1 Introduction

The health of mariculture systems directly determines aquaculture benefits and sustainable development of the sector. However, due to intensive culture, excessive feed, and biological excretion, the mariculture habitat was badly damaged (Xu et al., 2022) and a large hypoxia area occurred (Wang et al., 2021), which seriously destroyed the balance of the ecological environment. In this environment, the degradation of sulfur-containing proteins and the dissimilatory sulfate reduction lead to the generation and diffusion of substantial sulfide (Fan et al., 2020). Sulfide is reported as the silent killer in the mariculture ecosystem and it is also an important indicator in assessing the quality of marine sediments (Hu et al., 2018). In the mariculture sediment, a small part of sulfide may further form stable pyrite sulfur and metallic phase (FeS₂), while 99% of the soluble sulfide (including H₂S, HS⁻, S²⁻) continues to spread into overlying seawater (Jørgensen, 1982). It heavily impacts mariculture organisms, destroys intestinal microbial structures, or directly produces toxic effects on marine organisms (Pozsgai et al., 2019). Therefore, to prevent further deterioration of mariculture habitat, effective measures should be taken to control the production of sulfide.

The coastal mariculture region has a very high sulfide content. The investigation by Meng et al. (2019) revealed that mariculture sediments from Jiaozhou Bay of China contained up to 630 mg kg⁻¹ of sulfide. Furthermore, the sulfide production procedure is extremely complicated. It is subjected to the coordination of various biological and abiotic factors in mariculture habitats. Among them, sulfate-reducing bacteria (SRB) and other bacteria that mineralize sulfur-containing proteins mainly contribute to the production of sulfide (Jørgensen, 1982; Zhou et al., 2014; Fan et al., 2020). The latter covers a wide range of groups and is hard to be investigated by phylogenetics and functional genes (Jørgensen, 1982). Compared with other bacteria, SRB plays a more important role in terms of sulfide generation under an anoxic organic mariculture environment (Zhou et al., 2014). Studies have shown that SRB is the main group for sulfide production in the ocean, and only less than 3% of sulfide comes from the mineralization of organic sulfur (Jørgensen, 1977). SRB also contributes to the ocean carbon cycle, more than 50% of the mineralization of organic matter was conducted by SRB (Jørgensen, 1982; Boetius et al., 2000). Therefore, an in-depth understanding of the SRB involved in sulfide production in mariculture ecosystems and seeking safe methods to inhibit SRB activity are important for sulfidogenic inhibition.

Recently, the studies on the inhibition of SRB activity mainly focused on metal pipes corrosion in reservoir and oil production processes (de Andrade et al., 2020). To control such corrosion and reduce loss, chemical fungicides, such as glutaraldehyde, tetramethyl-phosphatesulfate, etc. were often used to inhibit SRB activity (Wen et al., 2009). Long-term use of chemical fungicides might induce SRB to develop drug resistance. Considering the actual situation of aquaculture, chemical fungicides have been severely prohibited due to their strong secondary pollution. Therefore, biological methods gradually become acceptable in controlling the production of sulfide. Investigation showed that the addition of nitrate increased the abundance of denitrifying bacteria in the oilfield-produced water, and led to an obvious reduction of SRB (Qi et al., 2022). Meanwhile, nitrate improved the biological competition of nitrate/nitrite reducing bacteria (NRB) over SRB for electron donors. Consequently, the SRB sulfate reduction activity was inhibited (Lai et al., 2020). Myhr et al. (2002) showed that NO₂⁻, a product of NO₃^- reduction, revealed a stronger inhibitory effect on SRB and it disturbed the expression of sulfite reductase, consequently, SRB was not able to reduce SO₄²⁻ to poisonous sulfide. The reduction of NO₃⁻ also improved the oxidation-reduction potential (ORP) of the environment, which made the sulfide production driven by SRB impossible (Fan et al., 2020). However, the NO₃⁻ inhibition mechanism on sulfide production in mariculture sediments and the response of microbial community should be further investigated.

Bait residue and animal waste lead to nitrogen accumulation in mariculture habitat (Thomas et al., 2010). Xin et al. (2019) showed that dissolved inorganic nitrogen increased by 7-fold over the past 60 years in Bohai Sea, China. The concentration of nitrate in some coastal waters in the Philippines increased by 90% over recent 10 years (San Diego-McGlone et al., 2008). According to the investigation of the sea cucumber culture pond, the annual accumulation of NO₃⁻ was up to 1.2 mg L⁻¹ (Kang and Xu, 2016). Therefore, NO₃⁻ is ubiquitous in the mariculture environment with a high concentration. However, how to make full use of nitrate to inhibit SRB activity still need further study.

In this study, the sulfate-reducing microbiota (SRM) enriched from mariculture sediments was used as the target to reveal the possibility of NO₃⁻ inhibiting the SO₄²⁻ reduction activity. Meanwhile, the response of the microbial community, dsrB gene, and SRM to NO₃⁻ dosage was investigated, providing a theoretical basis for the control of sulfide in the mariculture habitat.

2 Materials and Methods 2.1 Enrichment of Sulfate-Reducing Microbiota (SRM)

Sulfate-reducing microbiota (SRM) were enriched from mariculture sediments in Qingdao by using the modified Starkey medium. The medium included (g L⁻¹): K₂HPO₄ 0.5; NH₄Cl 1.0; Na₂SO₄ 0.5; 65% sodium lactate 5.0; CaCl₂ 0.076; MgSO₄·7H₂O 2.0; (NH₄)₂Fe(SO₄)₂·7H₂O 0.5; ascorbic acid 0.1; yeast powder 1.0; pH 7.0 – 8.0, and 30 g L⁻¹ NaCl was added to simulate seawater salinity. Among them, the reductants (NH₄)₂Fe(SO₄)₂·7H₂O and ascorbic acid were filtered through a 0.22 μm membrane. The prepared liquid medium was divided into 100 mL serum bottles and each bottle contained a 50 mL medium. High purity N₂ was used to degas for 3 min, followed by high-pressure steam sterilization (121℃, 20 min). When the medium was cooled to room temperature, sterilized reductants were added.

Each bottle of the medium was filled with 50 g of mariculture sediments. The whole process was operated in an anaerobic environment. Afterwards, the bottles were placed in an incubator at 37℃ in dark. When the upper liquid medium turned black and rotten eggs smell produced, the enriched upper liquid medium was inoculated into the fresh liquid medium with a sterile syringe at 1% (V/V). The bottles were placed in an incubator at 37℃ in dark. The above procedure was repeated three times and finally, the SRM was obtained.

2.2 Nitrate Inhibition of Sulfate-Reducing Microbiota (SRM) 2.2.1 Function of sulfate-reducing microbiota (SRM)

The enriched SRM was inoculated into the fresh liquid medium with a sterile syringe at 1% (V/V). It was incubated in dark at 37℃. Samples were daily taken to measure the optical density of solution (OD₆₀₀), and the concentrations of SO₄²⁻ and H₂S.

2.2.2 Inhibition of sulfate-reducing activity by nitrate

A fresh liquid medium was prepared and nitrate was added to approach the final concentration of NO₃⁻ 1, 3, 5, and 7 mmol L⁻¹, respectively, which were similar to some coastal waters with heavy nitrate pollution (Wu et al., 2019). The five treatments were marked as Treat-N0 (control), Treat-N1, Treat-N3, Treat-N5 and Treat-N7. The fresh SRM solution was centrifuged at 4000 g for 20 min. Discard the supernatant, add the same volume of fresh liquid medium, and resuspend. The bacteria solution was then transferred to the fresh culture medium containing nitrate with a volume ratio of 1% (V/V). The bottles were placed in an incubator at 37℃ in dark. Samples were daily taken to determine the concentration of H₂S, SO₄²⁻, NO₃⁻, NO₂⁻, and NH₄⁺. The samples with 7 mmol L⁻¹ NO₃⁻ and the control group were taken for DNA extraction on day 3 and day 7, respectively. The abundance of dsrB gene was measured by qPCR (Bio-rad CFX96, USA) to characterize the changes in the quantity of SRB. High-throughput sequencing was performed for two microbial community samples on day 7 to reveal the changes in the structure of SRM before and after inhibition.

2.3 Analysis Methods 2.3.1 Chemical and physical analyses

The optical density (OD₆₀₀), and concentrations of H₂S, NO₃⁻, NO₂⁻ and NH₄⁺ were determined by spectrophotometry as the protocols described by Chen et al. (2022). The concentration of SO₄²⁻ was assayed using a Dionexion chromatography (ICS-2100) system (Dionex Corporation, Sunnyvale, CA, USA). All the measurements of OD₆₀₀, H₂S, NO₃⁻, NO₂⁻, NH₄⁺ and SO₄²⁻ were carried out in triplicate, and values were presented as mean ± standard deviation. The analyses of other parameters were performed in duplicate and the associated statistical analyses agreed to within 95% confidence.

2.3.2 Molecular biological analysis

Sediment samples enriched SRM, and nitrate inhibited SRM (7 mmol L⁻¹ nitrate for 7 d) were obtained, and a DNA extraction kit (Mobio, USA) was used for DNA extraction.The composition of microbial communities was determined by 16S rRNA gene-based high throughput sequencing.

For each sample, 45 mL solution was centrifuged at 4000×g for 20 min. The supernatant was discarded, and the precipitated bacteria were used for DNA extraction. The partial 16S rRNA gene-based high-throughput sequencing was performed according to the previous method (Gao et al., 2014). Bacterial 16S rRNA gene universal primers were 515F: 5'-GTG CCA GCA GCC GCG GTAA-3' and 806R: 5'-GGA CTA CCA GGG TAT CTA AT-3', corresponding to V3 – V4 variable region of 16S rRNA gene. The sequencing was conducted by using the Illumina Miseq platform in Novegene, China.

As the number of dsrB genes, which encode the subunit of sulfite reductase in SRB, was directly proportional to the amount of SRB, DsrB was consequently used to estimate the quantity of SRB. Quantitative analysis of specific dsrB genes in SRB was performed with primers DSR-p2060F: 5'-CAA CAT CGT YCA YAC CCA GGG and DSR-4R: 5'-GTG TAG CAG TTA CCG CA as described in previous studies (Liu et al., 2016). The plasmid with dsrB gene fragment was used as the standard for fluorescence quantitative PCR reaction to draw the quantitative standard curve of dsrB gene. The extracted DNA samples were detected by quantitative PCR. The reaction system and procedure are consistent with those used in establishing the standard curve (Liu et al., 2016). The negative and positive controls were set. The dissolution curve was used to check whether there was non-specific amplification, and the number of detected cycles was substituted into the standard curve to obtain the gene copy number.

2.4 Data Analysis

One-way ANOVA in SPSS 17.0 software was used to conduct statistical analysis on the NO₃⁻ inhibiting SRM. High-throughput sequencing sequences were filtered according to Gao et al. (2014). To reveal the species composition of each sample, the operational taxonomic units (OTUs) clustering was performed on the valid sequences with 97% similarity. Then the OTUs sequence was annotated by species. The Alpha diversity of community sequences was analyzed (Shannon, Simpson, Chao1, ACE).

3 Results and Discussion 3.1 Characteristics of Sulfate-Reducing Microbiota (SRM)

The growth curve and sulfate reduction capacity of SRM is shown in Fig. 1. The SRM presented a high growth rate and reached a logarithmic phase in 20 h. On the 30th hour, OD₆₀₀ reached the maximum value of 1.642. The initial concentration of SO₄²⁻ was about 1700 mg L⁻¹. The concentration of H₂S increased to 421 mg L⁻¹ on day 3. According to the calculation of sulfur mass balance, the maximum theoretical yield of H₂S was 602 mg L⁻¹, a little higher than the actual value. The reduction process of SO₄²⁻ to H₂S consists of a series of enzymatic reactions. In this process, sulfate gains electrons and produces multiple intermediates, such as SO₃²⁻, S₃O₆²⁻, and S₂O₃²⁻. Therefore, intermediate metabolites were the main reason that the actual H₂S production was lower than the theoretical value. Finally, the sulfate conversion rate by the SRM reached 69.92%. Lai et al. (2020) enriched SRM from oil effluent and the concentration of H₂S increased to 122 mg L⁻¹ on day 6. Thus, the SRM enriched in this study presented a strong reduction capability of SO₄²⁻ and a high capacity of H₂S generation.

3.2 Inhibition of Nitrate on Sulfate-Reducing Microbiota

As shown in Fig. 2, NO₃⁻ presented a certain inhibitory effect on SRM activity and the inhibitory capacity became stronger with the increase of NO₃⁻ concentration. The consumption of SO₄²⁻ was accompanied by the generation of H₂S. When no nitrate was added, SRM generated H₂S on the second day. When the concentration of NO₃⁻ was 1, 3, and 5 mmol L⁻¹, it presented a slightly inhibitory effect on SRM. Compared with the control, the inhibition time lasted 1–3 d. However, higher NO₃⁻ (7 mmol L⁻¹) revealed a significant inhibitory effect on SRM. The inhibition duration was increased to 6 d, and the final maximum content of H₂S decreased by about 60 mg L⁻¹ compared with the control.

Further analysis revealed that all NO₃⁻ was reduced within 1 d (Fig. 3a). Meanwhile, NO₂⁻ and NH₄⁺ were generated accordingly (Figs. 3b, 3c). Taking the 5 mmol L⁻¹ NO₃⁻ inhibition group as an example, with the presence of nitrate reductase, 5 mmol L⁻¹ NO₃⁻ was reduced to 1.87 mmol L⁻¹ NO₂⁻ and 1.95 mmol L⁻¹ NH₄⁺ within 1 d. On the third day, all NO₂⁻ was reduced, and the content of NH₄⁺ in the medium was maintained at about 4.17 mmol L⁻¹ (Eq. (1)), which indicated that SRM reduced most of NO₃⁻/NO₂⁻ finally to NH₄⁺.

$ \mathrm{NO}_2^{-}+6 \mathrm{e}^{-}+8 \mathrm{H}^{+} \rightarrow \mathrm{NH}_4^{+}+2 \mathrm{H}_2 \mathrm{O}. $

(1)

With the presence of NO₃⁻ or NO₂⁻, there was no significant reduction of SO₄²⁻ (Fig. 2a), indicating that SRM could preferentially use NO₃⁻/NO₂⁻ as electronic acceptors.

Interestingly, the SRM enriched in this study has the ability of dissimilatory nitrate reduction to ammonium (DNRA). DNRA was divided into two steps: in the first step, dissimilatory nitrate reductase (Na R) reduces NO₃⁻ to NO₂⁻ (Pandey et al., 2020); in the second step, nitrite reductase (Ni R) reduces NO₂⁻ to NH₄⁺ (Pandey et al., 2020), which was observed in this study (Figs. 3a – 3c). Ni R enzyme is a kind of peritytoplasmy enzyme encoded by nrf A gene (Darwin et al., 1993). With Ni R enzyme, Desulfovibrio spp. reduce nitrite to NH₄⁺ (Price et al., 2011). Desulfovibrio spp. had been proved to have the ability of DNRA (Steenkamp and Peck, 1981). Thus, the inhibition of nitrite to Desulfovibrio spp. was relieved by this transformation as ammonia presenting less toxic than the sulfide. In this study, according to the results of high-throughput sequencing, high abundant Desulfovibrio was found in the SRM, thus resulting in the accumulation of NH₄⁺. Metatranscriptomics studies have shown that, besides some Desulfovibrio strains, the vast majority of SRB contains an enzyme system to reduce nitrate (Haveman et al., 2005; Zhou and Xing, 2021). From the perspective of thermodynamics, denitrification is easier conducted than sulfate reduction (Eckford and Fedorak, 2002). Okabe et al. (2003) investigated the inhibitory effect of nitrate/nitrite on the activity of sulfide production in sludge. They showed that 0.3 – 1.0 mmol L⁻¹ nitrate/nitrite were preferentially reduced by the bacteria. This pushed the sulfide reduction zone into the deeper area and inhibit the generation of sulfide.

As NO₂⁻ was present (day 3) and absent (day 7), qPCR was performed to reveal the dsrB gene abundance for 7 mmol L⁻¹ NO₃⁻ addition group. Combined Fig. 3d with Fig. 2b, compared with the control, dsrB gene abundance decreased by 3 orders from 9 to 6, which suggested that the quantity of SRB distinctly decreased. Meanwhile, the presence of NO₂⁻ inhibited the generation of H₂S. When all NO₂⁻ was eliminated on day 7, there was no signifycant difference in the abundance of dsrB gene compared with the control. This indicated that the SRM could immediately recover the capability of reducing SO₄²⁻ to H₂S under the condition of the absence of NO₂⁻. Therefore, the reason that NO₃⁻ inhibited the sulfate activity of SRM should be ascribed to the fact that NO₂⁻ could effectively inhibit the activity of SRB. This is more conducive to delaying the generation of toxic H₂S. In previous studies (Myhr et al., 2002; Zhou and Xing, 2021), the direct addition of NO₂⁻ inhibited SRB activity significantly. The number of denitrifying bacteria increased in the reaction system, which resulted in a significant decrease in the number of SRB. Pillay and Lin (2013) inhibited SRB activity by adding the same concentration of nitrite and nitrate and they found the nitrite inhibition effect was better than nitrate. Further analysis showed that NO₂⁻ and denitrifying intermediates (NO, N₂O) presented more obvious inhibition on bacteria than nitrate. Nitrite and its oxides NO₂, NO, and N₂O inhibit SRB activity by interfering with the sulfite reductase (Pillay and Lin, 2013). According to the specific analysis of intermediates, the inhibition of NO₂⁻ on bacteria was related to NO (produced in the process of denitrification) or nitroso acyl and NO+ complex. It inhibited the activity of enzymes during sulfite reduction to H₂S. However, NO contains an activated unpaired electron, which is in a highly active state and has non-obligate toxicity inhibition to many bacteria. The inhibition of N₂O on bacteria is due to the formation of complex bonds with the transition metals in the enzyme, which changes the activity of the enzyme and inhibits the metabolism of bacteria (Pillay and Lin, 2013).

By measuring the concentration of NO₂⁻ and NH₄⁺ in the medium (Figs. 3b, 3c), it was found that NO₂⁻ decreased along with NH₄⁺ concentration increased inversely. This process was very slow, and the presence of NO₂⁻ directly affected the inhibitory effect on SRM. In this stage, NH₄⁺ is generated continuously. When NO₂⁻ was absent, NH₄⁺ was maintained at relatively stable content. Although the toxicity of NH₄⁺ is not so strong as NO₂⁻, it still needs to be controlled.

In this study, as the toxic NO₂⁻ was used up, SRM recovered to reduce SO₄²⁻ to H₂S. This result indicated that NO₃⁻ inhibition on SRB activity was reversible. As shown in Fig. 2, all SO₄²⁻ was reduced along with H₂S generation within 1 – 2 d. The concentration of H₂S reached up to 340 – 512 mg L⁻¹ in the logarithmic phase. Under the inhibition condition with 7 mmol L⁻¹ nitrate, the SO₄²⁻ reduction efficiency reached 56.47%, which was 13.45% lower than the control.

3.3 Changes of Community Structure

16S rRNA gene-based high-throughput sequencing was performed for the mariculture sediments, enriched SRM and NO₃⁻ (7 mmol L⁻¹) inhibited SRM to explore the changes in community structure. The Alpha diversity index (Shannon, Simpson, Chao1, ACE) of different samples was analyzed (Table 1). The community richness and diversity index of SRM decreased compared with the mariculture sediments. However, The NO₃⁻ inhibition showed little effect on the richness and diversity index of the bacterial community.

Fig. 4 shows the microbial distribution and difference of three samples at the phylum level. Halobacterota and Firmicutes accounted for a relatively high proportion in mariculture sediments, with a relative abundance of 18.39% and 16.09%, respectively. Desulfobacteria accounted for a relatively low proportion of 3.86%. The abundance of Desulfobacteria was obviously improved as enrichment, it reached 71.51%. The abundance of Firmicutes also increased by enrichment and accounted for 22.6%. Desulfobacteria and Firmicutes contain species that are able to reduce sulfate (Kumar et al., 2017; Yan et al., 2022). Fig. 5a shows that most of the SRB in the SRM belonged to the genera Desulfovibrio and Halodesulfovibrio, with a relative abundance of 52.02% and 6.05%, respectively. Desulfovibrio spp. were typical sulfidogens and distributed widely in an anoxic or anaerobic environment (Meyer et al., 2013). The high salinity was exactly conducive to the survival of the Halodesulfovibrio spp., thus the abundance of this genus also showed a high level.

When 7 mmol L⁻¹ NO₃⁻ was added, an abundance of the recent proposed phylum Desulfobacterota decreased from 71.51% in SRM to 58.06% in nitrate inhibited SRM (Fig. 4). Obviously, the addition of nitrate inhibited the growth of SRB in phylum Desulfobacteria. At the genus level (Fig. 5), the abundance of Halodesulfovibrio and Propionigenium decreased significantly with nitrate inhibition. However, the abundance of Desulfovibrio and Citrobacter was remarkably increased.

Halodesulfovibrio spp. use different forms of SO₄²⁻ as electron receptors in the high salt environment (Shivani et al., 2017). The strain Halodesulfovibrio spirochaetisodalis JC271^T contains 6 genes related to sulfite reduction (Shivani et al., 2017). With the inhibition of NO₃⁻, the abundance of Halodesulfovibrio has decreased significantly, which indicates this genus was inhibited by the high nitrite (Fig. 3b). The specific mechanism of inhibittion remains to be further studied. Interestingly, Propionigenium spp. were reported that they were not able to reduce sulfate, sulfur, or nitrate (Schink and Pfennig, 1982), however, they were highly enriched by the sulfate dosage and inhibited by the nitrate dosage. This should be well investigated in our following research.

However, the abundance of Desulfovibrio and Citrobacter increased with the nitrate dosage comparing with other genera. Citrobacter spp. were reported to reduce nitrite by which they might avoid to be inhibited. Desulfovibrio species generate H₂S by reducing sulfates, sulfites, thiosulfate, or sulfur. Haveman et al. (2005) showed that Desulfovibrio vulgsris was different from other SRB strains, it lacked nitrate reductases and thus could not reduce nitrate. Meanwhile, Korte et al. (2014) revealed that with a high concentration of nitrate, nitrate consumption by Desulfovibrio vulgsris was not detected over the course of the experiment. Moreover, the gene encoding the putative Rex transcriptional regulator (DVU0916/ Dde_2702), as well as a cluster of genes (DVU0251-DV U0245/Dde_0597-Dde_0605) that is poorly annotated, is the key for Desulfovibrio vulgsris adapting better to the environment with the NO₃⁻. Korte et al. (2015) proposed that the inhibition effect of nitrate and nitrite on Desulfovibrio were two independent systems. Desulfovibrio were not able to utilize the nitrate due to the absence of relating enzyme system, while they were able to reduce nitrite by nitrite reductase or consume nitrite as electron acceptors to relieve the inhibition of nitrite. Finally, it is speculated that Desulfovibrio in the SRM adapts better to the environment containing nitrate/nitrite than other SRB species. Thus, Desulfovibrio showed an increase in abundance. Desulfovibrio is the main sulfide producer. In sulfide-inhibition practical application, more attention needs to be paid to Desulfovibrio spp. as they contained different nitrate/nitrite metabolism systems and presented different response.

According to above findings, if Desulfovibrio were not the dominant species, nitrate/nitrite could be dosed to regulate the production of sulfide. However, iron oxide could serve as a replacement if Desulfovibrio predominated in the SRM because it inhibits the formation of sulfide from the maricutulre without species selectivity (Wang et al., 2021).

4 Conclusions

In this study, different concentrations of NO₃⁻ were added to inhibit SRM activity. With the increase of NO₃⁻, the stronger inhibition for the production of sulfide appeared. SRM has the ability of dissimilatory nitrate reduction to ammonium (DNRA), and preferentially uses NO₃⁻ and NO₂⁻ as electronic acceptors. The presence of NO₂⁻ directly impacted the SRM activity. The dsrB gene abundance decreases by 3 orders when NO₂⁻ was present. With the dosage of NO₃⁻, the abundance of Halodesulfovibrio has decreased significantly. However, the abundance of Desulfovibrio was increased. Desulfovibrio contains a mechanism to avoid NO₃⁻/NO₂⁻ inhibition and it adapts better to the environment with the NO₃⁻/NO₂⁻. As Desulfovibrio spp. are the main sulfidogens in some mariculture area, iron oxide could serve as a replacement because it inhibits the formation of sulfide from the maricutulre without species selectivity.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 41977315), and the Fundamental Research Funds for the Central Universities of China (No. 201964004).