances of Massilia, Flavobacterium, and Pseudomonas in group S3 were lower than those in groups S1 and S2 during the experiment.
It was reported that salinity detracts vital water from bacteria through their cell walls, leading to microbial dehydration. Moreover, higher osmotic pressure can disrupt microbial cell structures (He et al., 2017). It is hypothesized that high salinity indirectly affects DOM compositions via its effects on bacterial activities.
To explore the interactions between bacteria and DOM components under salinity gradients, RDA was performed between the dominant bacterial genera and DOM components and salinity (Fig. 4). Thiobacillus showed a strong positive correlation with salinity, indicating that the genus Thiobacillus is salt-tolerant, which explained the dominance of Thiobacillus in the group with high salinity (Fig. 3). Moreover, Thiobacillus showed a negative correlation with the Fmax of C2 and C4. Zhu et al (2022) found that the chemoautotrophic bacterium Thiobacillus indirectly affects DOM release through its sulfur oxidation process. Previous studies have also found that DOM can act as a Thiobacillus photosensitizer, triggering apoptosis, while the photosensitizing DOM is constantly depleted by its oxidation (Huang et al., 2022b). Although Thiobacillus did not directly utilize DOM as its carbon source for biosynthesis, the increase in salinity may indirectly inhibit the release of protein-like components and promote the self-oxidation of DOM via its relationship with Thiobacillus, which was presumably the important reason for the lower Fmax of protein-like components in group S3. Flavobacterium showed a negative correlation with the Fmax of C1, indicating that fulvic acid-like substances were metabolized by Flavobacterium and acted as a carbon source for degradation. Flavobacterium prefers to utilize specific types of macromolecularly active DOM, and even some difficult-to-treat organic matter is consumed by Flavobacterium (Chen et al., 2021). In this study, there was a positive correlation between the relative abundance of Flavobacterium and salinity. However, previous studies have reported that the degree of activity of Flavobacterium was relatively high in riparian areas with lower salinity (Madetoja et al., 2003; Zhang et al., 2006), which was different from the results of this experiment. Consequently, it was shown that a salinity of 6000 mg/L had an inhibitory effect on the growth of Flavobacterium, while the organic matter in the pore water and the interroot oxygenation of reeds in this experiment may have alleviated the salinity stress. Pseudomonas, Brevundimonas and Polaromonas were negatively correlated with salinity, indicating that salinity exerted an inhibitory influence on their relative abundance. Pseudomonas, Brevundimonas and Polaromonas were negatively correlated with the Fmax of C1 and positively correlated with the Fmax of C2 and C4. Pseudomonas was capable of conversion using various organic compounds and exhibited high degradation efficiency for various aromatic compounds (Dong et al., 2019; Chen et al., 2021). Pseudomonas may breakdown highly aromatic fulvic acid-like substances (C1) and produce amino acids through protein degradation. The produced amino acid substances promoted organic matter degradation and transformation by Pseudomonas with a synergistic effect (Dong et al., 2019). The genus Brevundimonas is known for its ability to degrade organic matter associated with DOM transformation (Feng et al., 2021). Polaromonas is a microbial strain that still shows protein hydrolysis activity in extreme environments (Matsui et al., 2017) and is involved in the production of C2 and C4 fractions. Thus, their relative abundances were positively correlated with the Fmax of C2 and C4 and negatively correlated with the Fmax of C1. Overall, higher salinity inhibited the relative abundances of bacteria associated with the degradation of fulvic acid-like substances and the production of protein-like substances, resulting in higher HIX and lower BIX values.
The relative abundance of Desulfatiglans was higher in group S3 than in group S1 from the 55th to the 70th day; the relative abundance of Massilia in group S3 was lower in the first 40 d of the experiment than in the other groups; and the relative abundance of Algoriphagus showed a higher level on the 10th day in group S3. Combined with the fact that the relative abundance of these bacteria did not correlate with salinity, salinity slightly affected them, but also had some inhibitory or promoting effects. Desulfatiglans showed a strong positive correlation with the Fmax of C1 and a negative correlation with the Fmax of C2 and C4. Desulfatiglans uses a wide range of carbon substrates including aromatic organic compounds (Osterholz et al., 2022), suggesting its ability to degrade tannins and lignin-based complex aromatic compounds to produce fulvic acid-like substances. Both components C2 and C4 represented aromatic compounds, indicating that the bacterium degraded C2 and C4 as a source of nutrition. Massilia and Algoriphagus showed a strong negative correlation with the Fmax of C1 and the Fmax of C3. Massilia has been reported to exhibit strong hydrolytic activity, and its relative abundances were associated with DOM decomposition (Sedláček et al., 2022). Massilia reportedly produced indole derivatives by using tryptophan, which was the reason for the negative correlation with the Fmax of C3 (Myeong et al., 2016). Algoriphagus is also known for the catabolic transformation of difficult-to-degrade macromolecules due to its ability to secrete polycyclic aromatic hydroxylated dioxygenases (Ganiyu et al., 2022). The relative abundances of bacteria were at a low level at the end of the experiment, which may account for the insufficient carbon source associated with the low DOM contents.
3.5 Effect of salinity on the relative abundances of key biosynthetic and metabolic gene units related to DOM transformation
Tryptophan and tyrosine are the primary components of DOM. Therefore, the effects of salinity on the relative abundances of key metabolic genes associated with tryptophan and tyrosine transformation were determined by specific key KEGG modules. Three modules related to tryptophan synthesis and degradation were identified: M00023 was related to tryptophan biosynthesis, M00037 characterized melatonin biosynthesis and M00038 characterized tryptophan metabolism, which is mainly related to tryptophan degradation and conversion.
The abundances of key tryptophan metabolic function genes are shown in Fig. 5. During the experimental period, the relative abundance of M00023 in group S3 was higher than the corresponding values in groups S1 and S2 (P < 0.05). The relative abundances of M00038 in group S3 were significantly lower than those in groups S1 and S2 (P < 0.05) from the 40th to 70th day of the experimental period (Fig. 5b). A previous study revealed that autotrophic and heterotrophic bacteria coexist in wetland systems, with DOM production and degradation occurring simultaneously (Zhang et al., 2020). High-sodium environments diminish the ability of bacteria to absorb external substances, thus increasing Na+ accumulation in the microbial cytoplasm and interfering with metabolic processes (Napieraj et al., 2020). The tryptophan-like substances in high salinity environments decreased to a certain level, and microbial tryptophan metabolism in the external environment gradually shifted to tryptophan synthesis by itself to resist the effects of salinity stress. In addition, tryptophan alleviates the effect of salinity on microorganisms, which is responsible for increased tryptophan biosynthesis pathways (Kahveci et al., 2021). Bacteria (such as Actinobacteria and Proteobacteria) involved in amino acid transport/metabolism. It was assumed that higher salinity inhibits tryptophan degradation metabolic pathway genes in cells and further enhance tryptophan biosynthesis capacity for microbial cells to resist the effects of salt stress. Thus, tryptophan alleviated the effect of higher salinity on microorganisms, which was responsible for increased pathways of biosynthesis, clarifying the relatively high activity of the tryptophan synthesis gene module (M00023) and the relatively low activity of the tryptophan degradation gene module (M00038) in group S3.
The relative abundances of key metabolic function genes of tyrosine are shown in Fig. 6. Four modules were related to tyrosine synthesis and degradation: M00025 and M00040 were associated with tyrosine biosynthesis, M00042 characterized catecholamine biosynthesis and M00044 characterized tyrosine degradation, which was mainly related to tyrosine degradation and transformation.
The relative abundances of M00025 in group S1 were higher than those in groups S2 and S3 during the first 40 d of the experiment. The relative abundances of M00040 in group S3 were lower than those in groups S1 and S2, which accounted for the differences in tyrosine compositions between groups S1 and S3 at the beginning of the experiment. Tyrosine has a significant antioxidant effect that prevents cellular damage from osmotic stress generated by salinity (Tang et al., 2021). The relative abundances of M00044 in group S3 were significantly lower than those in groups S1 and S2 on the 40th to 70th day, suggesting that high salinity inhibited the metabolic degradation ability of tyrosine. High salinity disrupts microbial enzyme structures by increasing cellular osmotic pressure (He et al., 2017) and affects the degradation of protein-like substances (Zhu et al., 2022). Hence, the relative abundances of tyrosine metabolic pathway genes were lower under high salinity. Higher salinity stimulated the microbial cells in the environment; the microbial cells took corresponding measures to reduce their tyrosine metabolic activities, which may be the reason for the slow decline in tyrosine over time in groups S2 and S3.