Prevalence of Antibiotic-Resistant Bacteria in the Environment of Poultry Farms

The occurrence and spread of antibiotic-resistant bacteria (ARB) due to the high demand in poultry industries are of great public health concern. Indiscriminate and abusive use of various antibiotics on a large scale causes antibiotic resistance (AMR) in animal-associated bacteria that may be pathogenic to humans. These bacteria are widely disseminated in the environment via animal waste. This study was therefore designed to assess the prevalence of multiple antibiotic resistance (MAR) among the environmental bacteria in poultry farms and to determine the risk contamination category of poultry by calculating the multiple antibiotic resistance index (MARI).

(AMR), as it is a bloc of fast-developing and interconnected economies [5,9,10,11]. In recent decades, countries in SEA (Vietnam, Thailand, Indonesia) have been experiencing rapid growth in aquaculture and poultry production sectors, representing 48% of the global veterinary antibiotic market [11,12,13].
Prevention and therapy using antibiotics has become one of most favorable approaches for these countries to avoid any uncontrolled epidemic diseases that can jeopardize the economy [14].
The indiscriminate and abusive use of antibiotics can result in higher concentrations of antibiotic residues in the environment [15]. Resistance to antibiotics is a side effect associated with this pollution [15,16]. Several authors have investigated the prevalence of AMR in numerous environmental samples.
Similarly, Zhu et al. noted high levels of tetracycline in manure and soil samples collected from large commercial swine farms in China [18].
Undoubtedly, animal husbandries are an important component in understanding the interplay of ARB between humans, animals and the environment [19,20,21]. In summary, ARB diffuse between humans and animals through environmental sources. AMR can be transmitted from environmental bacteria to those that colonize humans and animals [22]. This paper aims to assess the prevalence of multiple antibiotic resistance (MAR) among environmental bacteria in poultry farms and to determine the risk contamination category of poultry by calculating the multiple antibiotic resistance index (MARI). The results provide preliminary ndings on ARB distribution in the poultry environment, as this environment is one of the potential sources of ARB.

Antibiotic resistance in environmental bacteria
A total of 511 bacterial isolates were sampled from soil and e uent samples. Figure 1 shows the AMR patterns in gram-negative and gram-positive bacterial isolates and their prevalence rates. Detailed information is provided in Table 1, which comprises a list of antibiotics tested for gram-negative and gram-positive bacteria, the total number of bacteria resistant to each antibiotic and the prevalence rate of this resistance, which was calculated based on the number of ARB divided by the total number of bacteria that were analyzed. Based on the Mann-Whitney U-test, there was no signi cant difference between isolates from soils and e uents in resistance against antibiotics (U = 32144, p-value = 0.943). Therefore, the samples were further discussed as overall environmental samples. According to Fig. 1, gram-negative bacteria isolated from poultry farms were highly resistant to ampicillin (79.3%), cefazolin (86.8%), ampicillin/sulbactam (46.65%) and trimethoprim-sulfamethoxazole (39.8%). To our surprise, 1.8% of the isolates were resistant to meropenem, a drug used clinically to treat against extended-spectrum beta lactamase (ESBL) organisms.
Prevalence of multidrug resistance in the poultry environment A total of 372 isolates of gram-negative bacteria and 138 isolates of gram-positive bacteria were tested for antibiotic susceptibility. Figure 2 below shows the number of antibiotic-resistant gram-negative and gram-positive bacteria in each district. In total, 58.2% of the isolates were resistant to at least 3 or more antibiotics tested. District C had the highest percentage, 85.7%, followed by district E, at 69.7% (Fig. 2).
There were 32 bacterial isolates with resistance to at least 10 antibiotics, 4 of which showed resistance to 16 antibiotics. Only 52 bacterial isolates (10.18%) were sensitive to all antibiotics tested.
A Kruskal-Wallis H test showed that there was a statistically signi cant difference in the resistance to different antibiotics among 9 districts (χ2(2) = 34.79, p < 0.001). A signi cant difference was also observed among 33 farms (χ2(2) = 80.54, p < 0.001). Details on the percentage of resistance in the farms and districts are available in Table 2. Additionally, the Mann-Whitney U-test was conducted to analyze the effect of the types of chickens reared (broiler, free range) on the farms as well as the types of coop systems (closed system, open system) being applied. This test showed that isolates from broiler chicken farms exhibited a signi cantly higher percentage of resistance than did isolates from free range chicken farms (U = 22515, p < 0.0001).
Regarding the effect of coop system, the open system exhibited a signi cantly higher percentage of resistance than did the closed system (U = 25402, p < 0.05). The line graphs in Fig. 3 show the comparison of the number of isolates with different amounts of AMR between broiler and free range chicken farms, as well as in the number of isolates with different amounts of AMR between open and closed systems.

Multiple antibiotic resistance index determination
Many isolates acquired resistance to multiple antibiotics. The MARI was calculated, and Table 2 and Figure 4 present the MARI values of all the isolates from the poultry farms. In general, half of the isolates (54.01%) from the soil and e uent samples had a MARI value higher than 0.2. The MARI values ranged widely, from 0.00 to 0.89. There were 100 isolates with very high values of more than 0.40 (49.5%), 10 of which had extreme values of more than 0.79 (10.89%). There were 49 isolates whose values were between 0.20 and 0.25; this range is considered to indicate ambiguity, and these values need extra scrutiny [23].

Discussion
This study is a preliminary approach for exploring the percentage of resistance in bacteria isolated from poultry farming environments. There were high levels of resistance to certain antibiotics, those commonly used for human health. Some of these antibiotics are classi ed by the WHO as critically important antibiotics. The resistance rates were substantially high for some of those antibiotics, especially ampicillin (79.3%), erythromycin (63.7%), linezolid (30.2%), rifampicin (23.1), gentamicin (16.3%) and cipro oxacin (15.4%). There was also a high level of resistance to antibiotics that are listed as highly important antibiotics in human medicine, including cefazolin (86.8%), fusidic acid (84.6%), clindamycin (65.5%), benzylpenicillin (51.9%) and trimethoprim-sulfamethoxazole (45.3%) [24].
According to Gumphol et al., it was suspected that compared with high-income countries, low-and middle-income countries used higher amounts of antibiotics categorized as medically important [7]. This was supported by the ndings of veterinary drug residues in food samples, which indicated that violation rates in developing countries were higher than those in developed countries [25].
In Malaysia, some studies have been conducted on ARB isolated from poultry [25][26][27][28]. Results similar to those in this study, which was centered in a poultry environment, were obtained. A high percentage of resistance to the antibiotics ampicillin, erythromycin and tetracycline, was found in both the poultry and the environment, and most of the bacteria isolated were resistant to multiple antibiotics. More than half of the isolates (58.2%) in this study showed MAR.
The MAR observed in this study may be attributed to the indiscriminate use of antibiotics in poultry production. The resistance percentage, which was signi cantly higher in broiler farms and closed systems, may have in uenced the presence of multi-antibiotic-resistant bacteria and increased the likelihood of contamination on farms. Other studies have shown that poultry type can in uence the contamination and the presence of multi-antibiotic-resistant bacteria [29,30]. This was further proven by the analysis in this study, which showed that 39.53% of the MARI values indicated high risk MARI and 14.48% were ambiguous. This MARI value distribution is alarming, as it shows that isolates originated from high-risk source(s) of contamination with antibiotics, which may increase the possibility of transfer of mobile antibiotic genes between bacteria [31].
This study showed that poultry farms have been contaminated with ARB. Chicken type, coop system and location played important roles in determining the signi cance of this resistance. Environmental bacteria have been exposed to antibiotics for a very long time, and now, there are multi-antibiotic-resistant bacteria, whose presence can affect the effectiveness of antibiotics overall, not only in animal husbandry [32].

Study limitations and opportunities for future research
Some limitations were found in this study. First, we were not able to obtain complete information regarding the types of antibiotics used in the poultry farms. This is because we believe that most of the antibiotics were mixed in with the feed and other supplements. In addition, the antibiotics used are controlled by the supplier, who has full control over food and medication, and it was di cult to obtain this information from the supplier. Second, due to time and budget constraints, we were not able to quantify the presence of antibiotics in the environmental sample; therefore, we could not link the ARB with the presence of trace antibiotics. Third, the classi cation of the chickens and farms was broad. Despite these limitations, our study provided multidisciplinary data that were su cient to analyze the prevalence of ARB in a poultry environment.

Conclusions
In conclusion, this study con rmed the presence of multi-antibiotic-resistant bacteria in poultry farms. The prevalence of MAR and the MAR index, together with the resistance patterns of each isolate, showed that AMR is present not only among humans and animals but also in environmental bacteria. Antibiotic usage in poultry farms, either directly or indirectly, could be one of the factors causing the increase in MAR, and there are always possibilities for various effects, including MAR in environmental bacteria as well as occupational-related risks.
Regulatory authorities play an important role in controlling the usage of antibiotics in farms. They are responsible for granting marketing authorization and quality control of the antibiotic agents. With proper and limited prescription of the antibiotics, including precise treatment regime, dose, treatment intervals and duration of treatment, the amount of antibiotics secreted could be controlled. We hope that the result of this study contribute to strengthening control measures and policies surrounding antibiotic usage among animals and proper disposal of antibiotics.
Basically, this study is an introduction to the AMR of bacteria in the environment. Further studies need to be done to observe the horizontal transfer of the resistance gene and the overall mobile genetic elements.

Environmental sampling
A cross-sectional study was conducted from January 2018 to October 2019 in 9 districts of Selangor state. This study was conducted in poultry farms that were registered with the state Department of Veterinary Services. In total, there were 213 registered farms, and based on the calculation of sample size, 33 farms were randomly selected. The farms were categorized into 2 types of chicken breeding farms, namely, broiler farms and free-range farms. All the sampling were conducted after getting permission from state Department of Veterinary Services and the farm owners.
Soil and e uent samples were collected for this study by trained personnel. Soil samples were collected from 3 different locations within each farm, namely, at the coop and in areas where the chickens clustered together [33,34]. A metal spade was cleaned, disinfected with 75% alcohol and amed prior to soil sample collection. Soil samples from all 3 locations were pooled together in sterile Ziplock plastic bags [35]. E uent water, if available, was also sampled from the drainage or water-pooled area. E uent samples were collected by using a scoop sterilized using the same method as for the spade and placed in a sterile Ziplock plastic bag. All the samples were transported at 4°C in an ice box back to the laboratory.

Isolation and enumeration of bacteria
All the samples were processed within 24 h. The soil sample was homogenized, and 10 g was weighed before it was dispensed into the rst dilution bottle containing 90 mL of Difco TM Peptone water. The e uent sample was used directly. This was recorded as a 10 -1 dilution. The dilution bottles were vigorously shaken and then left to settle for a few minutes. Tenfold dilutions were made by transferring a 1 mL aliquot from the dilution bottle to a fresh 9 mL peptone water bottle, vortexing and continuing until a ten-fold dilution was reached [36]. One milliliter of each dilution was poured into petri dishes containing trypticase soy agar and spread using a disposable loop. All the agar plates were incubated at 37°C for 24 h. Based on morphology, approximately three colonies were selected from plates containing 30 to 300 isolates and then isolated and puri ed before testing for identi cation and susceptibility.
Identi cation and susceptibility testing of bacteria Gram staining was conducted to classify the isolates into gram-negative and gram-positive bacteria, and for further identi cation and susceptibility testing, the VITEK®2 system was used. VITEX ®2 GN and VITEX ®2 GP cards (bioMérieux, Nurtingen, Germany) were used for bacterial identi cation, while AST-GN83 and AST-P592 cards (bioMérieux, Nurtingen, Germany) were used to determine the minimum inhibitory concentrations [37].

Statistical analysis
All statistical analyses were performed using Statistical Package for Social Sciences [(SPSS) IBM software version 20]. The resistance pro le for each antibiotic was used to calculate the percentage of AMR in the poultry environment. The Kruskal-Wallis H test and Mann-Whitney U-test were conducted to compare the amount of AMR in the bacteria between the farms and districts. A p-value below 0.05 was considered statistically signi cant.

MARI of isolates
MARI was calculated using the formula by Blasco et al. [38]:

Declarations
Ethics approval and consent to participate Informed consent was obtained from all the participants.

Consent for publication
Not applicable.