Performance of aerobic granular sludge for piggery wastewater treatment and its Cu²⁺ and Zn²⁺ removal mechanism

Abstract

Abstract

In this study, actual piggery wastewater was utilized to cultivate aerobic granular sludge (AGS), which was used to remove conventional and heavy metal pollutants from piggery wastewater. The removal mechanism of Cu²⁺ and Zn²⁺ by AGS was investigated using adsorption and high-throughput sequencing methods. The results showed that the extracellular polymeric substance (EPS) components and dominant bacterial species in AGS play important roles in the removal of Cu²⁺ and Zn²⁺. Both dynamic and static factors influence the abundance of microorganisms, which consequently affect the Zn²⁺ adsorption effect of AGS. Microorganisms that are adapted to the presence of heavy metals and possess competitive advantages contribute to the adsorption of Cu²⁺ and Zn²⁺ by AGS. There are positive and negative correlations within the microbial community after AGS adsorption of Cu²⁺ and Zn²⁺. The results of this study are of great significance for promoting the widespread application of AGS technology.

Key points

• The removal of Cu²⁺ and Zn²⁺ mainly depends on the EPS components in AGS

• The biological activity of AGS has a significant impact on the adsorption of Zn²⁺

• Dominant microorganisms play an important role in the adsorption of heavy metals

Supplementary Information

The online version contains supplementary material available at 10.1007/s00253-025-13611-1.

Introduction

With rapid economic development, the demand for livestock and poultry products has increased. By the end of 2021, China’s pig population reached 449.22 million, corresponding to the production of a large amount of piggery wastewater (source: https://www.chinairn.com/hyzx/20220607/160227765.shtml). During the breeding process, antibiotics and heavy metal elements such as Cu²⁺ and Zn²⁺ are mandatorily added to commercial feeds to promote pig growth and prevent disease (Cao et al. 2018; Li et al. 2020a, b). However, due to their extremely low assimilation rates (typically < 10% for Cu and < 15% for Zn in pig metabolism), over 90% of these additives are excreted into wastewater (Zhang et al. 2020). For instance, monitoring data from Chinese pig farms indicate that Cu²⁺ and Zn²⁺ concentrations in pig manure range from 988 to 1027 mg/kg and 1514 to 418 mg/kg, respectively (Shi et al. 2011a, b; Song et al. 2018), far exceeding the levels of other heavy metals. It is worth noting that when such high-concentration heavy metals enter receiving water bodies with flushing wastewater from breeding farms, their dissolved Cu²⁺ and Zn²⁺ concentrations can reach 50–300 mg/L (converted by sewage volume), far exceeding the limits specified in China’s “Pollutant Discharge Standards for Urban Sewage Treatment Plants” (Cu²⁺, 0.5 mg/L; Zn²⁺, 1.0 mg/L). If discharged without effective treatment, they will accumulate through the food chain, causing acute toxic effects and posing long-term threats to aquatic ecosystems and human health (Maszenan et al. 2011; Peng et al. 2019). If they cannot be effectively removed and directly enter receiving water bodies, they can cause long-term harm to human health and the entire ecosystem, as well as impact the reuse of water resources. Therefore, exploring effective methods to reduce the impact of heavy metal pollutants in piggery wastewater on water environments and water resources is an important research topic.

Heavy metal pollutants in piggery wastewater are generally not treated separately for cost control reasons. Therefore, it is necessary to find a low-cost technology that can simultaneously remove conventional pollutants and heavy metals. Biotechnology, due to its wide application range, large adsorption capacity, and inexpensive raw materials, has become the main method considered by researchers (Yan et al. 2024a, b). However, most biosorbents are in the form of suspended microorganisms or biological flocs, which are easily dispersed in the liquid phase. This makes the separation of biological adsorbents from wastewater challenging. Aerobic granular sludge (AGS) is a self-immobilized aggregate of microorganisms characterized by a dense microbial structure and excellent settling performance (Wang et al. 2004). Unlike anaerobic granular sludge (AnGS), which is constrained by strict anaerobic conditions, AGS thrives in aerobic environments, enabling simultaneous organic degradation and heavy metal adsorption—a critical advantage for treating oxygenated piggery wastewater. AGS can withstand the impact load of high-concentration organic wastewater (Derlon et al. 2016; Pronk et al. 2015), and its layered microbial community provides functional redundancy against metal toxicity, a feature absent in conventional biofilms. Additionally, AGS eliminates the need for carrier materials or complex separation units (Gao et al. 2011; Zhang et al. 2019), offering a cost-effective solution for rural pig farms.

At present, the factors affecting the removal effect of heavy metals by AGS include the ion concentration (C₀/X₀) of AGS, pH value, coexisting ions, contact time, and AGS particle size, among others. Liu et al. (Liu et al. 2003) demonstrated that AGS can adsorb the heavy metal Zn²⁺ and that the adsorption capacity is linearly related to the initial concentration of Zn²⁺ (C₀) and the AGS concentration (X₀), a finding that has also been confirmed by other researchers (Xu et al. 2004; Liu et al. 2010). Research has shown that the pH of wastewater affects the adsorption effect of AGS on heavy metals, mainly because pH changes the charge characteristics of heavy metals and the surface potential of sludge (Xu et al. 2006). Sun et al. found that the maximum Cd⁴⁺ adsorption capacity of AGS can reach 348.13 mg/g at a pH of 5.2 (Sun et al. 2011). Studies have shown that when multiple heavy metal ions coexist, the removal effect of AGS on target heavy metal ions decreases (Purba et al. 2020). This is mainly because the affinity of AGS adsorption sites for metal ions varies, and AGS microorganisms exhibit competitive effects in heavy metal ion adsorption (Hao et al. 2016). The particle size of AGS significantly impacts the adsorption of heavy metals. Both excessively large and excessively small AGS particles are not conducive to the adsorption of heavy metals (Hao et al. 2016). Generally, AGS with a smaller particle size has a larger specific surface area, resulting in more effective adsorption sites per unit mass of AGS. However, AGS with a small particle size also produces fewer extracellular polymeric substances (EPS), which significantly impacts the adsorption of heavy metals. AGS with an excessively large particle size usually has a smaller specific surface area, which may lead to a decrease in adsorption capacity, thereby reducing its removal effect for heavy metals. In addition, larger AGS usually has a lower porosity, which affects the internal channels of the AGS and limits the diffusion and adsorption of heavy metals inside the AGS.

Some studies have shown that the adsorption of heavy metals by AGS occurs not only on the particle surface but also in internal regions, which is related to the EPS secreted by AGS. The active groups in EPS, such as carboxylic acid, alcohol, and ether groups, serve as binding sites for heavy metals (Ahn and Hong 2015). Additionally, the removal of heavy metals by AGS involves processes such as ion exchange and chemical precipitation (Xu and Liu 2008). Heavy metal ions may undergo oxidation–reduction reactions on the surface of microorganisms and be removed, while some ions may be removed by precipitation or volatilization (Hao et al. 2016). Furthermore, some researchers have investigated the distribution of microbial communities within AGS. The results showed that tolerant microorganisms such as Propionibacteriaceae, Ochrobactrum anthropi, and Micropruina glycogenica gradually become dominant bacteria in the removal of Cr⁴⁺ by AGS and play an important role in the removal of heavy metals (Wang et al. 2015).

The removal mechanism of heavy metals by AGS is particularly complex for Cu²⁺ and Zn²⁺ due to their divergent adsorption behaviors in real wastewater matrices. Unlike monovalent ions, Cu²⁺ preferentially binds to carboxyl groups in EPS via inner-sphere complexes, while Zn²⁺ favors hydroxyl groups through outer-sphere interactions (Quenea et al. 2009). This competition for binding sites, combined with high organic ligand concentrations in piggery wastewater that form stable complexes with Cu²⁺/Zn²⁺, creates unique challenges for their simultaneous removal (Wang et al. 2024a, b). Currently, researchers primarily study this process through simulation methods in the laboratory, focusing on the behavior and mechanism of AGS in removing single heavy metal ions. This study used actual piggery wastewater as influent to cultivate and domesticate AGS and investigated the changes in sludge performance and pollutant removal effect during the AGS formation process. The key analyses focused on the adsorption kinetics of the heavy metals Cu²⁺ and Zn²⁺ by AGS and the adsorption effects of AGS microorganisms on the heavy metals Cu²⁺ and Zn²⁺, and the heavy metal removal mechanism of AGS was revealed. This research has important practical and scientific significance for in-depth studies on the removal of heavy metal pollutants by AGS. Moreover, the findings are highly relevant to the application of AGS biological treatment methods, which have significant advantages.

Material and methods

Experimental setup and operating conditions

In this study, the sequencing batch reactor (SBR) was constructed from acrylic material, with a diameter and effective height of 50 mm and 550 mm, respectively. The exchange ratio was set at 50%. It operated for eight cycles, and the duration of each stage is detailed in Table 1. The hydraulic retention time (HRT) of the reactor operation was 6 h, with a sludge retention time (SRT) of 25 days. The superficial air velocity within the reactor was maintained at 1.86 cm/s using an air flow meter. The bottom of the reactor was fitted with an aeration head, which could achieve aeration with an aeration rate of 2.5 L/min. During reactor operation, the dissolved oxygen (DO) level was maintained at 2.5–3.0 mg/L, and the temperature was controlled at 25 ± 2 °C. The pre-screened piggery wastewater was pumped to the SBR by a peristaltic pump. The peristaltic pump, air compressor, and outlet solenoid valve were all automatically controlled by microcomputer time-controlled switches.

Table 1.

Operating time of reactor in different operation stage

Reactor operation stage (day)	Substrate filling (min)	Aeration (min)	Settling (min)	Effluent discharge (min)	Idle (min)
Adaptation period (1–5)	6	150	20	1	3
Domesticated period (6–19)	6	165	5	1	3
Stable operation period (20–60)	6	169	1	1	3

Experimental sludge and experimental wastewater composition

The inoculated sludge was brown in color and came from the secondary sedimentation tank of a sewage treatment plant in a city in southern China. The mixed liquor suspended solids (MLSS) and sludge volume index (SVI) of the inoculated sludge were 5.0 g/L and 60 mL/g, respectively. Microscopic observation of the inoculated sludge revealed indicative organisms such as Vorticella, indicating that the inoculated sludge had good activity. The average particle sizes of active and inactive AGS measured in the experiment were 1.0 ± 0.1 mm and 0.9 ± 0.2 mm, respectively, with corresponding average densities of 1.05 ± 0.03 g/cm³ and 1.12 ± 0.05 g/cm³, respectively. According to literature reports, the specific surface area of AGS is 35.2 ± 1.8 m²/g, and the porosity is 0.43 ± 0.02. According to literature reports, the specific surface area (BET) of granule sludge ranges from 7.74 to 23.22 m²/g, with pore volumes (BJH) of 0.0047–0.0192 cm³/g (Wu et al. 2010). The experimental wastewater was pig farm wastewater that had been filtered through a 1.0-mm sieve, which was collected from a pig farm in Zhenjiang, China. The purpose of screening was to remove large floating particles and suspended solids. The concentrations of COD, TP, and NH₄⁺-N were 700 mg/L, 12 mg/L, and 90 mg/L, respectively.

Analytical methods

The water samples were filtered using qualitative filter paper before conducting water quality analysis. The DO and temperature were measured using a dissolved oxygen meter and a thermometer, respectively. The particle size of AGS was measured using the sieving method, and the density was determined by the specific gravity method. The TP, COD, and NH₄⁺-N of the samples were measured using HACH kits and a spectrophotometer (DR/3900). The SVI and MLSS were determined by the standard method (SEPA 2002). The extraction of extracellular polymeric substances (EPS) was performed according to a previous study (Sun et al. 2012). The determination of protein (PN) and polysaccharide (PS) was based on previous studies (Herbert and Fang 1996; Kwang Soo Kim 2014).

Batch adsorption experiments

The concentration ranges of Cu²⁺ (1–15 mg/L) and Zn²⁺ (5–25 mg/L) were selected based on test concentrations slightly exceeding Chinese discharge standards (Cu²⁺, 0.5 mg/L; Zn²⁺, 1.0 mg/L) to assess compliance potential; simulating diluted wastewater after pretreatment (e.g., precipitation) and accidental overload events; and ensuring adsorption site saturation could be observed without confounding effects from metal precipitation (Alyasi et al. 2024; Chen et al. 2022; Pagliaccia et al. 2022).

Take the matured AGS for adsorption experiments of heavy metals Cu²⁺ and Zn²⁺. A blank control group was set up in this process to demonstrate that the adsorption effect of the container on Cu²⁺ and Zn²⁺ can be ignored during the adsorption experiment. During the pre-experimental preparation, the AGS sample needs to be rinsed several times with deionized water to ensure that the AGS was in a complete particle state during the adsorption experiment. Quantitative AGS was placed in Cu²⁺ and Zn²⁺ solutions with concentrations of 5 mg/L and 15 mg/L, respectively, and reacted under both oscillating and non-oscillating conditions. Water samples were collected at predetermined time intervals. AGS with quantitative sludge concentrations were divided into two groups: One group consisted of inactivated AGS; this sludge was thermally inactivated by boiling in deionized water for 10 min to eliminate biological activity while retaining physicochemical adsorption properties. This inactivation process denatured microbial enzymes and proteins but preserved EPS and functional groups critical for metal binding. The other group comprised active AGS; this sludge remained untreated to retain full biological activity. The treated sludge was placed in Cu²⁺ solutions at 1, 2, 5, 10, and 15 mg/L, as well as Zn²⁺ solutions at 5, 10, 15, 20, and 25 mg/L, respectively. After 36 h of reaction, water samples were collected. AGS with sludge concentrations of 1, 5, 15, 20, and 25 g/L were selected and reacted in 5 mg/L Cu²⁺ and 15 mg/L Zn²⁺ solutions. Sampling was done according to the set time, and the initial and residual Cu²⁺ and Zn²⁺ concentrations were measured immediately after filtering with a 0.22-μm membrane.

The Cu²⁺ and Zn²⁺ adsorption quantity was calculated according to Eq. (1):

$$qt=\frac{Co-Ct}{\mathit{MLSS}}$$ 1

q_t is the adsorption capacity of Cu²⁺ and Zn²⁺ at time t (mg/g); C₀(mg/L) is the initial Cu²⁺ and Zn²⁺ concentrations; and Ct (mg/L) is the residual Cu²⁺ and Zn²⁺ concentrations at time t (Shi et al. 2011a, b). This approach aligns with prior studies on AGS-heavy metal interactions (Wei et al. 2019; Seyedein Ghannad and Lotfollahi 2019), where similar concentration ranges were used to avoid microbial inhibition while capturing adsorption capacity thresholds.

DNA extraction and microbial community analysis

After the batch adsorption experiment, AGS samples were taken and high-throughput detection technology was used to analyze the impact of heavy metals on the distribution of AGS microbial communities. S_Oscillation and S_{Non oscillation} represent AGS sludge samples after adsorbing heavy metals Cu²⁺ and Zn²⁺ under oscillating and non-oscillating conditions, respectively; Samples S_blank, S_1-A, S_1-B, S_1-C, S_1-D, and S_1-E represent AGS samples after adsorbing different concentrations of Cu²⁺ (0, 1, 2, 5, 10, 15 mg/L) and Zn²⁺(0, 5, 10, 15, 20, 25 mg/L), respectively; S_2-A, S_2-B, S_2-C, S_2-D, and S_2-E represent sludge samples adsorbed with Cu²⁺ and Zn²⁺ at different AGS sludge concentrations (1, 5, 15, 20, 25 mg/L MLSS). The DNA was extracted using a DNA kit (OMEGA E.Z.N.A™), and the extraction steps were carried out according to the kit’s instructions.

Using statistical analysis methods, the community structure of samples at different classification levels was observed. The community structure distribution figure was drawn using R language software. Heatmaps can be used for species, function, or inter-sample similarity clustering, which can cluster high and low abundance species or functions into blocks and reflect the similarity and differences in community composition or function among multiple samples at different taxonomic levels through color gradients and similarity levels. The heatmap was drawn using R’s gplots package. Correlation analysis is a classic method used to analyze the interactions within microorganisms, which can identify items with significant, strong, positive, and negative correlations within microbial communities. Select species with an abundance higher than 1% for bilateral testing during analysis. The correlation coefficient and p-value between species were calculated using SparCC, and the correlation matrix heatmap was plotted using R’s corrplot package.

The accession numbers of sequences including all described strains are shown in Table 2.

Table 2.

Results of sequences of 16S rDNA

Genes	Accession numbers	Genes	Accession numbers
Neomegalonema	PV037472	unclassified_Betaproteobacteria	PV037493
Chryseolinea	PV037473	unclassified_Chitinophagaceae	PV037494
Thauera	PV037474	unclassified_Proteobacteria	PV037495
Nitrosomonas	PV037475	unclassified_Xanthomonadales	PV037496
Comamonas	PV037476	unclassified_Pirellulales	PV037497
Pelomonas	PV037477	Zoogloea	PV037498
Acidovorax	PV037478	Aquincola	PV037499
Pseudoxanthomonas	PV037479	Runella	PV037500
Acetobacteroides	PV037480	Paraburkholderia	PV037501
Reyranella	PV037481	unclassified_Comamonadaceae	PV037502
Amaricoccus	PV037482	unclassified_Alphaproteobacteria	PV037503
Sediminibacterium	PV037483	unclassified_Clostridiaceae_1	PV037504
Bdellovibrio	PV037484	unclassified_Isosphaeraceae	PV037505
Reyranella	PV037485	unclassified_Gemmataceae	PV037506
Flavihumibacter	PV037486	unclassified_Clostridiales	PV037507
Hydrogenophaga	PV037487	unclassified_Bacteria	PV037508
Devosia	PV037488	unclassified_Bacteroidales	PV037509
Kofleria	PV037489	Niveibacterium	PV037510
Gemmobacter	PV037490	Roseibacillus	PV037511
unclassified_Bacteroidetes	PV037491	Clostridium_sensu_stricto	PV037512
unclassified_Rhodobacteraceae	PV037492	Ravibacter	PV037513
unclassified_Gammaproteobacteria	PV037514

Results

Impact of AGS on pollutant removal performance

The changes in sludge morphology during the cultivation of AGS using actual piggery wastewater are shown in Fig. S1, and the changes in MLSS and SVI during the sludge granulation process are shown in Fig. S2. The relevant data analysis is presented in Text S1. The removal of conventional pollutants such as COD, NH₄⁺-N, TN, and TP during the sludge granulation process was studied, and the results are shown in Fig. 1. As shown in Fig. 1a, the COD removal efficiency showed a stable upward trend. In the early stage of SBR operation, selective pressure flushing of the sludge was implemented, resulting in low biomass, and the COD removal efficiency was only approximately 75% at this time. The COD removal efficiency increased with the gradual formation and maturation of AGS and ultimately exceeded 95%, with an effluent concentration of less than 10 mg/L. The removal effects of AGS on TN and NH₄⁺-N are shown in Fig. 1b and c. In the early stage of reactor operation (1–11 days), the removal efficiency of NH₄⁺-N was poor, and the removal efficiency was maintained between 45 and 50%. The AGS gradually matured as the SBR operation time increased, and the NH₄⁺-N removal efficiency tended to stabilize, with the removal efficiency basically remaining within 80.0% and reaching over 95.0% in the later stage.

Fig. 1.

Removal efficiency of pollutants in reactor during different operation phases: a COD; b NH₄⁺-N; c TN; d TP

As shown in Fig. 1c, the removal efficiency of TN was not as high as that of NH₄⁺-N, and the final removal efficiency remained at approximately 70%. It could be explained by insufficient carbon availability for denitrifying bacteria, coupled with inadequate development of anoxic microenvironments within aerobic granules, both of which synergistically limited the denitrification process. Additionally, the complex composition of organic nitrogen in piggery wastewater (e.g., urea, proteins) could delay ammonification (Kuolin et al. 2018), indirectly exacerbating the TN removal bottleneck by reducing bioavailable NH₄⁺-N for subsequent nitrification–denitrification pathways. The removal effect of TP is shown in Fig. 1d. The figure shows that the removal effect of TP was poor. After 31 days of reactor operation, the removal efficiency of TP was only approximately 50%. With increasing reactor operation time, the removal efficiency of TP increased and ultimately remained within 70%.

Removal of Cu²⁺ and Zn²⁺ by AGS

The removal effects of AGS on Cu²⁺ and Zn²⁺ during the cultivation process are shown in Fig. 2a and b. The average influent concentrations of Cu²⁺ and Zn²⁺ during reactor operation were 5.0 ± 0.2 mg/L and 15.1 ± 0.2 mg/L, respectively. In the first and second stages, the average concentrations of Cu²⁺ in the effluent were 2.5 ± 0.1 mg/L and 1.8 ± 0.1 mg/L, with corresponding removal efficiencies of 49% ± 1 and 64% ± 1, respectively. During this period, AGS had not yet formed, and the sludge exhibited only granulation (Fig. S1b and c), so the removal effect of Cu²⁺ was not ideal. As AGS gradually formed and matured (Fig. S1d, e, and f), the removal efficiency of Cu²⁺ gradually increased. In the third stage, the removal efficiency of Cu²⁺ significantly increased to an average of 79.8%, and this removal effect remained stable throughout the fourth stage. The average effluent concentrations of Zn²⁺ in the four stages were 6.4 ± 0.2 mg/L, 4.1 ± 0.1 mg/L, 2.7 ± 0.3 mg/L, and 2.8 ± 0.2 mg/L, and the average removal efficiencies were 58% ± 1, 72% ± 1, 82% ± 1, and 81% ± 1, respectively. The removal trends of AGS for the two heavy metals were similar, and the removal efficiencies gradually increased with the formation of AGS. However, AGS had a better removal effect on Zn²⁺ than Cu²⁺.

Fig. 2.

The removal efficiency of AGS on Cu²⁺ and Zn²⁺: a, b the concentration and removal efficiency of Cu²⁺ and Zn²⁺ in the effluent; c, d the concentrations of Cu²⁺ and Zn²⁺ in proteins and polysaccharides of AGS

Adsorption of Cu²⁺ and Zn²⁺ by AGS

This experiment was designed to investigate the adsorption effects of mature AGS on Cu²⁺ and Zn²⁺. AGS activity has different adsorption effects on organic compounds. To investigate whether AGS activity has an impact on the adsorption of heavy metals, two groups of AGS samples of equal mass, with one group comprising active AGS and the other group comprising inactive AGS (the specific treatment measures are described in Section “Batch adsorption experiments”), were used in adsorption experiments with different concentrations of Cu²⁺ and Zn²⁺. The results are shown in Fig. 3a and b, respectively. Figure 3a and b indicate that the adsorption capacity of AGS for Cu²⁺ and Zn²⁺ is positively correlated with the concentrations of these two metal ions, indicating that the adsorption of Cu²⁺ and Zn²⁺ by AGS is strongly influenced by the heavy metal concentrations. Figure 3a demonstrates that the adsorption of Cu²⁺ by both active and inactive AGS was essentially equivalent, with a maximum difference of 0.003 mg/g. Figure 3b shows that the adsorption difference between active and inactive AGS for Zn²⁺ exhibited a concentration-dependent nonlinear trend within the tested range (5–25 mg/L). The maximum difference (0.063 mg/g) was observed at 15 mg/L Zn²⁺, followed by a sharp decline to 0.007 mg/g at 20 mg/L and a subsequent rebound to 0.052 mg/g at 25 mg/L. The experimental results indicate that AGS activity has little effect on the adsorption efficiency of Cu²⁺ but has an impact on the adsorption efficiency of Zn²⁺.

Fig. 3.

Adsorption law of AGS on Cu²⁺ and Zn²⁺: a, b the adsorption effect of AGS activity on different concentrations of Cu²⁺ and Zn²⁺; c the dynamic and static adsorption laws of AGS on Cu²⁺ and Zn²⁺; d the adsorption effect of different wet sludge concentrations on Cu²⁺ and Zn²⁺

Subsequently, static and dynamic adsorption experiments were conducted for Cu²⁺ and Zn²⁺ adsorption by AGS. The results are presented in Fig. 3c. The results indicate that the adsorption efficiency of AGS for Cu²⁺ remained relatively consistent under both static and dynamic conditions. There was a noticeable reduction in the adsorption of Cu²⁺ under both conditions when the adsorption time reached 8 h. The adsorption capacity of Cu²⁺ increased and then gradually stabilized after 24 h. In the later stage, its static adsorption effect was slightly better than its dynamic adsorption effect. The adsorption effect of AGS on Zn²⁺ was significantly different from that of Cu²⁺. Figure 3c shows that the dynamic adsorption effect of Zn²⁺ was better than the static adsorption effect in the first 24 h of the reaction. After 24 h of reaction, the static adsorption of AGS for Zn²⁺ increased with increasing reaction time and gradually stabilized after 96 h, with a maximum adsorption capacity of 0.46 mg/g. The dynamic adsorption of Zn²⁺ decreased with increasing reaction time after 48 h, and the final adsorption amount was only 0.04 mg/g. This reduction may be attributed to desorption from AGS during the reaction, resulting in a decrease in the adsorption of Zn²⁺ (Wu et al. 2019). Zn²⁺ binds to hydroxyl groups through outer-sphere interactions (Du et al. 2020), which aligns with its dynamic adsorption advantage under agitation (Fig. 3c). The adsorption effects of different AGS concentrations on Cu²⁺ and Zn²⁺ were investigated, and the results are shown in Fig. 3d. The adsorption capacities of both Cu²⁺ and Zn²⁺ increased and then decreased with increasing AGS concentration. When the mass concentration of AGS was 15 g/L, the adsorption capacities of AGS for Cu²⁺ and Zn²⁺ reached their maximum values, which were 4.5 mg/g and 8.9 mg/g, respectively. Bioactivity plays a critical role in Zn²⁺ removal, with active AGS showing higher adsorption than inactivated sludge at 15 mg/L Zn²⁺ (Fig. 3b), driven by ATP-dependent ZntA transporters and Zoogloea-dominated microbial communities, which secrete cysteine-rich metallothioneins for extracellular sequestration (Costa et al. 2023). The hierarchical adsorption mechanism progresses from initial physical/ion exchange dominance to intermediate EPS complexation, culminating in bioactivity-driven Zn²⁺ removal. These findings align with recent advances in AGS-metal interactions (Li et al. 2022).

Changes in the microbial community in AGS after the adsorption of Cu²⁺ and Zn²⁺

The changes in microbial communities after AGS adsorption of Cu²⁺ and Zn²⁺ under different conditions were investigated, and the results are shown in Fig. 4 and Fig. S3. Fig. S3 shows dilution curves indicating whether the sequencing depth was sufficient and whether the diversity of the samples reached a stable state. The curves of all the samples eventually tended to flatten, indicating that the sequencing depth was adequate (Liu et al. 2023). The Shannon evenness index of all samples ranged from 0.45 to 0.65, indicating that the microbial community in the AGS samples was relatively stable and that the competitive and symbiotic relationships between species were relatively balanced (Liu et al. 2023). Among the samples, the blank sample had the highest Shannon evenness index, which indirectly indicates that heavy metals interfere to varying degrees with the AGS microbial community, reducing the Shannon evenness indices of the AGS samples after heavy metal adsorption relative to that of the blank sample.

Fig. 4.

Variations in microbial communities at the genus level after AGS adsorption of Cu²⁺ and Zn²⁺ under different conditions: a vibrated and non-vibrated conditions; b different concentrations of heavy metals; and c different sludge concentrations

Figure 4a shows the distribution of microbial communities in AGS at the genus level. Figure 4a shows that the microbial species in the AGS samples that adsorbed Cu²⁺ and Zn²⁺ under vibration and nonvibration conditions were essentially the same, but the two conditions had a significant effect on microbial abundance. The relative abundances of 15 microorganisms, including Neomegalonema, Chryseolinea, and Comamonas, were significantly higher under vibration than under nonvibration, whereas the relative abundances of 9 microorganisms, including Thauera, Nitrosomonas, and Zoogloea, were significantly higher under nonvibration than under vibration. As shown in Fig. 3c, the two conditions had little effect on the adsorption of Cu²⁺ but had a greater effect on the adsorption of Zn²⁺.

Figure 4b shows the distribution of microbial communities in AGS after the adsorption of different concentrations of heavy metals. Compared with the blank AGS sample, the AGS samples with adsorbed heavy metals contained microorganisms with gradually decreasing relative abundances, such as Pseudoxanthomonas, Hydrogenophaga, and Kofleria. Some microorganisms, such as Pelomonas and Sediminibacterium, showed gradually increasing relative abundances. In addition, the relative abundance of some microorganisms, such as Neomegalonema, Sphingopyxis, and Nitrosomonas, first increased and then decreased. Correspondingly, the relative abundance of microorganisms such as Thauera, Zoogloea, and Bdellovibrio first decreased and then increased. Finally, some microorganisms, such as Chryseolinea, Acidovorax, and unclassified_Betaproteobacteria, showed a relatively constant abundance.

Figure 4c shows the distribution of microbial communities within AGS after the adsorption of Cu²⁺ and Zn²⁺ at different sludge concentrations. Figure 4c shows that the distribution of microorganisms was closely related to the AGS concentration and its adsorption of heavy metals. The relative abundance of Zoogloea in the blank sample was only 5.65%. However, after AGS adsorbed heavy metals, the relative abundance increased significantly; in particular, when the AGS concentration reached 20 g/L, the relative abundance of Zoogloea reached 45.44%, indicating that Zoogloea has a strong tolerance and adsorption capacity for heavy metal ions and can adapt to the environmental pressure caused by the presence of heavy metals, leading to a significant increase in its number (Liang et al. 2023). The increases in the relative abundances of genera such as Thauera, Nitrosomonas, and Aquincola were more affected by the increase in AGS concentration. In contrast, the relative abundances of unclassified_Bacteroidetes and unclassified_Betaproteobacteria significantly decreased after heavy metals were adsorbed, indicating that heavy metals are highly toxic to these microorganisms (Nguyen et al. 2019; Zhang et al. 2023). Similar microorganisms include Neomegalonema, Chryseolinea, and Pseudoxanthomonas. Similarly, for some microorganisms in the samples, the relative abundance first decreased and then increased (Bdellovibrio) or increased first and then decreased (Acidovorax). This may be due to the influence of heavy metals on the metabolic activity of microorganisms or to the competition between microorganisms for nutrients, space, and energy (Aponte et al. 2020; Olaniran et al. 2013). Under heavy metal exposure, the number of microorganisms with competitive advantages increases, while the number of microorganisms with competitive disadvantages decreases. It can be seen that the adsorption of heavy metals by AGS is closely related to the microorganisms within the AGS, and microorganisms that are adapted to the presence of heavy metals and have competitive advantages contribute to the effective adsorption of Cu²⁺ and Zn²⁺ by AGS. Figure 3d shows that the adsorption effect of AGS on heavy metals was not improved by simply increasing the sludge concentration, which is somewhat related to the decrease in abundance and extinction of certain microorganisms within the AGS.

Correlation analysis between microorganisms after AGS adsorption of heavy metals

Figure 5a shows a heatmap of the dominant species in the samples. The figure reflects the similarities and differences in microbial community composition at the genus level among multiple samples through color gradients and similarity degrees. The darker (more red) the color is, the higher the relative abundance of a species, and the lighter (more green) the color is, the lower the relative abundance of a species. The microbial distribution of all the samples in Fig. 5a can be divided into three regions: I, II, and III. Region I corresponds to the relatively high-abundance genera of bacteria in all samples and includes Thauera, Zoogloea, and Neomegalonema, as well as unclassified_Bacteroidetes and unclassified_Rhodobacteraceae genera that cannot be cultured. The succession pattern of microorganisms is clearly displayed in region II.

Fig. 5.

The correlation between microorganisms after AGS adsorption of Cu²⁺ and Zn²⁺: a heat map of dynamic changes in microbial communities; b corrplot of microorganisms’ correlation

The genera with higher relative abundances in all samples were taken as examples to analyze the correlations between microorganisms, and the results are shown in Fig. 5b. The genera that were positively correlated with Thauera included unclassified_Betaproteobacteria, Bdellovibrio, and Nitrosomonas, and their positive correlations with Thauera decreased in sequence. These findings suggest that changes in the abundance of the Thauera bacterial community are correlated with fluctuations in the populations of these other communities. This finding is consistent with the results shown in Fig. 4c, as the evolutionary patterns of Thauera and Bdellovibrio are aligned. This correlation indicates the potential for a mutually beneficial or symbiotic relationship between Thauera and these other communities, which may rely on similar environmental conditions or resources for survival (Nelson and May 2017). The correlation coefficients of Thauera with unclassified_Betaproteobacteria and Bdellovibrio were greater than 0.9, indicating a strong correlation between these genera. These findings suggest that the functional changes in Thauera are closely related to these two microorganisms. Conversely, the genera that have a negative correlation with Thauera include unclassified_Clostridiales, Clostridium_sensu_stricto, and Acetobacteroides, with their negative correlations with Thauera decreasing in sequence. This finding indicates that the increase in the number of Thauera genera is related to the decreases in the number of these negatively correlated genera. There may be a competitive relationship between Thauera and these genera, such as competition for the same resources, or the growth of Thauera may inhibit the growth of these other genera (Hibbing et al. 2010).

Figure 5b shows that other bacterial genera may be either positively correlated or negatively correlated. For example, the positively correlated genera of Zoogloea include unclassified_Betaproteobacteria, Thauera, Gemmobacter, etc., and the negatively correlated genera include Neomegalonema, unclassified_Xanthomonadales, Devosia, etc. The genera positively correlated with Neomegalonema include unclassified_Alphaproteobacteria, unclassified_Xanthomonadales, and Devosia, while the negatively correlated genera include Zoogloea, Gemmobacter, and Nitrosomonas. Pseudoxanthomonas was positively correlated with genera such as unclassified_Bacteroidetes, Kofleria, and Devosia, but it did not have any negatively correlated genera. This may be because Pseudoxanthomonas and other microorganisms occupy different ecological niches in environments where heavy metals are present, and this niche differentiation can reduce competition, allowing different populations to coexist (Kubota et al. 2022).

Discussion

Analysis of the removal of COD, NH₄⁺-N, TN, and TP

The improvement of COD removal efficiency is related to the microorganisms in AGS. Microbial activity increased as the AGS gradually matured, which played an important role in the degradation of COD (Pérez-Bou et al. 2024). The SEM image shows that the mature AGS surface was uneven (Fig. S1h), which resulted in a large specific surface area and facilitated the adsorption of organic matter on the AGS surface, ultimately enabling the removal of organic matter through biodegradation. NH₄⁺-N in wastewater is usually removed through nitrification and denitrification reactions. In this process, nitrifying bacteria convert NH₄⁺-N into nitrite and nitrate, and finally, denitrifying bacteria convert nitrate into nitrogen (Zou et al. 2024). However, the AGS had not yet formed in the early stage of reactor operation, and the microbial community in the AGS had not reached a stable equilibrium state, lacking nitrifying bacteria and denitrifying bacteria, which resulted in a low NH₄⁺-N removal efficiency during the initial operation stage of the reactor. In the later stage of reactor operation, the AGS gradually matured. The unique spherical structure of AGS, which has external aerobic and internal hypoxic conditions, provides an ideal growth environment for various nitrogen removal microorganisms (Jingfeng et al. 2007), thereby promoting NH₄⁺-N removal.

The experimental results revealed that when actual piggery wastewater was used to cultivate AGS in an SBR, the removal effects of COD and NH₄⁺-N in the wastewater were good during the sludge granulation process, but the removal effects of TN and TP were not ideal. This may be related to the competition among microbial populations in AGS. The removal of TP depends on absorption by polyphosphate-accumulating organisms (PAOs) and the accumulation of biological phosphorus, while the microbial populations in AGS mainly target the removal of NH₄⁺-N, which results in a lower removal efficiency of TP (Shi et al. 2024). Additionally, the influent contained heavy metal pollutants, which can inhibit the activity of PAOs, ultimately affecting the removal effect of TP by AGS (Kedves and Konya 2024).

Analysis of removal of the Cu²⁺ and Zn²⁺

The removal mechanism of heavy metals by AGS is relatively complex. Some studies have shown that the removal of heavy metal ions by AGS is mainly achieved through EPS adsorption, ion exchange, metal chelation, chemical precipitation, and so on (Purba et al. 2020). The organic complexation of microbial cells and EPS with heavy metal ions is an important way for AGS to adsorb heavy metal ions (Xu et al. 2006). Figure 2c and d show the changes in the contents of the two heavy metals in the sludge EPS samples from the reactor at different operating stages. As shown in Fig. 2c, the contents of Cu²⁺ in proteins and polysaccharides gradually increased with increasing reaction time, while the content of Cu²⁺ in proteins was significantly higher than that in polysaccharides, indicating that the adsorption of Cu²⁺ by EPS mainly depends on protein components, possibly because proteins contain atoms such as O, N, and P that can form chelates or complexes with heavy metal ions and thereby adsorb heavy metals (Purba et al. 2020). The dominant chemical adsorption occurs via coordination complexation with EPS, where Cu²⁺ preferentially forms inner-sphere complexes with carboxyl groups (Ai et al. 2019), as evidenced by the higher Cu²⁺ adsorption in EPS proteins compared to polysaccharides and a certain linear correlation between EPS content and adsorption capacity. Figure 2d shows that the variation trends for Zn²⁺ in proteins and polysaccharides were similar to the trend for Cu²⁺, but the Zn²⁺ concentrations in proteins and polysaccharides were higher than those of Cu²⁺. This is mainly due to the differences in chemical properties between Zn²⁺ and Cu²⁺. The radius of Zn²⁺ is larger than that of Cu²⁺, which makes its metallic bond weaker than that of Cu²⁺. When metallic bonds are weaker, metal atoms are more likely to lose electrons and exhibit stronger chemical reactivity (Hao et al. 2016; Jagaba et al. 2024). The results also indicated that AGS had a better removal effect on Zn²⁺ than Cu²⁺ due to the effects of EPS.

Removal mechanism of Cu²⁺ and Zn²⁺ by AGS

As shown in Fig. 3b, the adsorption effect of active AGS on Zn²⁺ was better than that of inactive AGS. At low concentrations (< 10 mg/L), active AGS prioritized energy allocation toward basal metabolism rather than metal adsorption, resulting in marginally lower adsorption capacity compared to inactive AGS (− 0.008 mg/g difference at 5 mg/L), likely due to Zn²⁺ efflux mechanisms in metabolically active cells (Cabral et al. 2016). As concentrations increased to 10–20 mg/L, ATP-dependent transport systems and enhanced secretion of EPS enriched with thiol groups synergistically improved active adsorption, achieving peak performance at 15 mg/L. However, this advantage diminished at 20 mg/L as energy resources were redirected to detoxification pathways such as metallothionein synthesis. Beyond 20 mg/L, stress-induced adaptive responses dominated (Yan et al. 2024a, b), which restored adsorption superiority at 25 mg/L. This triphasic behavior highlights the dynamic trade-off between microbial survival strategies and functional performance under escalating metal stress, governed by threshold-activated metabolic regulation (Zhu et al. 2024).

The removal of heavy metal ions by AGS is generally the result of multiple mechanisms working together. Zn²⁺ and Cu²⁺ ions may undergo oxidation‒reduction reactions on the surface of AGS microorganisms and be removed, with Cu²⁺ being more readily adsorbed due to its stronger oxidizing nature (Luo et al. 2016). This may be the reason why the activity of AGS has a small effect on the adsorption of Cu²⁺ and a large effect on the adsorption of Zn²⁺. Cu²⁺ exhibits strong oxidative capacity (standard reduction potential: + 0.34 V, significantly higher than Zn²⁺ at −0.76 V). In the microaerophilic/anaerobic core region of AGS, microorganisms (such as heterotrophic bacteria or sulfate-reducing bacteria) can reduce Cu²⁺ to metallic copper or low-toxicity forms through metabolic activities (such as secreting reducing agents or utilizing electron transport chains), leading to its preferential immobilization (Li et al. 2020a, b; Wang et al. 2022). In contrast, due to its weaker oxidative capacity, Zn²⁺ removed primarily via physicochemical adsorption on microbial cell surfaces (e.g., binding to carboxyl or amino groups in EPS) or intracellular accumulation, making its removal more sensitive to AGS metabolic activity. In addition, the cell walls and EPS of microorganisms (e.g., bacteria, archaea) in AGS contain abundant functional groups (e.g., –OH, –COOH, –NH₂). Due to its higher charge density and selective binding capacity, Cu²⁺ readily forms stable complexes with sulfur/nitrogen-containing groups (e.g., cysteine or histidine side chains), resembling the chelation observed in metalloenzymes (Che et al. 2024). Conversely, Zn²⁺ binds weakly to these functional groups, particularly under dynamic flow conditions, and is more susceptible to environmental fluctuations (e.g., pH, ionic strength) or microbial metabolic byproducts, resulting in active adsorption–desorption processes. This may be the reason for the significant difference in adsorption trends between the dynamic and static adsorption of Zn²⁺ and Cu²⁺ by AGS.

The adsorption effects of different AGS concentrations on Cu²⁺ and Zn²⁺ were investigated, and the results are shown in Fig. 3d. The adsorption capacities of both Cu²⁺ and Zn²⁺ increased and then decreased with increasing AGS concentration. When the mass concentration of AGS was 15 g/L, the adsorption capacities of AGS for Cu²⁺ and Zn²⁺ reached their maximum values, which were 4.5 mg/g and 8.9 mg/g, respectively. The main reason for the significant difference in the adsorption effect of AGS on Cu²⁺ and Zn²⁺ may be that the initial concentration of Zn²⁺ was greater than that of Cu²⁺. Under the same conditions, due to the large amount of Zn²⁺ per unit volume, Zn²⁺ has more opportunities to come into contact with AGS, resulting in the relatively large adsorption capacity of AGS for Zn²⁺. Furthermore, when the concentration of metal ions is high, the driving force for their diffusion to the AGS surface increases, which helps overcome the resistance on the AGS surface and facilitates Zn²⁺ adsorption (Yang et al. 2024). When the mass concentration of AGS exceeded 15 g/L, the adsorption capacity of AGS for Cu²⁺ and Zn²⁺ significantly decreased. This may be because when the AGS concentration is too high, the distance between particles decreases, leading to an increase in the mass transfer resistance of Cu²⁺ and Zn²⁺ between AGS particles, which slows the migration rate of heavy metal ions to the AGS surface and reduces the adsorption capacity of heavy metals (Feng and Huang 2024). In addition, as the concentration of AGS increases and the adsorption sites become saturated, if the concentrations of Cu²⁺ and Zn²⁺ remain unchanged, the amount of heavy metal ions adsorbed per unit specific surface area decreases, resulting in a decrease in the adsorption of Cu²⁺ and Zn²⁺ (Hao et al. 2016). Therefore, when AGS removes heavy metal ions, the sludge concentration needs to be adjusted according to the actual situation, and the operating conditions need to be optimized to maximize the adsorption effect.

The relationship between microorganisms and the removal of Cu²⁺ and Zn²⁺

To further explore the adsorption mechanism of AGS for Cu²⁺ and Zn²⁺, the changes in microbial communities after AGS adsorption of Cu²⁺ and Zn²⁺ under different conditions were investigated, and the results are shown in Fig. 4 and Fig. S3. The experimental results indicate that vibration or nonvibration conditions affect the abundance of microorganisms, which in turn influence the adsorption effect of AGS on Zn²⁺. There are three possible explanations for this phenomenon. First, microorganisms can change the pH value of the surrounding environment through their metabolic activities, which affects the solubility and speciation of Zn²⁺ (Meng et al. 2016). Under these conditions, microbial metabolic products such as organic acids and EPS can chelate with metal ions, influencing the adsorption behavior of Zn²⁺ (Han et al. 2015). Second, there is a biological transformation effect; for example, Thauera, Nitrosomonas, and Zoogloea can convert metal ions into insoluble precipitates (Priyadarshanee and Das 2021). This biological transformation effect is beneficial for the removal of metal ions. When the number of these microorganisms increases, the degree of transformation also correspondingly increases, thereby improving the adsorption effect of Zn²⁺. Finally, the affinity of microorganisms for Cu²⁺ and Zn²⁺ ions is different, and changes in the number of microorganisms may alter their competitive equilibrium, thereby affecting the adsorption effect of AGS on Zn²⁺ (Yin et al. 2019).

The quantitative AGS adsorption results for different concentrations of heavy metals show that the adsorption effect of AGS on Cu²⁺ and Zn²⁺ decreased with increasing metal ion concentration. This is closely related to the succession pattern of microorganisms. Different concentrations of heavy metals have toxic effects on microorganisms, damaging their cellular structures, interfering with their metabolic activities, and leading to a decrease in the number of certain microorganisms, such as Pseudoxanthomonas, Hydrogenophaga, and Kofleria (Giller et al. 1998). Some microorganisms may have a strong tolerance and adsorption capacity for specific heavy metal ions, and the abundances of these microorganisms, such as Pelomonas and Sediminibacterium, may gradually increase in the presence of heavy metals to adapt to environmental pressure (Wang et al. 2024a, b). Under the long-term influence of heavy metals, the microbial community may undergo succession, where certain types of microorganisms are replaced by others that are more adapted to the environment. This succession may lead to an initial increase followed by a decrease in the relative abundance of some microorganisms, such as Neomegalonema, Sphingopyxis, and Nitrosomonas, or a decrease first and then an increase, as observed for Thauera, Zoogloea, and Bdellovibrio.

Figure 5a shows that after AGS adsorbed heavy metals under different conditions, the relative abundance of microbial communities, including common and unculturable bacterial genera, changed significantly, with some increasing and others decreasing in relative abundance. Microorganisms with strong adaptability can grow and reproduce better when heavy metal ions are adsorbed, and there are also cases where the synergistic effect of microorganisms leads to an increase in the number of certain microorganisms while the number of other microorganisms decreases (Giller et al. 1998; Wang et al. 2024a, b). Due to the influence of heavy metals, some microorganisms may proliferate because of enhanced metabolic activity, while others may decrease due to inhibited metabolic activity (Aponte et al. 2020). The adsorption of heavy metals by AGS is closely related to the types and quantities of microorganisms in the AGS. Figure 5b reveals the presence of both positive and negative correlations among microbial communities after the removal of Cu²⁺ and Zn²⁺ by AGS. A positive correlation indicates that the microbial community has high stability, as species can support each other and jointly cope with the adverse environment caused by the presence of heavy metals (Faust and Raes 2012). A negative correlation indicates that different microorganisms compete for advantageous resources for survival and reproduction, which can drive niche differentiation, allowing different species to reduce direct competition and adapt to the environmental conditions where heavy metals coexist (Kubota et al. 2022).

In conclusion, AGS had excellent removal effects on COD and NH₄⁺-N in piggery wastewater, with removal efficiencies that eventually stabilized at over 85.0% and 96.0%, respectively. However, the removal effect of AGS for TN and TP was average, with final removal efficiencies reaching 72.7% and 69.1%, respectively. The removal trends of Cu²⁺ and Zn²⁺ by AGS were similar, and the final average removal efficiencies reached 79.8% and 81.0%, respectively. The removal of Cu²⁺ and Zn²⁺ mainly depends on the EPS components in AGS. The biological activity of AGS had almost no effect on the adsorption of Cu²⁺, with a maximum adsorption capacity difference of 0.003 mg/g. However, the adsorption effect of active AGS on Zn²⁺ was better than that of inactive AGS, with a maximum adsorption capacity difference of 0.063 mg/g. The adsorption effects of AGS on Cu²⁺ and Zn²⁺ were influenced by both the concentrations of heavy metals and the concentration of AGS. When the mass concentration of AGS was 15 g/L, the adsorption amounts of Cu²⁺ and Zn²⁺ on AGS reached their maximum values, which were 4.5 mg/g and 8.9 mg/g, respectively. Vibration and nonvibration conditions affect the abundance of microorganisms, which further affects the adsorption effect of AGS on Zn²⁺. The concentrations of Cu²⁺ and Zn²⁺ significantly affect the microbial succession process in AGS. Microorganisms that adapt to the presence of heavy metals and have competitive advantages, such as Pelomonas and Sediminibacterium, contribute to the adsorption of Cu²⁺ and Zn²⁺ by AGS. There are positive and negative correlations among the microbial communities after the adsorption of Cu²⁺ and Zn²⁺ by AGS. Microorganisms with a positive correlation, such as Thauera and unclassified_Betaproteobacteria, can jointly cope with adverse environments caused by the presence of heavy metals. Microorganisms with a negative correlation, such as Thauera and Acetobacteroides, can reduce direct competition between species due to niche differentiation, ensuring the diversity of AGS microbial populations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

Xiaochun Wang wrote the main manuscript text and prepared figures 1-5. Shutao Xiao, Mingyang Li, Biming Wang and Yun Zhou designed experiments and provided experimental datas. Zhonglin Chen, Xiaolei Zhang and Xiangtong Zhou reviewed the manuscript.

Funding

The work was supported by Open Project of State Key Laboratory of Urban Water Resources and Environment (Grant No. ES202220); the China Postdoctoral Science Foundation (Grant No. 2023M741423); the Youth Program Natural Science Foundation of Jiangsu Province (Grant No. BK20230552); and the Shenzhen Science and Technology Innovation Commission (Grant No. GJHZ20220913143007014).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Competing interests

The authors declare no competing interests.