Antimicrobial resistance (AMR) is one of the leading causes of morbidity and mortality worldwide. It has a huge negative impact on economic sustainability (FAO, UNEP, WHO, and WOAH, 2020; Walsh et al., 2022). The large worldwide use of antimicrobials for animals (73% of all administered antibiotics) is correlated with increased AMR spread in humans and animals (Boeckel et al., 2017; Ardakani et al., 2023). On one hand, this could be due to the fact that certain antibiotic classes are used in animals and humans [Rahman and Hollis, 2023; European Food Safety Authority (EFSA) and European Centre for Disease Prevention and Control (ECDC), 2024]. On the other hand, it could be because resistant pathogens and/or genes can be transmitted to humans through food chains or direct contact with livestock (Bennani et al., 2020). This alarming trend underscores the need for effective resistance control and prevention strategies to protect human and animal health and sustain the economy.

In order to effectively combat resistance spread, a comprehensive understanding of the emergence and transmission of resistance is crucial (Berendonk et al., 2015). This can be possible through phenotypic and genotypic surveillance and monitoring of resistance patterns in bacterial populations over time (Graham et al., 2019). Both are very useful for understanding the resistance mechanisms and are important for developing strategies against the spread of antibiotic-resistant bacteria and protecting public health (Diallo et al., 2020).

Escherichia coli (E. coli) is commonly used as an indicator organism in many antimicrobial resistance surveillance programs (Anjum et al., 2021; Safety, European Food, and European Centre for Disease Prevention and Authority Control, 2021). Because E. coli is a potential reservoir and transmitter of many different AMR genes. As a commensal bacterium in the intestines of humans and animals, it also facilitates the acquisition and accumulation of resistance genes (Pissetti et al., 2021). Similarly, E. coli is ubiquitous and is found in various environments including the gastrointestinal tract, water, soil, and food (Jang et al., 2017). These properties make it a valuable sentinel organism for monitoring AMR trends and transmission in various environments. Hence, by tracking the presence and patterns of AMR in E. coli, surveillance programs can provide valuable insight into the overall prevalence and spread of AMR, thereby supporting the development of combating strategies to this global health threat (Suwono et al., 2021).

In pig production, antibiotics are primarily used to treat and combat diseases (Cromwell, 2002; Hallenberg et al., 2020). They are particularly often applied in the early weaning period, as the piglets are very vulnerable for infections (Dewulf et al., 2022). Weaning, which is conventionally done at 3 to 4 weeks of piglet age, is an extremely stressful time marked by dietary and environmental changes, often resulting in decreased feed intake, inadequate weight gain, and potential problems such as diarrhea, morbidity, and mortality. During this period, piglets are very susceptible to infections due to their underdeveloped immune system, compounded by the rapid decline in maternal passive immunity and the early onset of active immunity (Campbell et al., 2013).

The status of AMR resistance varies across pig farms depending on antibiotic usage, production type, hygiene practices, and biosecurity measures (Österberg et al., 2016; Mencía-Ares et al., 2021). There are studies indicating that farmers use large amounts of antibiotics to combat diseases associated with poor biosecurity and unhygienic conditions in farms (Raasch et al., 2018; Albernaz-Gonçalves et al., 2021). Incorrect and excessive use of antibiotics contributes to the widespread emergence and spread of AMR in livestock farms (Aarestrup et al., 2008; Holman and Chénier, 2015; EMA and EFSA, 2017). Although, AMR occurrence can be decreased by reducing antimicrobial use (Friedman and Whitney, 2008; European Centre for Disease Prevention and Control (ECDC) et al., 2017). The spread of AMR can also be counteracted through strict biosecurity and hygiene management (Laanen et al., 2013; Dohmen et al., 2017; Raasch et al., 2018).

In our preliminary work (Behrens et al., 2023), a dramatic increase in the prevalence of resistant bacteria (from <1% of resistant E. coli to almost 100%) in fattening piglets was observed within the first 4 weeks after weaning at farm arrival. This was not solely due to selection pressure from the use of antibiotics alone, but also possibly due to the colonization of antimicrobial-resistant microbes from the piglet environment in the barn (Yun et al., 2021; Saladrigas-García et al., 2022; Smith et al., 2023). In addition to the use of antimicrobial medications and management, factors such as barn hygiene can influence the emergence and persistence of drug-resistant intestinal bacteria in pigs (Dewulf et al., 2007). Some reviews also speculate about a connection between biocide and antibiotic resistance in bacteria, particularly through the mechanism of numerous efflux systems that cause co-(cross-)resistance to a number of structurally unrelated antimicrobials such as antibiotics and biocides (Ortega et al., 2013; Jean-Yves, 2018). Similarly, Puangseree et al. (2021) recommend routine monitoring of biocide tolerance because it plays a role in the emergence and spread of AMR when used frequently, acting as a non-antibiotic selection pressure. By implementing effective hygiene measures such as thorough cleaning, disinfection, and biosecurity measures, it is possible to curb the development and reduce the spread of AMR (Yun et al., 2021). In this study, we improved hygiene in the piglet barn through increased hygiene management techniques and compared the impact on the spread of AMR with routine practices. We analyzed biocide resistance of barn AMR E. coli to the farm regular disinfectant that had been used in the study farm for many years.

In this study, we investigated the consequences of hygiene improvement on the spread of AMR E. coli in piglets and their environment. Poor hygiene and biosecurity in intensive livestock farming play a central role in the spread of diseases, including antimicrobial resistance, by serving as a route of transmission for pathogens (Dewulf et al., 2007; Postma et al., 2015). By improving hygiene management in the piglet house, we were unable to find a significant differences to the conventional practices in terms of reducing the presence of antibiotic-resistant bacteria.

In our mitigation trials, a high prevalence of ampicillin-resistant E. coli was observed in feces and fly samples from both groups (experimental and control group), while resistances were moderate in feed and relatively low in dust (Table 1). In fact, this is not surprising because enteric pathogens are primarily excreted in feces and spread via the fecal-oral route (Gerba, 2009). Flies also play an important role in the transmission of resistant pathogens through circulating pathogens from animals and/or their excretions to other individuals, feed, equipment and environment in the house, as they roam around indiscriminately and also spread outside of barns (Meerburg et al., 2007; Usui et al., 2015; Behrens et al., 2023).

Ampicillin and sulfonamide resistance were prevalent in nearly all types of samples, while very low cefotaxime (only in feces) and no ciprofloxacin resistance were detected. Ampicillin-resistant E. coli were present in all 24 (100%) fecal samples with a proportion between 5% and 100% to the amount of non-resistant E. coli (Table 1). This suggests a widespread ampicillin resistance, might be related to previous colonization (before arrival) or possibly due to the use of the drug (Dupamox; amoxicillin derivatives) against rare bacterial infections in individual piglets on the studied farm, particularly in experimental groups (Supplementary Table S1). Antibiotic treatment of individual diseased pigs resulted in increased E. coli resistance that can last up to some weeks (Sali et al., 2021; Tams et al., 2023). On the other hand, ampicillin is also an old antibiotic widely used in humans and animals and therefore has been heavily exposed to Enterobacteriaceae for many years, resulting in a high basic rate of resistance and the according resistance genes, particularly bla_TEM-1 (Tran-Dien et al., 2018). The bla_TEM-1 gene confers resistance to old beta—lactam antibiotics such as penicillin derivatives, and early cephalosporin (Salverda et al., 2010). Although no single treatment with sulfonamide drug administered to piglets in our experimental unit, sulfonamide-resistant E. coli were frequently detected in their fecal samples. This indicates that direct exposure to the drug is not the only factor for resistance and is more likely a consequence of sulfonamides were widely used in agriculture and livestock production over many decades (Lesch, 2007). Likewise, a previous longitudinal study indicates that the amount of antibiotics from the penicillin group (amoxicillin and ampicillin) used in weaned piglets in Germany is approximately more than 75% compared to the other antibiotics, highest treatment frequency (Schaekel et al., 2017) and also the proportion of resistance to this group of antibiotics was also the highest (Van Rennings et al., 2015; Mesa-Varona et al., 2021).

The prevalence of cephalosporin resistance in the current study was quite low (Figure 2), only 5 of 12 (42%) fecal samples from the experimental group had cefotaxime-resistant E. coli with a proportion less than 0.5%. Similarly, a very low frequency of bla_CTX-M-1 (resistance gene for extended-spectrum cephalosporins) in fecal samples was also detected by qPCR in both groups. A study by Cavaco et al. (2008) showed that the harboring of bla_CTX-M-1 resistant coliform bacteria and the use of cephalosporin is the main cause of the spread of indigenous bla_CTX-M-1 producing E. coli strains and the possible emergence of strains producing bla_CTX-M-1 genes acquired through horizontal transfer. Therefore, the few detections of cefotaxime-resistance in the experimental groups in cultivation and qPCR could be related to exposure to bla_CTX-M-1-producing bacteria before arrival or treatment with cephalosporins.

Although there was no significant difference in ampicillin resistance between improved and conventional hygiene management, some factors may have overshadowed the effect of hygiene improvement on the occurrence of AMR. As shown in Figure 1B, the proportion of ampicillin resistance observed across weeks and groups varied greatly. One such factor could be the pre-arrival colonization of piglets with ampicillin-resistant bacteria, which were detected in fecal samples on the first day of arrival (12–24 h; Supplementary Figure S2). A high proportion of ampicillin-resistant E. coli was observed on the arrival day, particularly in the experimental groups. In addition we observed that AMR levels can vary strongly between individuals in the same herd and of the same origin (data not shown). AMR pathogens from infected piglets can easily contaminate feed and water sources in the flat deck. The contamination of feed and water can lead to the ingestion of AMR bacteria by other piglets, thus promoting the spread of resistance easy and fast (Hutschemaekers et al., 1976). If AMR bacteria were initially detected in piglets this might have impacted hygiene management measures.

Quantitative PCR results showed that the total bacterial load in pooled fecal samples was significantly lower in the experimental group than in the control group on the first day of arrival (Figure 3). This indicates that soaking with detergents during cleaning significantly reduced the overall bacterial load in the experimental groups than just soaking with water alone in the control groups. Similarly, a former study found a significantly greater reduction in total bacteria in swab samples collected in pig houses that were first soaked with detergent compared to houses cleaned without detergent (Hancox et al., 2013). Another reason for the difference in bacterial load between the experimental and control groups can be the properties of the disinfectants. FL-des Allround (disinfectant for experimental groups) forms a foamy consistency when applied and retains the active ingredients exposed to the microbes for a longer time. On the contrary, Sorgene does not form foam and therefore drains faster and the active ingredients exposed to the microbes have comparatively less time to act. Foamy disinfectants offer a longer exposure time compared to their liquid counterparts (Cadnum et al., 2020). The time the disinfectant is exposed to the microbes is very important to the effectiveness of disinfection (Wales et al., 2021).

Selected ARGs were detected and quantified by qPCR in all fecal samples of both groups. There was no significant difference for ARGs between the samples from the experimental group and the control group, except for bla_TEM-1 and tet(A) (p < 0.04 for both) in the fourth week (Figure 3). In Addition, a significant difference in the relative abundance of bla_TEM-1 between the experimental group and the control group was detected (p < 0.0086; Figure 4). One possible explanation for this is that 1 week after arrival ampicillin treatment (Supplementary Table S1) given to individual piglets experimental groups 1 and 2 led to resistance selection. In the study by Zeineldin et al. (2019), authors suggested that early interventions with penicillin’s in piglets could promote resistance selection in herds. The other reason could be related to the previous colonization of the piglets with the resistant bacteria before arrival, as roughly indicated in Supplementary Figure S2.

Trends in resistance genome tracking during experiments indicate that the emergence and spread of AMR over time is not related to persistent AMR from the previous piglet batch (Figure 5). This can be determined from the ARGs in swab samples collected before the arrival of the piglets and the ARGs in fecal samples on arrival date and in the first week after arrival. The resistance gene prevalence in swab samples were very low as expected compared to those in feces. The resistance patterns in swab and fecal samples were inconsistent. We conclude that the emergence and spread of antimicrobial resistance in farms is not due to poor cleaning and disinfection. A difference between the two groups was found in the proportion of the two beta-lactam resistance genes bla_CTX-M-1 and bla_TEM-1. This could be related to earlier colonization before arrival, and the use of antibiotics in piglets (in this study Duphamox; an amoxicillin antibiotic, was used in experimental group one) to treat bacteria results in large variations in the overall resistance frequency. Burow et al. (2019) indicated that antibiotic treatment and the spread of antibiotic resistance in the production chain have a major impact on the prevalence of antibiotic resistance in pig farms. It confirmed the difference in beta-lactam resistance of E. coli between treated and untreated pigs. The likelihood that piglets will carry E. coli resistant to ampicillin is quite high if their mothers developed resistance to the same antibiotics. Many previous studies have shown a possible link between antibiotic resistance in sows and their offspring, which could have potential effects on individual animals (Callens et al., 2014; de Greeff et al., 2020; Pholwat et al., 2020). Microorganisms originating from the maternal and surrounding environment may significantly contribute to the microbial succession observed in newborn piglets following their birth (Chen et al., 2018).

Similar to the previous argument, the distribution patterns of sulphonamide and tetracycline resistance genes could indicate that they are not group-specific, due to hygiene management or the use of both antibiotics (as shown in Figure 3; Supplementary Table S1). This shows that antibiotic exposure cannot be the only important factor in the appearance of AMR genes to these antibiotics (Holman and Chénier, 2015).

According to our hypothesis and some speculations (Ortega et al., 2013; Jean-Yves, 2018; Puangseree et al., 2021), biocide tolerance could be related to long-term use of disinfectants on farms and in some occasions linked to antibiotic resistance. However, the results of the disinfectant tolerance test showed that the E. coli strains isolated from both the control groups (disinfected by Sorgene) and the experimental groups (disinfected by FL-des Allround) were phenotypically susceptible to both disinfectants. Nevertheless, bacteria are able to develop resistance to disinfectants; especially when they are in the state of a spatially organized biofilm (Sanchez-Vizuete et al., 2015; Todorić et al., 2023) or are present after flushing. There was a significant difference in the minimum inhibitory concentration as a percentage of the recommended dose between the two disinfectants. To inhibit E. coli, a higher concentration of FL-des Allround was required than of Sorgene (Figure 6). In addition, Sorgene had a better inhibitory effect against E. coli that had been isolated on cefotaxime-suppelmented agar, compared to other E. coli, but FL-des Allround has a similar inhibitory effect on all E. coli isolates, irrespective of isolation. Furthermore, the bactericidal activity of Sorgene against E. coli isolated from the farm is quite high and no phenotypic cross-resistance with E. coli was observed. Similarly, Wieland et al. (2017) reported that residual concentrations of disinfectants were not able to select ESBL-/AmpC producing E. coli. This also agrees with Maertens et al. (2020) suggestions that repeated use of disinfectants in animal housing does not lead to antibiotic resistance or reduce susceptibility to disinfectants. Although some previous findings (Ortega et al., 2013; Schwaiger et al., 2014; Jean-Yves, 2018) indicate opportunities for bacteria to resist disinfectants, no co-(cross-)resistance between biocide and antibiotic was observed in our current study.

Barn cleaning and disinfection is of utmost importance to control the spread of antimicrobial resistance from the previous animal batch to the next one in livestock production (Martelli et al., 2017). However, the epidemiology of AMR is influenced in complex ways by a combination of factors that include antimicrobial drug use, biosecurity level, the emergence of cross-resistance, and many non-antimicrobial risk factors (Dewulf et al., 2007). Therefore, reducing the use of antibiotics and strict biosecurity measures after careful initial cleaning and disinfection have a major long-term impact on reducing antimicrobial resistance in pig farms [European Centre for Disease Prevention and Control (ECDC) et al., 2017; Mencía-Ares et al., 2021].

Compared to our previous work (Behrens et al., 2023) on the same farm, no fluoroquinolone-resistant CFUs were observed in the current study, although a proportion of fluoroquinolone-resistant CFUs of up to 10% was observed in some cases in the study before 3 years. This could be because the farm stopped using preventative antibiotics after the piglets arrived. This shows that management practices have a major influence on the emergence of antibiotic resistance. Furthermore, although all piglets in control group 2 (Supplementary Table S1) were treated with enrofloxacin during the third week of the experiment, phenotypic resistance to ciprofloxacin was not observed, even though qnrS1 (fluoroquinolone resistance gene) was detected in the fourth week.

This research contributes to understanding how hygiene and other factors influence AMR dynamics in the piglet barn. The lack of clear reduction of AMR bacteria in the improved hygiene groups compared to the control groups and the lack of resistance to the regular used disinfectants suggest that the hygiene level in the studied conventional farm was already high. The effects of an improved hygiene might be better visible in farms with a lower level of hygiene. In conclusion, while managing hygiene is essential, it alone is insufficient to significantly reduce AMR in piglet rearing. Our findings highlight the multifaceted nature of AMR spread in piglet barns and the need for a comprehensive management strategy that addresses the various contributing factors to effectively combat AMR.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

MJ: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. VJ: Data curation, Investigation, Methodology, Validation, Writing – review & editing, Resources. BK: Formal analysis, Software, Visualization, Writing – review & editing. MB: Data curation, Investigation, Writing – review & editing, Validation. DW: Conceptualization, Methodology, Resources, Supervision, Funding acquisition, Writing – original draft. CD: Project administration, Resources, Writing – review & editing. JS: Data curation, Formal analysis, Investigation, Visualization, Writing – review & editing, Resources. CH: Conceptualization, Methodology, Resources, Supervision, Writing – review & editing, Visualization. JZ: Supervision, Writing – review & editing. TA: Conceptualization, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing, Resources. UN: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Software, Supervision, Writing – review & editing, Validation. TK: Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing, Conceptualization.