Influence of Alkalinity Enhancement with Olivine or Steel Slag on a Bacterial Community in Activated Sludge Systems

Abstract

A promising strategy to mitigate Carbon dioxide emissions involves the addition of finely ground alkaline minerals to activated sludge systems. However, the release of bioactive substances from these minerals alters the microenvironment of the sludge, with the potential to induce shifts in the bacterial community. In this study, the responses of the bacterial communities in an aerobic activated sludge system to two alkaline minerals (olivine and steel slag) were investigated. This study demonstrated that the addition of olivine and steel slag to activated sludge could selectively shape the microbial community structure. The results indicated a higher diversity of the attached bacterial community in the olivine and steel slag group compared to the glass group. Olivine significantly enriched the bacterial genera that were involved in organic matter degradation and denitrification, such as SC-I-84 and Candidatus Competibacter, thereby potentially enhancing the system’s efficiency in removing carbon and nitrogen pollutants, while the steel slag promoted the growth of iron-dependent denitrifying phosphorus-accumulating bacteria including Thermomonas and Arenimonas, thus establishing a microbial consortium with the potential for nitrogen and phosphorus removal in activated sludge systems. These findings provide crucial evidence for developing mineral–microbe synergistic strategies toward carbon capture and low-carbon sewage treatment.

1. Introduction

Large amounts of Carbon dioxide (CO₂) are emitted as a result of the microbial decomposition of the organic carbon in activated sludge systems. The organic carbon comprises both biogenic carbon (originating from sources such as food and feces) and fossil carbon (originating from sources such as detergents, cosmetics and pharmaceuticals) in sewage [1]. Studies have indicated that fossil organic carbon could constitute up to 28% of the total organic carbon (TOC) in the influent of sewage treatment plants [1,2]. Furthermore, Schneider et al. [3] found that CO₂ emissions resulting from the decomposition of fossil carbon accounted for 11.4–15.1% of the total TOC-derived CO₂ emissions. Fossil carbon originates entirely or partially from petroleum-based chemicals. Conventional activated sludge systems lack effective in situ carbon capture technologies, and CO₂ derived from fossil carbon represents a significant source of greenhouse gas emissions from sewage treatment, which should be included in the greenhouse gas inventory [2]. One promising strategy for effective carbon capture within activated sludge systems is the application of alkaline minerals (e.g., olivine and steel slag) as CO₂ capture agents. In practical applications, distinguishing and selectively capturing CO₂ released from fossil carbon decomposition is nearly impossible. Therefore, the application of alkaline minerals should be designed to capture the total CO₂ emitted from sewage, thereby securing the capture of CO₂ from fossil carbon. Zheng et al. [4] found that the sewage containing activated sludge accelerated olivine dissolution and exhibited a significant increase in total alkalinity under aerobic conditions, and it was estimated that this process could sequester up to 18.8 ± 6.0 teragrams of CO₂ per year globally.

However, alkaline minerals are not inert fillers. Olivine and steel slag are two widely studied alkaline materials with distinct origins and compositions [5]. Olivine is a magnesium silicate mineral predominantly composed of (Mg, Fe)₂SiO₄, which is abundant in the upper mantle [6], while steel slag is an industrial byproduct composed primarily of CaO, SiO₂, and Fe₂O₃ [5]. Dissolution of both materials increases the pH of the solution, creating an alkaline microenvironment and releasing their characteristic ions: olivine primarily releases magnesium (Mg), iron (Fe), and silicon (Si), whereas steel slag mainly releases calcium (Ca), iron (Fe), and silicate (Si), while also releasing trace quantities of heavy metals [5]. The differences result in distinct reaction kinetics and microenvironmental impacts, suggesting that their interactions with microbial communities might be distinct [7,8]. The stability of the microbial community structure was one of the key factors ensuring the efficient removal of pollutants [9]. Zhang et al. [10] demonstrated that pH was a significant environmental factor influencing the distribution of bacterial communities in municipal sewage treatment plants. The microbial community found in sewage, accustomed to low levels of metal ions, had to undergo an acclimation period to tolerate elevated metal ion concentrations [11]. Surface properties of alkaline mineral particles, together with the alkaline microenvironment and the accumulation of metal ions, might reshape microbial community structure, affect adhesion behavior and metabolic pathways, and potentially induce sludge bulking. A previous study showed that the addition of olivine had no discernible influence on bacterial abundance [12]. Ren et al. [13] observed that olivine addition restructured the marine particle-attached bacterial community, elevating the relative abundance of biofilm-forming bacteria, but had no significant influence on the free-living bacterial community. The observation of reacted grains via SEM revealed the presence of bacteria and biofilm coverage [14]. Current findings are primarily based on studies of non-sewage systems, and the ecological impacts of alkaline minerals on activated sludge microbial communities remain poorly understood.

This study investigated how the alkaline minerals reshaped the microbial structure in activated sludge by comparing the differential responses induced by glass, olivine and steel slag. The study addressed the following key questions: (1) which key functional taxa experienced significant abundance shifts; and (2) how the relationships within the community were restructured by the alkaline minerals. The results provide crucial ecological evidence for evaluating environmental risks and selecting appropriate minerals for sewage treatment applications.

2. Materials and Methods

2.1. Experimental Design

Three 20 L polyethylene buckets served as cultivation reactors in the experiment. Six replicates of commercially available glass beads, olivine [15], and steel slag (Table S1) (all with particle sizes of 0.5–1 mm) were prepared and placed in Petri dishes (60-mm-diameter, BKMAMLAB, Changde, China) separately, with the glass group serving as the control. Each material was assigned to a dedicated reactor filled with the collected wastewater, and the Petri dishes were fully submerged during incubation. Each dish occupied a negligible fraction of the reactor volume, and six Petri dishes of each treatment were physically separated within the reactor to minimize potential cross-interference. The incubation lasted 30 days, and samples were collected on days 15 and 30. Throughout the experiment, the sewage in each reactor was completely renewed every 24 h. Sewage used in this study was collected from the aerobic activated sludge tank of the sewage treatment plant of Shandong Urban Construction Vocational College, where treatment was achieved through the anoxic/oxic process and biological aerated filter. The sewage primarily originated from student domestic activities, and the initial pH was 7.00 ± 0.01. At each time point, three Petri dishes were aseptically collected for each material. Residual activated sludge was gently removed by rinsing with ultrapure water, and the samples were transferred into 50 mL sterile centrifuge tubes (BKMAMLAB, Changde, China). All samples were immediately stored at −20 °C for subsequent analyses.

2.2. DNA Extraction and High-Throughput Sequencing

Microbial DNA of the samples (glass, olivine and steel slag) was extracted using the DNeasy PowerSoil Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. The V3–V4 region was amplified using primers 341F (5′-CCTAYGGGRBGCASCAG-3′) and 806R (5′-GGACTACNNGGGTATCTAAT-3′). Sequencing was performed on the Illumina HiSeq 2500 platform. Following quality filtering, the high-quality reads were clustered into amplicon sequence variants (ASVs) using the QIIME2 pipeline (v2025.4). The SILVA database v138 was used as a reference database for taxonomic affiliation of the sequences. Sequences assigned to chloroplasts and mitochondria were removed prior to downstream analysis.

2.3. Statistical Analysis

All statistical analyses in this study were performed using the R language (version 4.3.0). Alpha diversity differences among the three materials were evaluated by ANOVA with Tukey’s HSD applied for multiple-comparison correction. The vegan package was used to calculate the alpha diversity, including Shannon diversity index, richness, and Pielou index. Principal coordinate analysis (PCoA) was conducted based on the Bray–Curtis distance dissimilarity, and the significance of differences between groups was assessed via the analysis of similarities (ANOSIM). Species with significant differences in abundance between different treatment groups were identified by Linear discriminant analysis Effect Size (LEfSe) analysis (LDA score > 2.0).

In order to evaluate the effects of alkaline mineral addition on bacterial interactions, only ASVs that were present in all six samples with a relative abundance higher than 0.1% was retained for network construction. Spearman’s correlation coefficients between ASVs were calculated using the Hmisc package in R. Correlations with an absolute value greater than 0.7 and statistically significant (p < 0.05) were retained for further network construction. The networks were constructed and analyzed using the igraph and ggClusterNet packages. Networks were imported into Gephi (v0.10.1) and visualized with the Frucherman–Reingold algorithms. To identify the topological roles of each node in the network, we calculated the within-module connectivity (Zi) and among-module connectivity (Pi) [16]. According to the threshold values of Zi and Pi, the nodes were divided into four parts: (1) peripherals (Zi ≤ 2.5, Pi ≤ 0.62), indicating nodes that have limited interactions both within and between modules; (2) module hubs (Zi > 2.5, Pi ≤ 0.62), indicating nodes that interact with others within the same module; (3) connectors (Zi ≤ 2.5, Pi > 0.62), indicating nodes that interact with nodes in other modules; and (4) network hubs (Zi > 2.5, Pi > 0.62), indicating nodes that interact with others within and among modules.

3. Results

This study compared the alpha diversity of bacterial communities attached to three materials (glass, olivine, and steel slag). The results demonstrated that at day 15, the Shannon index of both the olivine and steel slag groups was significantly higher than that of the glass group (p < 0.05), while no significant difference was observed between the olivine and steel slag groups. By day 30, the olivine group showed a significantly higher Shannon index than both the steel slag and glass groups (p < 0.05) (Figure 1a). Regarding richness, at day 15, richness was highest in the olivine group, followed by the steel slag and the glass group (p < 0.05); at day 30, the olivine group maintained significantly higher richness than both the steel slag and glass groups (p < 0.05) (Figure 1b). At day 15, the olivine group had significantly greater evenness than the glass group (p < 0.05); however, the steel slag group showed no significant differences in evenness compared to either the olivine or the glass group. By day 30, no significant differences of evenness were detected among the three treatment groups (Figure 1c).

Principal Coordinates Analysis (PCoA) was employed in this study to examine differences among the treatment groups (Figure 2). The results revealed that the first two principal coordinate axes, PCo1 and PCo2, explained 32.65% and 28.67% of the total variation in bacterial community structure, respectively. The separation among the three treatment groups was primarily captured on the PCo1 axis, whereas the variation associated with temporal factors was predominantly reflected along the PCo2 axis. This observation suggests dynamic shifts in the bacterial community structure, driven by the different treatments. Furthermore, ANOSIM confirmed statistically significant differences among treatment groups (r = 0.58, p = 0.001).

A total of 45 bacterial phyla were identified in this study. The primary bacterial taxa included Proteobacteria, Actinobacteriota, Bacteroidota, Chloroflexi, Firmicutes, Acidobacteriota, Patescibacteria, Bdellovibrionota, Verrucomicrobiota, and Myxococcota (Figure 3a). Proteobacteria consistently emerged as the most predominant phylum in all groups. The proportional distribution of these phyla demonstrated temporal dynamics, with distinct successional patterns observed. The compositions of the bacterial communities at genus level are presented in Figure 3b. The predominant genera included PeM15, Run-SP154, Terrimonas, Ottowia, OLB8, Rhodobacter, SC-I-84, Massilia, Thermomonas, Brevundimonas, Ellin6067, Candidatus Competibacter, Denitratisoma, OLB14, Dechloromonas, Saccharimonadales, Devosia, Arenimonas, Nitrosomonas, and Candidatus Microthrix. Distinct dominant genera were observed among the different treatment groups. In the glass group, the dominant genera were PeM15, Massilia, and Terrimonas at day 15 and were succeeded by PeM15, Terrimonas, and OLB8 by day 30. In the olivine group, the dominant genera at day 15 were Run-SP154, PeM15, and Rhodobacter; by day 30, the predominant genera shifted to SC-I-84, Terrimonas, Denitratisoma, and PeM15. In the steel slag group, the dominant genera at day 15 were Run-SP154, Thermomonas, and PeM15; at day 30, the dominant genera were PeM15, Run-SP154, and Brevundimonas.

LEfSe identified significant differences in bacterial communities between the treatment groups at different time points. At day 15, the olivine group was characterized by a higher abundance of the bacterial phylum Myxococcota and the genus Nitrosomonas. The slag group showed an elevated abundance of the phylum Proteobacteria and the genera Run-SP154, Thermomonas, Brevundimonas, and Arenimonas (Figure 4a and Figure S2a). By day 30, no bacterial phylum was significantly enriched in the olivine group, whereas the slag group remained enriched in Proteobacteria. At the genus level, the olivine group was enriched with SC-I-84, Candidatus Competibacter, Denitratisoma, and Candidatus Microthrix. The slag group was enriched with Run-SP154, Thermomonas, Devosia, and Arenimonas. Notably, the glass group did not exhibit a genus that was evidently dominant at either day 15 or day 30 (Figure 4b and Figure S2b).

Bacterial co-occurrence networks were constructed under different treatment conditions (Figure 5). In the glass group, the network comprised 1384 positive correlations and 735 negative correlations, with negative interactions accounting for 34.7%. The average degree of the glass group was 32, and the network exhibited a moderate level of modularity (Modularity = 0.309). The olivine group showed a higher proportion of negative correlations (37.1%), with an average degree of 21 and a modularity of 0.526. The steel slag group demonstrated the highest proportion of negative correlations (47.4%), with an average degree of 30 and the lowest modularity (0.284). Nodes were classified into different topological roles based on within-module connectivity (Zi) and among-module connectivity (Pi) (Figure 6). The glass group contained 8 connector nodes, the olivine group contained 4, and only 1 connector node was identified in the steel slag group.

4. Discussion

Previous studies suggested that the application of alkaline minerals (olivine and steel slag) had the potential to enhance pH and alkalinity, thereby contributing to CO₂ emissions reduction [4,5], which was consistent with the clear pH elevation observed in our study (Figure S1). The suspended alkaline mineral particles in sewage adsorbed onto activated sludge, causing microenvironmental changes that ultimately reshaped the bacterial community and could potentially affect treatment efficiency. This study therefore investigated the microbial feasibility of alkaline mineral application in activated sludge systems.

The addition of alkaline minerals enhanced bacterial community diversity. The minerals provided rougher attachment surfaces compared with glass, which facilitated rapid bacterial colonization. Olivine (silicate mineral) likely generated a chemically milder environment than steel slag (carbonate mineral), creating favorable conditions for a more diverse microbial community.

In the olivine group, the genera Nitrosomonas (Day 15), SC-I-84 (Day 30), Candidatus Competibacter (Day 30), Denitratisoma (Day 30), and Candidatus Microthrix (Day 30) exhibited a significantly higher relative abundance. Previous studies have indicated that the secretion of polysaccharides into the extracellular environment by Candidatus Competibacter accelerated sludge formation [17], and the surface properties of olivine might have facilitated microbial colonization in this study. Candidatus Competibacter was one of the prevalent glycogen-accumulating organisms in activated sludge systems [18]. Although an increase in Candidatus Competibacter might hinder phosphorus removal, a previous study showed that this bacterium played a key role in denitrification, which is an important process in activated sludge systems [19]. SC-I-84 is known as an anaerobic bacterium, and anoxic zones within the biofilm of olivine due to diffusion gradients create a favorable microenvironment for it [20]. Genomic evidence further suggests that SC-I-84 possesses the capacity to produce trehalose, which could potentially enhance resistance to environmental stressors, and was characterized as a partially nitrifying bacterium carrying nitrate reduction genes, such as napA, napB, narV and narY, which could contribute to the effective removal of nitrogen [21]. The presence of a gene set related to the TCA cycle within its genetic map further indicates the potential for carbon source utilization in sewage [22]. The aerobic denitrifier Denitratisoma was more abundant in the olivine group, and previous studies have demonstrated that this genus could directly utilize complex organic matter in sewage as electron donors for denitrification processes [23]. Candidatus Microthrix is one of the most common filamentous bacteria in activated sludge, and its excessive proliferation is a major factor contributing to sludge bulking and deteriorated settleability [24]. It can utilize long-chain fatty acids under both aerobic and anaerobic conditions, and is also able to grow under anoxic conditions by reducing nitrate to nitrite [24,25]. A previous study indicated that the abundance of Candidatus Microthrix correlated strongly and positively with the removal efficiencies of organic matter and total nitrogen, highlighting its potential role in pollutant removal [26]. In our study, olivine likely enhanced the proliferation of Candidatus Microthrix by providing abundant attachment sites and by enabling the formation of biofilm microenvironment characterized by reduced oxygen availability, which might simultaneously promote denitrification and carbon removal while increasing the risk of sludge bulking. Nitrosomonas, a genus of ammonia-oxidizing bacteria, was identified as the dominant ammonia oxidizer within ammonia-rich systems such as activated sludge or biofilm reactors, which contributed to the removal of ammonia in the activated sludge systems [27]. In the slag group, Run-SP154 (Day 15, Day 30), Thermomonas (Day 15, Day 30), Arenimonas (Day 15, Day 30), Brevundimonas (Day 15), and Devosia (Day 30) showed a significantly higher relative abundance. The genera Brevundimonas and Devosia, which predominated in the slag-associated microbial communities, belong to the phylum Proteobacteria. Brevundimonas was isolated in a wide range of matrices and have the metabolic capacity to degrade aromatic compounds [28,29]. Brevundimonas was found in biofilm bacterial communities in rural drainage system ancillary facilities and played a key role in the initial stage of biofilm formation [30]. Devosia, identified as an aerobic nitrifying bacterium [31,32], could produce extracellular polymeric substances that contribute to maintaining the aggregation of activated sludge [33]. Run-SP154 was enriched in the slag group, and it was also highly abundant in the olivine group. Although the bacteria associated with the Run-SP154 sequence have not yet been isolated or identified, this genus plays an important functional role in sewage treatment and is widely distributed in sludge–biofilm composite systems [34]. As a representative denitrifying phosphorus-accumulating bacterium, Run-SP154 removed phosphorus effectively under oxygen-limited conditions [19], and previous analysis showed that Run-SP154 exhibited significant positive correlations with total nitrogen and total phosphorus [35]. The genus Thermomonas was confirmed to be one of the aerobic denitrifiers in sewage treatment, and their abundance increased prominently in the presence of iron [36]. Arenimonas is a mixotrophic denitrifying bacterium [37], and studies have shown that it is adept at both heterotrophic and iron-based autotrophic denitrification processes [38,39]. The iron rapidly released from steel slag enhanced the growth of Thermomonas and Arenimonas. In summary, olivine primarily enriched bacterial taxa with the potential to participate in organic matter decomposition and nitrogen transformation, whereas steel slag selectively stimulated the growth of iron-dependent bacteria involved in the removal of nitrogen and phosphorus. The major divergences in bacterial community structure among the three materials were found during the later stable phase of the experiment, whereas the early phase was characterized by more dramatic fluctuations in the relative abundances of specific genera.

Distinct microbial co-occurrence networks were formed in the glass, olivine, and steel slag groups. The olivine group exhibited the highest modularity, demonstrating that its co-occurrence network could be partitioned into several tightly interconnected modules. This pattern is frequently linked to niche differentiation and potential functional specialization within the microbial community [40]. Furthermore, connectors facilitated interactions between bacteria in different modules within the network. Following the addition of olivine, the bacterial community demonstrated the highest Shannon diversity index and richness. These characteristics indicate that the addition of olivine facilitated the formation of a structurally complex, functionally coordinated, and stable microbial community [41]. In the steel slag group, the proportion of negatively correlated edges was the highest (47.4%), and it exhibited the lowest modularity and the smallest number of connectors. These findings suggest that environmental stress resulting from the rapid dissolution of steel slag might lead to competitive interactions among bacteria [42]. In contrast, the glass group possessed the largest number of connected nodes and the highest average degree, reflecting closer species connectivity. This enhanced connectivity is directly attributable to the low-stress microenvironment on the glass surface.

Alkaline minerals can continuously release heavy metal ions under the combined action of hydraulic shear from sustained aeration and the acidic microenvironment within activated sludge. These ions can be discharged into external water bodies, posing a potential risk to aquatic ecosystems [43,44]. To mitigate this potential ecotoxicological risk, the dosage, particle size distribution, and dosing strategy of alkaline minerals should be systematically optimized based on sewage treatment capacity, influent characteristics, and effluent standards. Furthermore, microbial respiration generates organic acids and CO₂ during the subsequent treatment of dewatered sludge. The resulting acidic environment enhances the dissolution of olivine previously introduced into the activated sludge systems, thereby facilitating additional CO₂ absorption. The addition of olivine and steel slag not only modulated the microbial community structure but also altered the physicochemical characteristics of the residual sludge, raising important considerations for its downstream handling and potential reuse. The enrichment of metal ions such as Mg²⁺, Fe²⁺, and Ca²⁺, together with the possible formation of carbonate-associated phases, suggest opportunities for resource recovery. However, the presence of trace heavy metals remains a key constraint for safe utilization. Future work should therefore include a systematic assessment of long-term leaching behavior and focus on optimizing sludge-treatment strategies to balance treatment performance with safe and sustainable sludge valorization.

5. Conclusions

The application of alkaline minerals to activated sludge systems offers a promising route to reduce carbon emissions from sewage. The dissolution of olivine and steel slag altered the microenvironment of the activated sludge, leading to a restructuring of the bacterial community. These changes increased species diversity and selectively enriched functional taxa involved in carbon degradation and denitrification, while steel slag further promoted the growth of iron-dependent bacteria involved in phosphorus removal. Network analysis revealed contrasting ecological strategies, with olivine enhancing modularity and functional specialization, whereas steel slag intensified interspecies competition. Collectively, these findings elucidate the microbial mechanisms underlying mineral–microbe interactions and provide a microbiological foundation for developing mineral-based, carbon-neutral sewage treatment technology. To capture the dynamics of microbial community adaptation, future studies should incorporate higher-resolution temporal sampling. Furthermore, disentangling the specific contributions of mineral dissolution products to microbial community assembly and function will be critical for optimizing the application strategy.