Efficacy of anaerobic digestion wastewater treatment systems in reducing antimicrobial resistance bacteria in wastewater from swine farms in Thailand

Global public health is grappling with the issue of antimicrobial resistance (AMR). The AMR epidemic is a result of the improper and excessive use of antimicrobials in agriculture, animals, and humans. The rising global demand for animal products has increased antimicrobial use. Globally, antimicrobial usage in animal production cattle, sheep, chicken, and swine was estimated at 99,502 tonnes in 2020 and is projected, based on current trends, to increase by 8.0–11.5% to 104,079–107,472 tonnes by 2030 (Tiseo et al. 2020; Mulchandani et al. 2023). Recent studies estimate that swine farms use approximately 31,120 tonnes of antimicrobials annually, which accounts for about 40.9% of the total veterinary antimicrobial usage (Ardakani et al. 2024). The widespread use is primarily focused in Asia, which makes up 67% of total global usage (Mulchandani et al. 2023). In Thailand, 187,272 swine producers raised over 10 million swine in 2019. The use of antimicrobials on farms has been increasing rapidly, and it is projected that 843 tonnes of antimicrobials will be used by 2025 (Lekagul et al. 2020). The extensive use of these antimicrobial agents significantly contributes to the emergence of multidrug-resistant (MDR) bacteria, particularly in swine populations and their environments (Changkaew et al. 2015; Usui et al. 2016; Phongaran et al. 2019).

Antimicrobials are used in livestock to treat infections and prevent diseases. It is estimated that 30–90% of antimicrobials are distributed to various body tissues, with the rest excreted through urine and animal feces. Waste containing these anti-microbials is treated in wastewater treatment systems (WWTPs) before being released into the environment (Zhou et al. 2012). Managing wastewater in pig farms is of utmost importance due to the various sources on the premises. Urine, food scraps, swine feces, and other waste materials are the main sources of wastewater. According to Rizzo et al. (2013), as a result, WWTPs are a hotspot for microbes, drug residue, and animal agricultural waste, which encourages the development of AMR through selective pressure and gene transfer (Rizzo et al. 2013). Research and development for WWTPs in swine farms was previously mainly concentrated on improving the effectiveness of treating wastewater with high levels of organic pollution, with a particular focus on total Kjeldahl nitrogen (TKN) and biochemical oxygen demand (BOD), as well as minimizing offensive odors. Further research and development into developing WWTPs that can efficiently remove ARB and antimicrobials is becoming more and more necessary; this is a relatively new field of study.

Since 1995, animal farms in Thailand have implemented anaerobic digestion (AD) WWTPs, which have gained significant adoption. These systems are able to clean wastewater while also producing biogas (Aggarangsi & Teerasountornkul 2011). AD WWTPs are also effective in treating antimicrobial-resistant bacteria (ARB), resistance genes, and leftover antimicrobials, as demonstrated by recent studies. The integration of specific bacterial communities, operational conditions, and reactor types plays a crucial role in enhancing the removal efficiency (Wallace et al. 2018; Hosseini Taleghani et al. 2020; Zhu et al. 2020; Fan 2023). Swine farms in Thailand employ several types of AD WWTPs, such as covered lagoons, plug-flow anaerobic digesters (PFAD), and up-flow anaerobic sludge blankets (UASB), as well as combinations of PFAD and UASB. Recent research has shown that both PFAD and UASB systems can treat solid and liquid portions of swine wastewater and can help reduce the levels of pathogenic bacteria (Nuengjamnong & Rachdawong 2016). However, in Thailand, data on the efficiency of AD WWTPs in removing ARB is limited. In addition, while AD systems show promise in removing ARB, challenges remain, particularly regarding the variability in wastewater composition and the need for further optimization of operational parameters to enhance treatment efficacy.

Escherichia coli and Enterococcus spp. are used as indicators of fecal contamination in environmental investigations. These bacteria are normally found in the human intestines, but they have also been associated with diseases in both humans and animals (Gupta et al. 2001; Hammerum 2012). The AMR surveillance programs have also included E. coli and Enterococcus spp. as indicator species to assess the state of AMR in the clinical and environmental fields (Łuczkiewicz et al. 2010; Morris et al. 2023). Consequently, this study aimed to evaluate the AMR patterns among the included E. coli and Enterococcus spp. in the wastewater. Furthermore, characteristics of wastewater originating from swine farms were observed. This information could indicate the efficacy of AD WWTPs in removing ARB and reducing water pollution.

A cross-sectional study was conducted in three swine farms in Nakhon Pathom (Farm A) and Kanchanaburi (Farms B and C), Thailand, from October 2022 to January 2023. The research areas are in Regional Livestock 7, which includes 10 provinces in western and central Thailand. This region is recognized for its significant production and supply of swine. Furthermore, the farms included in the study were randomly selected from those utilizing AD WWTPs and voluntarily participated in the project. The study area is known for its high swine production and supply. Additionally, all farms utilize AD systems for wastewater treatment and voluntarily participate in a project. The study involved collecting wastewater samples before (influent) and after treatment (effluent) to analyze the AMR patterns of indicator bacteria and the characteristics of the wastewater. Research protocols were approved by the Biosafety Committee of Thammasat University, Thailand (no. 052/2564). Microbiological agents strictly followed the recommendations in the Biosafety Guidelines for Work Related to Modern Biotechnology or Genetic Engineering of Thailand's Technical Biosafety Committee.

To isolate E. coli, a 50-mm water sample underwent double filtration through Whatman No. 1 filter paper. The filtrate was applied to Chromocult Coliform Agar (Merk, Rahway, NJ) and then incubated at 37 °C for 24 h. Colonies displaying a blue-purple-red coloration, often accompanied by a discernible cloudy zone, were reserved for further validation through IMViC (Indole, Methyl Red, Voges-Proskauer, and Citrate) testing. Isolates exhibiting an IMViC pattern of positive–positive negative–negative were identified as E. coli. To isolate Enterococcus spp., a 0.1 mL sample was spread on Hicrome Enterococcus agar (Merk, Rahway, NJ) and then incubated at 37 °C for 24–48 h. Green colonies were suspected of Enterococcus faecium, while blue colonies were suspected to be another Enterococcus spp. Enterococcus spp. suspected colonies were streaked onto TSA slants and incubated at 37 °C for 18–24 h. Biochemical tests, including Catalase, L-arabinose fermentation, Yellow pigment formation, and PYRase tests, were conducted to confirm and classify the Enterococcus spp. (Murray 1990).

All Enterococcus spp. isolates were extracted DNA by the boiling method. Subsequently, DNA quantity and quality will be determined using a spectrophotometer (Microdigital, Gyeonggi-do, Korea). The components for performing PCR include 30 cycles with the following reagents: 0.25 μM primers, 1X GoTaq buffer, 1.5 mM MgCl₂, 0.25 mM dNTPs, 0.5 U Taq Polymerase, and 2 μL (1 ng/mL) DNA template, in a total volume of 20 μL. The primers used in this study are E.Faecium-F (5′-TTGAGGCAGACCAGATTGACG-3′) and E.Faecium-R (5′-TATGACAGCGACTCCGATTCC-3′) (Cheng et al. 1997). Amplification was done with an initial denaturation step of 95 °C for 5 min, followed by 30 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s and elongation at 72 °C for 1 min, with a final elongation of 10 min. The PCR product was analyzed using electrophoresis with a 2% agarose gel stained with GelGreen for visualization (Biotium, Fremont, CA).

In assessing antimicrobial susceptibility, three isolates of E. coli, Enterococcus spp., and E. faecium from each water sample were randomly selected and subjected to antimicrobial susceptibility testing. All isolates were included in samples with fewer than three identified isolates. The agar disc diffusion method should strictly adhere to the Clinical Laboratory Standards Institute Guidelines (CLSI) (Clinical & Laboratory Standards Institute 2023). For testing E. coli, a panel of antimicrobials was tested: gentamicin (GM), cefoxitin (FOX), cefotaxime (CTX), ceftazidime (CAZ), ampicillin (AM), ciprofloxacin (CIP), norfloxacin (NX) tetracycline (TE), doxycycline (DO), and sulfamethoxazole (SXT). For testing Enterococcus spp., the following antimicrobials were utilized: AM, CIP, NX, TE, vancomycin (VA), and erythromycin (E). In addition, minimum inhibitory concentrations (MICs) of colistin (CL) against E. coli were determined using the microbroth dilution assay according to the CLSI guidelines (Clinical & Laboratory Standards Institute 2023). Briefly, CL sulfate (Sigma-Aldrich, Burlington, MA, USA) was diluted in sterile 96-well plates using cation-adjusted Mueller–Hinton broth (Difco, BBL^TM, Franklin Lakes, NJ, USA) to create concentrations ranging from 0.5 to 128 μg/mL in sterile 96-well plates. A bacterial suspension at 10⁶ CFU/mL was inoculated into the broth with varying CL concentrations and incubated at 37 °C for 18–24 h. The MICs were defined as the lowest concentration exhibiting no visible bacterial growth upon evaluation by visual inspection. The susceptibility test results were interpreted according to susceptibility and resistance clinical breakpoints suggested by the EUCAST (European Committee for Antimicrobial Susceptibility Testing 2023) for Enterobacteriaceae (susceptible, MIC ≤ 2 mg/L; resistant, MIC > 2 mg/L). E. coli ATCC25922 and Staphylococcus aureus ATCC25923 were standard bacteria for quality control during E. coli and Enterococci spp. tests, respectively.

Water quality analysis for physical and chemical parameters was conducted using the Standard Methods for the Examination of Water and Wastewater (Baird et al. 2017). The parameters measured encompassed pH, BOD, total suspended solids (TSS), chemical oxygen demand (COD), TKN, and total phosphorus (TP), as detailed in Table 1.

Statistical analyses were performed using the SPSS program version 18 for Windows (PASW serial no. 5082357) (SPSS Inc, Chicago, IL). Data were analyzed using descriptive statistics, e.g. frequency distribution, minimum–maximum values, and percentage. Pearson's chi-square and Fisher's extract test were employed to analyze the statistical differences in the resistant rate of indicator bacteria isolated from influent and effluent.

The farms under study have the following characteristics: Farm A is a large farm holding more than 600 livestock units that utilizes an anaerobic WWTP, specifically the CMU-CD (Chiang Mai University-Chanel Digester) systems, which comprise a sand-trapped tank, chanel digester; CD, and UASB blanket; UASB. Farm B and Farm C are medium-sized pig farms holding 60–600 livestock units that employ the cover lagoon system for wastewater treatment. Table 2 displays the influent and effluent quality analyses for all three farms. The results indicate high levels of pollution, including BOD, TSS, and COD, which surpassed the wastewater quality recommended by the Pollution Control Department, Thailand (Pollution Control Department, n.d.). When comparing the effluent to the effluent discharge standards (Ministry of Natural Resources & Environment 2020), it was found that Farm A and Farm B showed a BOD level in the effluent that exceeded the standard for large-scale and medium-scale swine farms. In contrast, Farm C demonstrated a BOD value that met the standard for medium-scale swine farms. Only Farm B had effluent that exceeded the suspended solids (SS) standard. Additionally, the effluent from all three farms had COD, TKN, and TP levels that exceeded the standard.

The study found susceptibility results for all tested E. coli (95 isolates) and enterococci (76 isolates). The enterococci included E. faecium (six isolates) and other Enterococcus spp. that could not be species characterized (70 isolates). In the influent and effluent, E. coli showed the highest resistance to AM (influent 100%; effluent 70.7%), followed by TE (influent 83.3%; effluent 70.7%), SXT (influent 74.1%; effluent 43.9%), and DO (influent 63.0%; effluent 34.1%). Enterococcus spp. isolated from influent showed the highest resistance to TE (80%), followed by CIP (76.7%), NX (76.7%), and E (66.7%). The resistance rates of tested antimicrobials in the effluent were quite similar (7.5–17.5%). E. faecium isolates from influent showed resistance to CIP, NX, and TE at the same rate (100%). However, only TE-resistant E. faecium (33%) was found in the effluent. Note that resistance to CL and VA, the last-resort drugs, was found in this study. CL-resistant E. coli was found in both influent and effluent (3.7 and 7.3%), while VA-resistant Enterococcus spp. (7.5%) was in the effluent (Table 3).

For resistant enterococci, E, AM, CIP, NX, and TE resistance was distributed in the influent, and the resistant rate of effluent in all farms decreased, unlike the VA resistance that occurred in the effluent but was not found in the influent of Farm A (Figure 2). At Farm A, the prevalence of CIP, NX, and TE-resistant Enterococcus spp. isolated from the influent significantly differed from those collected from the effluent (p-value < 0.05). In Farm B, the E and CIP-resistant Enterococcus spp. isolated from the influent showed significant differences from the effluent samples (p-value < 0.05). Similarly, the TE-resistant Enterococcus spp. isolated from the influent at Farm C also had significant differences from the effluent samples (p-value < 0.05).

Table 4 shows the patterns of AMR in MDR E. coli. In Farm A, MDR E. coli resisted up to 10 antimicrobials, namely GM-CTX-CAZ-AM-CIP-NX-TE-DO-SXT-CL. This resistance was found in two isolates in the influent. Additionally, MDR E. coli showed resistance to nine types, including GM-CTX-CAZ-AM-CIP-NX-TE-DO-SXT, in two isolates (11.1%) in the influent. An analysis of the effluent samples revealed that MDR E. coli exhibited only one resistance pattern (GM-AM-SXT). In Farm B, MDR E. coli exhibited resistance to up to five antimicrobials, including GM-AM-TE-DO-SXT, found in both influent and effluent at 11.1% each, and GM-CTX-AM-TE-SXT found in influent at 11.1%. The resistance patterns of MDR E. coli in Farm B were less diverse than in Farm A. Similar to Farm B, our findings in Farm C also showed less diverse resistance patterns of MDR E. coli compared to Farm A. MDR E. coli in Farm C exhibited resistance to up to six antimicrobials, namely GM-CTX-AM-TE-DO-SXT and AM-CIP-NX-TE-DO-CL, found in one isolate each in the effluent.

The AMR patterns of enterococci are shown in Table 5. When comparing the three farms, it is observed that MDR enterococci are most prevalent in Farm A, which has the highest number of enterococci resistant to six antimicrobials, namely VA-E-AM-CIP-NX-TE, found in the effluent at 5.6%. Following that, enterococci resistant to five antimicrobials, including E-AM-CIP-NX-TE, were found in the influent at 38.5%. In Farm B, enterococci resistant to up to four antimicrobials, including E-CIP-NX-TE, were found in the influent at 10%, and E-AM-CIP-NX in the effluent at 4.5%. In Farm C, enterococci resistant to up to four antimicrobials, namely E-CIP-NX-TE, were found in the influent at 35.7%, and no MDR enterococci were found in the effluent.

The excessive use of antimicrobials in animal husbandry has led to the emergence of ARB in animals and the surrounding environment, including wastewater treatment plants, where the antimicrobial residue, ARB, and the genetic element controlling AMR are gathered. In order to assess the effect of wastewater treatment plants on AMR, this study evaluated the effects of AD WWTPs, the common wastewater treatment plants operating in Thai swine farms, on AMR patterns among indicator bacteria in swine farm wastewater. In addition, the characteristics of swine farm wastewater and effluent quality were monitored.

The analysis of influent quality revealed that all three farms had higher levels of COD, TKN, and TP compared to the recommended standards for pig farms set by the Pollution Control Department of Thailand. The BOD values were consistent with other research, while the COD and TP values were somewhat higher. Additionally, the amount of TKN exceeded the findings of Pongthornpruek (2017). The results of this study on influent quality underscore the effective waste management practices, particularly in swine manure management. Swine wastewater combines manure and water used to clean the swine sheds. It contains high concentrations of nitrogen, phosphorus, and organic matter, expressed as COD and BOD (López-Sánchez et al. 2022). However, previous studies reviewed that the proportions of pollutants in livestock wastewater depend on the animal species and age, the feed composition, housing methods, and environmental factors (Lv et al. 2018; Nagarajan et al. 2019).

It should be noted that the high discharge of swine feces into wastewater treatment plants could result in elevated levels of antimicrobial residue and ARB in the systems. The study in Belgium demonstrates that swine raw manure often contains various antimicrobial residues. The most frequently detected antimicrobial residues are doxycycline, sulfadiazine, and lincomycin. The study also reported the co-occurrence of AMR Salmonella and E. coli (Rasschaert et al. 2020). Furthermore, high levels of COD and TKN reported in our study indicated a high concentration of organic pollutants and nitrogen that could promote bacterial growth, including resistance strains. The study of bench-scale biological reactors showed that wastewater quality parameters, such as COD, ammonia nitrogen, and turbidity, positively correlated with the reductions of ARB and ARGs. Therefore, improving the quality of wastewater is likely to contribute to a reduction in ARB and ARGs (Yuan et al. 2016).

This study used E. coli and Enterococcus spp., the fecal indicator bacteria, to monitor AMR in AD WWTPs. E. coli represented AMR against Gram-negative bacteria and broad-spectrum antimicrobials, while Enterococcus spp. represented AMR against Gram-positive bacteria and broad-spectrum antimicrobials (Łuczkiewicz et al. 2010). This study found that Farm A has the highest AMR distribution compared to the other two farms. E. coli resistant to ampicillin, tetracycline, doxycycline, sulfamethoxazole, and gentamicin was commonly found in the influent of all farms. Enterococcus spp. isolated from influent showed resistance to tetracycline, ciprofloxacin, and norfloxacin. This resistance reflects the various antimicrobial uses on farms, especially in Farm A. Informal interviews revealed that all studied farms administered antimicrobials to swine for bacterial infections such as respiratory tract infections, gastrointestinal tract infections, and arthritis. The antimicrobial agents used on the studied farms are ceftiofur, amoxicillin, fluoroquinolones, enrofloxacin, tiamulin, and neomycin. These antimicrobials were commonly used in Thai swine farms due to their effectiveness, safety, and legal use in treating animal diseases (Supawadee & Issarapong 2021).

Besides the commonly used AMR patterns in animal husbandry, this study also found resistance to antimicrobial agents that are considered to be of concern both in Thailand and globally, including CL and VA. These are last-resort drugs for treating infections caused by MDR bacteria in humans (Mühlberg et al. 2020; Mondal et al. 2024). CL is classified as a hazardous drug that is not allowed to be mixed into animal feed or drinking water for disease prevention. This study found E. coli resistant to CL in the influent of Farm A, with multidrug-resistance patterns, GM-CTX-CAZ-AM-CIP-NX-TE-DO-SXT-CL. E. coli resistant to CL was also found in the effluent of Farm C, with resistance patterns including AM-TE-CL, AM-TE-DO-CL, and AM-CIP-NX-TE-DO-CL. Additionally, VA-resistant Enterococcus spp. were found in the wastewater of Farm A, with MDR patterns, VA-E-AM-CIP-NX-TE. As reported in this study, MDR indicator bacteria resistant to the last resort are significant for public health and environmental safety. These bacterial strains, including their genetic elements, can spread into other environments through discharges and runoff (Takawira & Mbanga 2023). Moreover, the evolution of these pathogens can enhance their virulence and resistance profiles, complicating treatment options (Zhao et al. 2024).

The observed resistance patterns might indicate the use of colistin and VA on the farms, though informal inquiries revealed that neither CL nor VA was used on the farms studied. It is suggested that the emergence of CL and VA resistance may be due to co-selection from the use of other drugs on the farms, as the patterns observed are MDR. Part of the resistance mechanisms for CL and VA includes acquired AMR genes, such as the mcr-1 gene for CL and the vanA gene for VA, which may be transferred along with other AMR genes via mobile genetic elements, such as plasmids, transposons, and integrons. Therefore, resistance to CL and VA can be found even without their use on farms (Mühlberg et al. 2020; Li et al. 2022; Mondal et al. 2024). A study treating municipal wastewater found that cefotaxime-resistant E. coli exhibited significant resistance to CL (76.5%). Multiple beta-lactamase genes suggested the potential co-selection of CL resistance in wastewater treatment (Adegoke et al. 2020). Co-selection of mcr-1 with beta-lactam resistance genes (bla_CTX-M-55, bla_CTX-M-14) of E. coli isolated from boiler farms was demonstrated through conjugation experiments, indicating that mcr-1 can be transferred alongside other AMR genes on IncI2 and IncHI2 plasmids (Cao et al. 2020). Based on our interview, beta-lactam antimicrobials are commonly used on the studied farms. These antimicrobials may contribute to co-selection. To test our hypothesis, we need to investigate the ARGs, the residues of beta-lactamase, and other frequently used tetracyclines and quinolones in the effluent.

This study investigated the decreasing trend in AMR among E. coli and enterococci in effluent. The decline suggests that AD treatment systems might be effective in removing AMR. Our finding is evidence to support previous ones that emphasize the efficacy of anaerobic WWTPs in addressing AMR (Wallace et al. 2018; Hosseini Taleghani et al. 2020; Zhu et al. 2020; Fan 2023). Although AD treatment systems studied seem to affect ARB removal, the AMR proportion of GM (Farms B and C) and CTX, DO, and CL-resistant E. coli (Farm C) in effluent is increasing. The increasing AMR was evident in sewage and sludge from municipal wastewater treatment plants using biological treatment systems, where MDR enterococci were found (Da Costa et al. 2006). It is hypothesized that sub-inhibitory concentrations of antimicrobials and disinfectants, often detected in sewage, might promote the growth of ARB or gene transfer between bacteria in the sewage treatment tanks or the biofilm lining of the pipelines (Aiello & Larson 2003; Lindberg et al. 2004). The sub-inhibitory concentrations of antimicrobials that persist in the wastewater treatment pond can influence changes in bacterial characteristics favoring AMR, such as biofilm formation, SOS system response, and disruption of primary metabolism (Berglund 2015). Regarding these hypotheses, the concentration of antimicrobial residue and ARGs in the influent of the tested farms needs to be investigated.

The current study focuses on the effectiveness of AD WWTPs commonly utilized in Thai swine farms, Chiang Mai University-Chanel Digester (CMU-CD), and covered lagoons, highlighting the variations in their ability to reduce ARB. CMU-CD systems combine the fundamentals of a rate UASB digester and the construction simplicity of a covered lagoon with the benefit of relatively warm and humid ambient weather (Aggarangsi & Teerasountornkul 2011). Covered lagoons are the most direct AD technology, where feedstocks are stored in an underground lagoon covered with a gas-tight flexible cover (Uddin & Wright 2023). AD WWTPs can remove ARB through various mechanisms and reactor configurations. Integrating specific bacterial communities, operational conditions, and reactor types plays a crucial role in enhancing the removal efficiency of ARB (Wallace et al. 2018; Hosseini Taleghani et al. 2020; Zhu et al. 2020; Fan 2023). However, none of the AD WWTPs can eliminate ARB. To improve the efficacy of the systems in reducing ARB, additional disinfection processes in treated wastewater before discharge are essential (Yuan et al. 2016). This suggestion conforms with the findings of the current study. All studied farms informed that disinfection processes were not operated before the discharge of the influent, which might affect the occurrence of ARB in the influent.

AMR is a major public health problem. This study demonstrated AMR fecal indicator bacteria (E. coli and Enterococcus spp.) distribution in wastewater from studied swine farms, suggesting widespread antimicrobial use on farms. CL-resistant E. coli and VA-resistant Enterococcus spp. were reported. The resistance to the last-resort antimicrobials may be due to the co-selection of other antimicrobials used on farms. Therefore, it is important to consider the potential for transmission and risk assessment of these MDR bacteria to humans, other animals, and the environment. Because AD WWTPs can treat wastewater and simultaneously produce biogas, they are frequently used in Thai livestock operations, notably swine farms. This study found that the declining AMR in E. coli and Enterococcus spp. in effluent suggested that AD might effectively remove AMR. On the contrary, high levels of BOD, TSS, and COD found in swine wastewater indicate significant levels of pollution. Swine farms should concentrate on waste management initiatives, particularly collecting pig feces, to ensure that the AD functions effectively for AMR and pollutant removal. Maintaining the treatment system on a regular basis is also essential. In addition, the operation of AD combined with another treatment method, such as disinfection processes, could enhance AD efficacy in ARB reduction. The data from three swine farms were analyzed in this study. To validate that the results are more typical of swine farms and appropriate treatment methods, future research should include more farm samples. In addition, to ascertain the efficacy of AD for the removal of ARGs and antimicrobial residues from the wastewater, more information on antimicrobial residues in the WWTP is required.

This work was supported by the Thailand Science Research and Innovation Fundamental Fund for the fiscal year 2023.

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

The authors declare there is no conflict.