The annual mean DSi concentrations of the influent and the effluent to the treatment plant were 235.4 ± 42.8 µmol L−1 (annual mean concentration ± standard deviation) and 193.9 ± 47.6 µmol L−1, respectively (Fig. 1). The DSi concentration of the effluent was significantly lower than the concentration of the influent in 40 of the 46 observations (you-test, p−1 and 12,093 ± 3,542 mol day−1respectively, and the effluent had a significantly lower load than the influent (you-test, p 6 soft year−1 for the influential and 4.43 × 106 soft year−1 for effluents (Fig. 2). On the other hand, the average annual PSi concentrations of the influent and the effluent to the treatment plant were 67.7 ± 23.9 µmol L−1 and 2.5 ± 1.6 µmol L−1, respectively (Fig. 1). In contrast to the DSi concentration, the effluent PSi concentration was significantly lower than the influent concentration in all observations (you-test, p−1 and 153 ± 104 mol day−1respectively, and the effluent had a significantly lower value than the influent (you -test, p 6 soft year−1 for the influential and 5.61 × 104 soft year−1 for effluents (Fig. 2). The annual removal rates of DSi and PSi were 29.5 and 96.9%, respectively. The PSi was of course eliminated by precipitation during the primary treatment with PBS. However, previous studies have shown that DSi concentration and load do not decrease significantly during primary and secondary processing in STPs.14.
Two possible reasons for the decrease in DSi concentration at the PTS of this study are physico-chemical and biological elimination. The coprecipitation of DSi with metal ions, such as magnesium, aluminum, and iron ions, is one factor. Magnesium hydroxide has generally been reported to precipitate particles more effectively under alkaline conditions17. Additionally, metal ions, such as magnesium, aluminum, and iron, at pH 6 and 9, accelerate the polymerization of DSi18. Therefore, the co-precipitation of DSi with metal ions, which are rich in STP influent, may have reduced the DSi concentration of the effluent. Magnesium hydroxide is generally used for neutralizing sulfur oxides and neutralizing wastewater in factories. However, excessive addition of magnesium hydroxide can cause an influx into the STP.
The second factor, biological elimination, consists of the absorption of DSi by bacteria of the genus Bacillus . The spores of species in this genus have a layered structure to protect nucleic acids and proteins, with a layer of nanoscale particulate matter on the outside of the coating layers19. DSi is embedded in these layers19. Bacillus species are ubiquitous in a wide variety of habitats, including soil20fresh water21and marine sediments22. In addition to these natural environments, Bacillus spp. are present in activated sludge used for secondary treatment of wastewater at sewage treatment plants23.24. Bacillus occupies 29 µmol L−1 h−1 of DSi during growth19. According to Murakami et al.25 Bacillusspp. represent 92 to 98% of the total bacteria (5 × 107 at 5 × 10ten mL cells−1) in secondary treatment activated sludge. Specifically, Bacillus thuringiensisis indispensable in STPs due to its ability to degrade starch and oil26. These results suggest that Bacillusbacteria living in the secondary treatment environment of sewage treatment plants absorb DSi and precipitate it into the wastewater influent as sludge, thereby reducing the DSi load in the effluent.
It is unclear from this study whether physico-chemical or biological elimination significantly influences the reduction of DSi in STP. Physico-chemical elimination occurs during the primary treatment process while biological elimination occurs during the secondary treatment process. Therefore, it is possible to identify the main factors by evaluating the changes in DSi concentrations during each treatment process. Additionally, it is still unclear what kinds of changes occur in the STPs that run the advanced processing process. Moreover, the reason why the reduction of DSi in some STP (this study) and not the other14 was unclear (eg, a difference in pH, alkalinity, and activated sludge quality). We need to pay more attention to DSi changes in STPs and evaluate them in more detail.
In the PSi dissolution experiment using the influential PBS (Table 1), the DSi concentration increased by an average of 8.1 ± 6.8% and the PSi concentration decreased significantly by an average of 20, 3 ± 7.5% The total Si concentration (DSi + PSi) was 304.0 ± 35.2 µmol L−1 before incubation and 297.2 ± 42.6 µmol L−1 at the end of the incubation, indicating no significant difference. Dissolution experiments suggest that some of the PSi entering the PTS is soluble, about 20.3%. Therefore, the charge of soluble PSi in the influent STP would be about 0.36 × 106, and most of this soluble PSi should be removed by STP. This amount is comparable to the 20.0% DSi removal amount of 1.85 × 106 soft year−1 by the STP (Fig. 2). Therefore, the removal of DSi and soluble PSi attributable to the construction of the STP may have reduced the DSi load in the watershed by 2.21 × 106 soft year−1.
In this study, the DSi and PSi loads decreased due to the progress of sewage maintenance in the watershed. Thus, the future construction of sewage treatment plants in coastal watersheds causes a “new silica deficiency hypothesis” for coastal waters around the world (Fig. 2). However, at present, time series observations of DSi in many coastal waters are less available than those of nitrogen and phosphorus. Therefore, data on DSi, as an essential parameter of environmental change, must be accumulated continuously from the present moment before any change occurs.