Occurrence and removal of 25 antibiotics during sewage treatment processes and potential risk analysis

Antibiotics usually encompass a variety of compounds with antibacterial activity, including natural, synthetic and semisynthetic compounds (Sarmah et al. 2006). They can be used to treat human diseases (Qian-Qian et al. 2015), promote growth in animal husbandry, and improve feed efficiency (Dan et al. 2020). Therefore, in the past few decades, to meet human health needs and increase economic benefits, antibiotics have played a vital role in medical treatment (Zhi et al. 2020), animal husbandry (Gaballah et al. 2021) and agriculture (Parente et al. 2019). However, most organisms have incomplete absorption and metabolism of antibiotics (Sarmah et al. 2006). As a result, antibiotics and their metabolite residues can be directly or indirectly discharged into surface water or farmland irrigation water through livestock waste, threatening crop safety and human health (Shi et al. 2020).

With the gradual increase in the global demand for antibiotics (Klein et al. 2018), some scholars estimate that the global consumption of antibiotics is in the range of 100,000–200,000 tons (Cheng et al. 2014). Antibiotics continue to enter the environment in a variety of ways, making them a new organic pollutant that is ubiquitous in aquatic environments. STPs, as channels for domestic sewage discharge, are a more prominent issue in terms of concentrating antibiotic residues. Therefore, the occurrence, removal and transformation of antibiotics in STPs has become a popular issue in related research fields (Chen et al. 2019; Li et al. 2021; Wang et al. 2021).

Current STPs are mainly designed for conventional pollutants and do not take into account the removal of trace pollutants such as antibiotics. Therefore, most sewage treatment processes cannot effectively remove antibiotics from sewage and prevent them from entering water bodies (Yao et al. 2021). For example, in three STPs in Switzerland, the average clarithromycin removal efficiency was 81%, and the average tylosin removal efficiency was 76% (McArdell et al. 2003). The average erythromycin removal efficiency in two STPs in Hong Kong was 43%, and the average ciprofloxacin and aureomycin removal efficiencies were 66 and 83%, respectively (Li & Zhang 2011). As a result, antibiotic and metabolite residues can be detected when most treated sewage is discharged into the aquatic environment (Zhang et al. 2021), resulting in ecological risks to the surrounding environment of aquatic and terrestrial organisms to a certain extent (Morosini et al. 2020).

According to the data, there are studies related to antibiotics in STPs in the eastern and northern regions of China (Zhao et al. 2020; Bao et al. 2021) and fewer studies in the southern part of the country. The two STPs in this study are located in the Jiuzhou River basin in southern China. The aquaculture industry in this basin is developed, and sewage is directly discharged into the Jiuzhou River after treatment. Therefore, two STPs in the Jiuzhou River basin are studied. The occurrence and fate of this study are of great significance for understanding the risks and potential effects of antibiotics in the Jiuzhou River aquatic environment. This study involves the following aspects:

• (1)

  Investigate the occurrence and distribution of antibiotics in the water phase of STP 1 and STP 2 sewage treatment;

• (2)

  Compare the antibiotic removal effects of two different sewage treatment processes;

• (3)

  Conduct a redundancy analysis of antibiotics in the water phase of two STPs, using water quality parameters;

• (4)

  Carry out an ecological risk assessment of the typical antibiotics in STP 1 and STP 2.

The results can provide a reference for antibiotic pollution control and environmental hazard assessments in sewage treatment systems.

During the treatment processes at the two STPs, sewage water samples were collected from different units. The STPs are all located on the Jiuzhou River in southern China. They mainly receive and treat local domestic sewage. The designed treatment flows are 500 m³/d and 1,300 m³/d, respectively. The first STP is a multistage composite mobile bed membrane bioreactor (MBBR) process, and the second is a continuous-flow transmission biological reactor (IBR) process. Water samples were collected from different treatment units in STP 1 and STP 2. The first STP consists of original influent, MBBR tank effluent, artificial wetland effluent (AWE) and final effluent; the second STP consists of original influent, IBR tank effluent, sedimentation tank effluent (STE) and final effluent. The specific sampling locations and sampling points are shown in Figure 1. The sampling plan was carried out in January 2021. During the sampling process, the STPs were in normal operation, and there was no rainfall projected for the near future. For each sampling, 1 L of each water sample was put into brown glass bottles with three repetitions, and the samples were transported to the laboratory within 3 days for preprocessing and chemical analysis.

The 25 target antibiotics were divided into four categories: (1) 10 kinds of quinolones (QNs): ciprofloxacin (CFX), danofloxacin (DNX), difloxacin (DFX), fleroxacin (FRX), marbofloxacin (MBX), norfloxacin (NFX), ofloxacin (OFX), oleandomycin (ODC), sarafloxacin (SFX), and sparfloxacin (SPX); (2) four kinds of macrolides (MLs): erythromycin (ERC), josamycin (JSC), kitasamycin (KSC), and tylosin (TLS); (3) four kinds of tetracyclines (TCs): chlortetracycline (CTC), doxycycline (DCC), oxytetracycline (OTC), and tetracycline (TCC); and (4) seven kinds of sulfonamides (SNs): sulfachlorpyridazine (SPZ), sulfadiazine (SDZ), sulfamethazine (SMZ), sulfamethazine (STZ), sulfamethoxazole (SXZ), sulfaquinoxaline (SQX), and sulfathiazole (SAZ).

All antibiotic analysis standards were purchased from Tianjin Alta Technology limited company. Detailed information and the chemical properties of the 25 antibiotics are listed in Table S1. Methanol (chromatographically pure) and acetonitrile (chromatographically pure) were purchased from Fisher Company in the United States, formic acid (chromatographically pure) was purchased from CNW Company in China, and other material information is listed in Table S2. Ultrapure water was used throughout the experiment, and methanol was used to dilute the antibiotic standard stock solution to prepare standard liquids of different concentrations.

Based on the importance of the analysis step, water samples should be pretreated immediately when returned to the laboratory (Sharip et al. 2020). In the reference research, the preparation method of the water samples was modified (Pan et al. 2020), and the pretreatment steps were as follows: (1) Activation: use 10 mL methanol and 10 mL ultrapure water to carry out the solid phase extraction column activation so that the sample solution can smoothly flow through the solid phase extraction column; (2) Acidification: add H₂SO₄ or NaOH to 200 mL of the water sample to make the pH = 3 ± 0.1; (3) Sample loading: transfer the adjusted water sample, pass it through the solid phase extraction column at a flow rate of 5 mL/min; (4) Rinse: rinse the solid phase extraction column with 10 mL ultrapure water to remove impurities; (5) Dry: add nitrogen to dry the solid phase extraction column; (6) Elution: elute the solid phase extraction column with 3 mL formic acid/methanol mixture; (7) Collection: soak the solid phase extraction column with 3 mL formic acid/methanol mixture for 5 minutes, then use 9 mL formic acid/methanol mixture at a rate of 1 mL/min elute the sample and collect the eluate; (8) Constant volume: concentrate the above eluent to less than 1 mL in a nitrogen concentrator, dilute it to 1 mL in methanol, mix well, and place the water sample in a 2 mL brown sample flask through a 0.22 μm organic nylon filter, and preserve the water sample at −4 °C.

A ZORBAX SB-C18 (2.1 × 100 mm, 3.5 μm) column and ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC–MS/MS) were used to analyse the target antibiotics. Mobile phase A used for QNs and MLs was 0.5% formic acid, and mobile phase B was acetonitrile. The flow rate was 0.5 mL/min. The column temperature was 25 °C, and the injection volume was 5.0 μL. See Table S3 for the gradient elution procedure. The mass spectrometry conditions were the following: capillary voltage 4.0 kV, drying gas temperature 350 °C, and atomizing gas pressure 25 psi. The electrospray positive ion source (ESI+) multiple reaction detection (MRM) mode was used for mass spectrometry data acquisition. The mass spectrometric parameters are shown in Table S4. Mobile phase A used for TCs and SNs was 1% formic acid, and mobile phase B was acetonitrile. The flow rate was 0.3 mL/min. The column temperature was 30 °C, and the injection volume was 3.0 μL. The elution procedure is shown in Table S3, the mass spectrometry conditions were the same as those used for the QNs, and the mass spectrometer parameters are shown in Table S4.

Using the external standard method, ultrapure water was used as a laboratory blank, and blank and label experiments were carried out on the target compound. According to all the steps of sample analysis, the blank experiment was repeated seven times, and the recovery rate and method detection limit were calculated. The results are shown in Table S5, none of the 25 target substances were detected in blank samples, LOQ was 6.9–21.6 ng/L, sample recovery was 60.2%–144.0%, and the detection curves showed good linear relationship (R² > 0.9997) at the 0.5–100 μg/L concentration of each antibiotic.

The ecological risk level was evaluated according to the RQ grading method proposed by other studies (Straub et al. 2019; He et al. 2020). When RQs < 0.1, there is a low ecological risk, and there is no need to take risk reduction measures at present; when 0.1 < RQs < 1.0, there is a medium ecological risk, indicating the need for further observation of the pollutant; when RQs > 1.0, there is a high ecological risk, indicating that the pollutant has ecological risk and corresponding measures should be taken to reduce the risk (Pan et al. 2020).

In the sewage collected from the different treatment units of STP 1 and STP 2, the occurrences and concentrations of 25 target antibiotics are shown in Table 1 and Figure 2. Fourteen and 16 antibiotics were detected in STP 1 and STP 2 sewage, respectively, and the detection rates were 56 and 64%. Among the antibiotics, TCs and QNs were the most frequently detected, followed by SNs, and the least frequently detected antibiotics were MLs. In addition, the concentrations of TCs and SNs in the raw water samples were much higher than those of QNs and MLs. This result may reflect the general use of TCs and SNs in the study area. This scenario is opposite that for the TCs and QNs frequently detected at the STP in Jinan (China) (Kun et al. 2021), but the situation is similar to that at the STPs in Guangdong (China) (Zhou et al. 2013). It may be that different regions have differing dependences on different classes of antibiotics (Phoon et al. 2020).

There were differences in the occurrences of antibiotics in the different treatment units. In the influent water, the concentrations of antibiotics detected in STP 1 were 13.7–4,265.2 ng/L, the concentrations of antibiotics detected in the MBBR tank were 15.0–2,773.6 ng/L, the concentrations of antibiotics detected in AWE were 12.0–693.3 ng/L, and the concentrations of antibiotics detected in the effluent were 8.2–269.2 ng/L. In each processing unit, TCs and QNs were ubiquitous. It can be clearly seen in Figure 2 that OTC, TCC and SXZ mainly existed in the influent of STP 1. They are TCs and SNs, respectively. QNs and MLs accounted for 2.77% and 2.58% of the total influent concentration, respectively. This phenomenon may be caused by the inhibitory effect of TCs and SNs on both Gram-positive and Gram-negative bacteria (Qian-Qian et al. 2015). These antibiotics are mainly used to treat swine asthma, pasteurellosis, brucellosis, anthrax, and acute respiratory infections (Jinmei et al. 2021). However, SMZ is only used as an animal promoter in China (García-Galán et al. 2012) and should not be detected at high concentrations. Thus, it is possible that the raw water under investigation was affected by the migration and diffusion of antibiotics from nearby farms.

In the study area, QNs are detected at a high frequency, but they have low detected concentrations. Compared with other antibacterial drugs, QNs are widely used in human medicine, aquaculture and veterinary medicine for the treatment of bacterial diseases due to their wide antibacterial spectrum, strong antibacterial activity, lack of cross resistance and minimal toxic effects and side effects (Backhaus et al. 2000). However, QNs have a long half-life in water and do not easily degrade, so their detection rate in aquatic environments is high.

MLs are a typical antibiotic used by humans and animals. Only ERC was detected in STP 1, and the detected concentration was high and higher than that detected in a sewage plant in Hangzhou (China) (Bao et al. 2021). Studies have shown that ERC is an ML with a high detection rate in aquatic environments in China, but the detection rate of clarithromycin is higher than that of ERC in other developed countries (Göbel et al. 2005). This scenario indicates that there are certain differences in the uses of these antibiotics in different regions.

Similar to those in STP 1, the main classes of antibiotics in STP 2 were TCs and SNs, but there were differences in the contents and compositions of the four classes of antibiotics. In contrast to STP 1, in STP 2, the total detected concentration of antibiotics was 25,265.0 ng/L, which was more than twice that in STP 1. Compared with STP 1, STP 2 had concentrations of 12 antibiotics, in addition to CPX, SMZ, and SAZ, that were generally the same or higher. In addition, the ERC (620.7 ng/L), CTC (1,349.0 ng/L), DCC (1,200.0 ng/L), OTC (10,761.7 ng/L), TCC (7,193.0 ng/L), and SXZ (2,169.3 ng/L) detected in STP 2 were significantly higher than the ERC (277.3 ng/L), CTC (739.5 ng/L), DCC (452.4 ng/L), OTC (4,265.2 ng/L), TCC (3,120.3 ng/L), and SXZ (1,026.9 ng/L) detected in STP 1, and this result may be related to their processing scales and surrounding environments. Studies have shown that when TCs come into contact with activated sludge, they are easily adsorbed by the sludge in a short time and accumulate in the sludge. Most SNs have weak adsorption potential but good water solubility, while QNs and MLs are not easily degraded (Kun et al. 2021). Therefore, in the effluent, the detected concentrations of TCs were lower than those of the other three classes of antibiotics.

The removal efficiency of each treatment unit in the two STPs was calculated, as shown in Table 2 and Figure 3.

As shown in Table 2, the antibiotic removal efficiency of the MBBR tank was −56.73% to 100%. In addition, the MBBR tank had a higher SN removal efficiency than the removal efficiencies of other three classes of antibiotics. Because TCs were the main antibiotics in STP 1 and the QN and ML removal efficiencies of this unit were not good, the total antibiotic removal efficiency was only 47.21% for the MBBR tank. It is worth noting that in the MBBR tank, negative removal of DNX, MBX, SFX, and ERC occurred, and their removal efficiencies were −13.54%, −9.49%, −8.71%, and −56.73%, respectively. Studies have shown that due to the low detection concentrations of QNs in winter, the QN removal efficiency fluctuates greatly, resulting in negative removal (Zhou et al. 2013). ERC, as an ML that is frequently detected in sewage, is frequently found to be negatively removed in different units throughout the STP process (Chang et al. 2010; Lin et al. 2018; Park et al. 2020). The phenomenon of negative removal may be caused by the following two reasons: first, the presence of other sewage water quality parameters, such as NH₄⁺ and toxic compounds, may affect the degradability of antibiotics (Han et al. 2018). Second, although biofilm treatment technology is considered suitable for the treatment of biodegradable micropollutants (Petrovic et al. 2009), the scale of sludge flocculation is small. When the mass load discharged from the bioreactor is higher than that flowing into the bioreactor, part of the degraded or decomposed antibiotics may be reduced to the parent compound (Polesel et al. 2016), which produces antagonism. It is worth noting that the sulfonamide antibiotic removal efficiency of the MBBR tank is high. In addition to the SXZ removal efficiency of 77.56%, the SMZ and SAZ removal efficiency of the MBBR tank reached 100%. The reason for this result may be that in the biofilm system, the biological process may also involve highly active microorganisms with aerobic and anaerobic particles (Sabri et al. 2020; Kun et al. 2021). Particles increase the adsorption of antibiotic compounds in the MBBR tank.

When the four classes of target antibiotics passed through the AWE, except for that of CTC, OTC and TCC, the removal efficiency of other antibiotics ranged from 0.41% to 62.93%, and the total removal efficiency was 62.80%. Some studies have shown that the occurrence of redox is the main factor affecting the removal of antibiotics in an AWE because in the cold winter, the plants existing in an AWE cause the anaerobic process of antibiotics through radial oxygen loss (Dan et al. 2020). Therefore, the overall antibiotic removal efficiency of the AWE was not high. After UV disinfection, the concentration of the antibiotics in STP 1 also decreased significantly, indicating that UV disinfection can effectively remove antibiotics in water, a result that is consistent with those of previous studies (Phoon et al. 2020). After the treatment, the concentrations of 16 antibiotics were in the range of n.d. ∼269.2 ng/L, and the concentrations of the target antibiotics all decreased.

When passing through the IBR tank, except for the SPZ and SAZ removal efficiencies, which were 100%, the removal efficiencies of the remaining 14 antibiotics were −8.97% to 60.13%, and the total antibiotic removal efficiency was 45.55%. In the IBR tank, CFX, NFX, and OFX had a negative removal trend, and the removal efficiency of other antibiotics was not high in this unit. Studies have shown that antibiotics can be adsorbed in a short period of time when they are in contact with activated sludge in IBR tanks (Kun et al. 2021). However, when antibiotics accumulate over a long-term cycle in this unit, high concentrations of antibiotics will inhibit the activity of microorganisms in activated sludge (Yang et al. 2020; Bao et al. 2021). This scenario leads to a low removal efficiency or even negative removal in the IBR tank (Hou et al. 2019).

The QN and ML concentrations in the sedimentation tank did not decrease significantly, which was consistent with the results of previous reports (Zhang et al. 2021) and may be related to the solid–liquid distribution coefficient (Tang et al. 2019). Chlorination is the last stage of STP 2. In this stage, the concentrations of all antibiotics decreased to different degrees compared with those in water, and the removal rate was 19.01%–99.84%. Although chlorinated disinfectants can react with dissolved organics such as antibiotics in water to produce a removal effect, they will produce a wide range of disinfection byproducts (Feng et al. 2019). Therefore, chlorination is not recommended for high-concentration antibiotic sewage.

Figure 3 shows that there were significant differences in the CFX and ERC removal efficiencies between the two STPs, and the removal efficiencies of the other QNs were 40.15%–70.67%. Studies have shown that the main QN removal mechanism is sludge adsorption. During the adsorption process, metal ions (such as Ca²⁺ and Mg²⁺) promote the combination of carbonyl and carboxyl groups in QNs; the chelates formed are stable, so they are not easy to remove (El-Kommos et al. 2003).

In STP 1 and STP 2, the ERC removal efficiencies were 2.92% and 77.01%, respectively, with a large gap between the two processes. Studies have shown that ERC has strong hydrophobicity, and it is more likely to be removed by sludge adsorption (Park et al. 2020). Therefore, the ERC removal efficiency of the MBBR process was lower than that of the IBR process. It is worth noting that the detected concentrations of ERC in the effluent of the two STPs were 269.2 ng/L and 142.7 ng/L, respectively, which were relatively high antibiotic concentrations and similar to the high ERC concentrations detected in the STPs in South China and the United States (Yang & Carlson 2004; Zhou et al. 2013). This result also reflects that the ERC removal effect cannot be effectively removed during various processes, which has a great influence.

Comparing the two STPs, STP 1 had better SN removal efficiency than STP 2, mainly because of the biodegradation in STP 1 and STP 2. The TC removal efficiency of s the final effluent was more than 99%, which indicates that the removal efficiency of TC antibiotic water treatment is very high; this result showed that all units had a good removal effect, possibly because TC antibiotics are greatly affected by pH (Jiao et al. 2008). Under different pH conditions, the molecular structures of TCs have different forms of polar functional groups. With increasing pH value, the degradation ability of TCs is enhanced (Dai et al. 2020).

As mentioned above, changes in antibiotics in various treatment processes may be correlated with various environmental parameters, so the potential relationship between antibiotic concentration and environmental parameters (relevant parameters are shown in Table S6) was analysed through multiple factors.

The pH, water temperature (WT), chemical oxygen demand (COD), and ammonia nitrogen (NH₃-N) of the sewage in each treatment unit were measured, and Canoco5 software was used to directly analyse the relationship between the target antibiotic and the physical and chemical parameters of the water. According to detrended correspondence analysis (DCA), the gradient length was 2.0, which was less than 3. Therefore, the redundant analysis (RDA) model was used for analysis (Wang et al. 2017; Zhao et al. 2020). The analysis results are shown in Figure 4.

Although COD and WT are environmental factors, there had strong correlations with antibiotics. COD and most antibiotics had trends in the same direction and thus were positively correlated. However, WT was negatively correlated with antibiotics because its trend was not in the same direction.

Table S7 shows that the p values of COD, WT, pH, and NH₃-N were 0.042, 0.06, 0.152, and 0.91, respectively. These results indicate that the antibiotics in the study area were significantly related to COD, and there was no correlation with WT, pH, or NH₃-N. The reason for this phenomenon may be that the main antibiotics in the study area are TCs, and the COD removal process in each treatment unit is similar to that of TCs. Studies have shown that all organic matter is composed of C and H, so most organic matter can be chemically oxidized or oxidized by microorganisms in wastewater. Organic matter needs to consume oxygen during the chemical oxidation process. Therefore, the more antibiotics there are in wastewater, the higher the oxygen consumption; the two are positively correlated (Fu et al. 2020). The influence of many environmental factors results in this being an extremely complex water environment and system. Therefore, the relationship between traditional pollutants and antibiotics needs further exploration.

The toxicity levels of antibiotics can be divided into acute toxicity and chronic toxicity (Marina et al. 2005). The former mainly occurs in green algae, bacteria and small daphnia, while chronic toxicity occurs mainly in large daphnia and fish, which is related to the trophic level at which antibiotics are located (Li et al. 2020). Under normal conditions, green algae, bacteria and small daphnia are at a low trophic level, while large daphnia and fish are at a high trophic level (Liu et al. 2015). Species at different trophic levels have different enrichment effects on pollutants, which makes the sensitivity of pollutants different, and this result is consistent with those of previous studies (Morosini et al. 2020). Discharging effluent containing antibiotics into the Kyushu River may adversely affect aquatic organisms in the aquatic environment and even human health (Wu et al. 2021). Therefore, an ecological risk assessment of the discharged effluent is of practical significance.

Current studies have shown that the toxic effects of TCCs on aquatic animals in aquatic environments rarely result in acutely toxic effects, but long-term exposure to antibiotics in the environment may be chronically toxic to nontarget organs (Dai et al. 2020). Antibiotics such as CFX, ERC, CTC, OTC, and TCC cause high ecological risks to some regions. For example, studies have found that in comparison to other antibiotics, OFX and CFX pose a higher risk to algae (Zhi et al. 2020). In ecosystems, different trophic levels have different sensitivities to target pollutants. For example, some scholars have studied the ecological risks to fish, invertebrates, and algae in rivers in the karst region of southwestern China (Huang et al. 2019), and the RQs of fish, invertebrates, and algae were 0–0.0009, 0–0.3031, and 0–14.6857, respectively, in this area. Therefore, different nutrient levels should be classified and discussed in an ecological risk assessment.

The RQs of antibiotics are shown in Table S8 and Figure 5. Except for the RQs of fish and invertebrates that show a medium risk in ERC, other target antibiotics pose a low risk. However, for the RQs of algae, CFX and ERC presented high risk levels, and OFX and TCC presented medium risk levels. Although individual concentrations of most target antibiotics were low, the RQs were in the range of 0–0.1 and showed a low risk. However, when multiple antibiotics coexist, the risk values of STP 1 and STP 2 to fish, invertebrates and algae change from low risk values to medium risk and high risk values. This result shows that, in comparison to the risk of other antibiotics, the risk of ERC in effluent in the study area is more prominent, and ERC has been included in the candidate list of drinking water pollutants by the United States Environmental Protection Agency in 2010 (Li & Zhang 2011); thus, attention should be given to the problem of ERC concentration residues. The results show that algae are the most sensitive species in the aquatic environment, followed by invertebrates, while fish sensitivity to antibiotics is generally not significant. Therefore, algae were the most sensitive aquatic organisms to the detected antibiotics, followed by invertebrates and fish, which was consistent with the results of previous reports (Suli et al. 2020).

Although TCs do not pose as high of a risk as other organic pollutants (such as polycyclic aromatic hydrocarbons), the residues of tetracycline antibiotics may promote the emergence of resistant bacteria (Agerso et al. 2006; Xiang et al. 2016); for example, soil bacteria with resistance to TCCs will develop, and bacterial activity will decrease after a period of time (Luo et al. 2010). In the past 70 years, soil antibiotic resistance genes have increased overall. It was also reported that 11 ARGs have been detected in the drinking water source of the Huangpu River (Jiang et al. 2013). Therefore, it is necessary to combine RQs with resistance genes.

This study investigated the occurrence and removal of 25 typical antibiotics at two STPs in Guangxi Province, China. The results showed that the detected concentrations of antibiotics were different in the different treatment units. Overall, in each treatment unit, in comparison to those of the other antibiotics, the detected concentrations of TCs accounted for the highest proportion, and it was the main antibiotic present. The SN removal efficiency of the MBBR process was better than that of the IBR process, but the TC and ML removal efficiencies of the IBR process were better than those of the MBBR process, which was based on the chemical structures, properties and environmental conditions of the sewage treatment plant. However, this result originates from a limited study site, so more investigations are needed to understand the fate of antibiotic occurrence and removal in STPs. In addition, the ecological risk assessment results indicated that ERC and SXZ may have toxic effects on aquatic organisms in this area. Risk mitigation measures therefore need to be developed. In addition to the management and elimination of antibiotics that occur in the environment, it is more important to encourage appropriate use of antibiotics and wastewater management to reduce the environmental risks caused by antibiotics from this source.

This work was supported by the Science and Technology Research and Development Program of Guangxi Zhuang Autonomous Region of China (Files no. AB21196039, no. AB21196039).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

All relevant data are included in the paper or its Supplementary Information.