Prevalence and Antibiotic Resistance Prole of Salmonella spp. in Broiler Carcasses From Dominant Poultry Production Areas in Bhutan

Background Salmonella is an important zoonotic pathogen and its infections are considered among the most commonly and widely distributed food-borne illness reported worldwide. Poultry products have been identied as important sources of Salmonella infection to humans. While there have been reports of high prevalence of Salmonella and multidrug resistance in imported chicken meat in Bhutan, the safety of nationally produced broiler meat with regards to Salmonella contamination is not known. In the absence of national surveillance of Salmonella in food animals in Bhutan, this study assessed the prevalence of Salmonella and its serotypes in broiler chicken carcass in the dominant poultry production areas of the country and determine its antibiotic susceptibility patterns. in in reported 13% of Salmonella with the that is and good quality. a for Salmonellosis in no studies conducted to establish Salmonella prevalence in produced Evidence on the epidemiology of Salmonella in produced chicken meat in Bhutan, including data on antibiotic resistance patterns of Salmonella isolates, will help improve the understanding of occurrence and burden of Salmonellosis at food animal level and subsequently leverage the need for routine monitoring and surveillance in poultry. study was carried out to assess the prevalence of Salmonella and its serotypes in broiler chicken meat in high poultry production areas in Bhutan and investigate antibiotic sensitivity proles of recovered Salmonella poultry serotypes. The ndings will help inform the national Salmonella control plan to avert the public health burden associated with Salmonella infection. S. S. S. Our results also indicate a high proportion of Salmonella isolates were sensitive to gentamycin 73.91% (17 isolates) and streptomycin 56.52% (13 isolates) but high resistance to tetracyclines 95.65% (22 isolates), trimethoprim 86.96 (20 isolates) and amoxicillin 65.20 (15 isolates). Our study ndings were in consistent with the ndings of (Murugkar et al., 2005) which labelled doxycycline, ampicillin, amoxicillin and tetracycline resistant to Salmonella isolates but, sensitive to noroxacin, enrooxacin, gentamicin and ciprooxacin. In the Asia region, high resistance of Salmonella was reported for tetracycline (97.14%) and chloramphenicol (94.28%), in chicken farm in Bangladesh (Alam et al., 2020; Mridha et al., 2020), trimethoprim sulfamethoxazole (70.3%) and tetracycline (54.3%) in chicken meat in Myanmar (Moe et al., 2017), sulphonamide compounds (98.9 %) and tetracycline (96.9 %) in chicken carcass in Nepal (Goncuoglu, Ormanci, Uludag, & Cil, 2016). The emergence of high resistance of Salmonella to tetracycline, amoxicillin and trimethoprim antimicrobials in the study may be contributed to indiscriminate use of these antimicrobials in poultry production as growth promotion, prophylactic and therapeutic purposes (Nourouzi, Mirzaii, & Norouzi, 2004) and Bhutan is no exception. As per the drug distribution report of Bhutan, 2018, the tetracycline trimethoprim A cross-sectional study conducted from 2016 to assess the prevalence of Salmonella and its serotypes in broiler chicken meat in high poultry production areas in Bhutan and to investigate their antibiotic susceptibility prole. The two major areas that produce and supply commercial chicken broilers in the country – the gewogs of Samphelling and Darla located in the Chukha district – were selected for this investigation (DoL, 2015). The study focused on sampling chicken broilers produced in commercial poultry enterprises. For the purposed of the study, commercial chicken broiler farms were dened as farms which have more than 2,000 birds with a permanent shed. As per the sampling frame, there were a total of ve commercial broiler farms in both the study areas fullling the commercial farm denition. All ve commercial broiler farms (Farm A, B, C, D and E) were included in the study. These enterprises were run by different farmers under different production management. We collected farm biosecurity management data including number of years in operation, presence of perimeter fencing, presence of disinfection points, presence of biosecurity sign boards, presence of visitor logbook, source country of DOCs, vaccination and deworming protocols, mortality and treatment record, source of feed, treatment of water, presence of a bio pit, utilisation of own farm equipment, use of PPE (ie. gloves, mouth covers, boots and apron), presence of meat processing facility, cleaning and disinfection protocols (animal shed, farm disinfection, disinfectant use, premise clean, drainage clean). The information on farm biosecurity management was collected by the lab technicians using pretested questionnaires. The ling of questionnaires involved an interview, observations and validation methods. of as a The epidemiological analyses of antibiogram using WHONET which is an tool that facilitates of the local epidemiology of microbial and of antimicrobial agents Each farm-level biosecurity indicator was scored a value of “1” when the response represented an adequate biosecurity practice and “0” when the response was less adequate. A total farm-level biosecurity score was estimated as the sum of all responses to the questionnaire. Results were compared descriptively against estimates of prevalence.


Abstract
Background Salmonella is an important zoonotic pathogen and its infections are considered among the most commonly and widely distributed food-borne illness reported worldwide. Poultry products have been identi ed as important sources of Salmonella infection to humans. While there have been reports of high prevalence of Salmonella and multidrug resistance in imported chicken meat in Bhutan, the safety of nationally produced broiler meat with regards to Salmonella contamination is not known. In the absence of national surveillance of Salmonella in food animals in Bhutan, this study assessed the prevalence of Salmonella and its serotypes in broiler chicken carcass in the dominant poultry production areas of the country and determine its antibiotic susceptibility patterns.

Methods
A cross-sectional study was conducted in January to April 2016 in all ve major commercial broiler meat producers and suppliers in Bhutan located in the Samphelling and Darla gewogs. Following a systematic random sampling method, 36 broiler carcasses were sampled from each of the ve broiler commercial farms producing the total sample size of 180 samples. Salmonella was isolated and identi ed following the International Organization for Standardization methods (ISO 6579:2002). All Salmonella isolates were subjected to antibiotic susceptibility testing (AST) following the procedures of the Clinical and Laboratory Standards Institute (CLSI) method.

Conclusion
The estimates of Salmonella prevalence and multidrug resistance in nationally produced commercial broiler chicken carcasses from the two main broiler meat producing areas in Bhutan, raises signi cant public health concerns highlighting the need for an integrated national Salmonella surveillance program across the poultry market chain.

Background
Salmonellosis caused by non-typhoidal Salmonella enterica, Salmonella Paratyphi and Salmonella Typhi have been recognized as a globally important foodborne illness of public health and economic signi cance in both humans and livestock (Westrell et al., 2009). It was estimated that poultry products such as eggs and fresh meat are the most commonly implicated animal products linked to human Salmonellosis (Control & Prevention, 2013), Pires et al., 2009) and have been identi ed as the main vehicles for Salmonella infection in humans (Authority, Prevention, & Control, 2018).
Salmonellosis is the third leading foodborne cause of death worldwide behind norovirus (120 million cases) and Campylobacter spp. (96 million cases) and the leading foodborne illness (7.6 million cases) in terms of global disease burden (WHO, 2015). The global deaths from the food borne illness due to non-typhoidal Salmonella enterica accounted for 59,000 deaths of the total 2,30,000 deaths due to foodborne diarrhoeal disease agents (WHO, 2015). Likewise, the same study reported Salmonella typhi as the major non-diarrhoeal causes of foodborne deaths (52000) globally (WHO, 2015 (Havelaar et al., 2015). In the South East Asia subregion, the median burden due to non-typhoidal Salmonella enterica was reported at 1,042,715 (95% CI, 225,416-2,824,443) DALYs which translates to 910 (95% CI, 89 − 4,760) DALYs per 100,000 population and ranked second behind African subregion (Havelaar et al., 2015). Similarly, the foodborne burden due to Salmonella typhi and Salmonella paratyphi A was reported at 250 illness and 58 illness per million population respectively in SEARO (Havelaar et al., 2015). The study on global burden of invasive nontyphoidal salmonellosis (iNTS) conducted in 2010, reported 3.4 million cases of iNTS per year which corresponds to 49 cases per 100,000 population (Ao et al., 2015).
Bhutan is largely an agrarian economy with 79% of its population engaged in agriculture and livestock farming (NSB, 2017). Approximately, 46% of households in Bhutan rear poultry for producing eggs and chicken (NSB, 2017). Since 2013, Bhutan has achieved 100% self-su ciency in egg production (DoL, 2018). Yet, Bhutan still imports large volumes of chicken meat primarily from India to meet national consumer demand. In 2018, Bhutan imported more than 1500MT of chicken meat to meet the national consumer demand (RSD, 2018). Reports from Bhutan's neighbouring countries Bangladesh and  (Helke et al., 2017). In Bhutan legislation prohibits the use of antibiotics in feed as a growth promoter. The Drug Regulatory Authority (DRA), and Bhutan Agriculture & Food Regulatory Authority (BAFRA), Bhutan carries out regular inspections and monitoring of the feed plant to ensure antibiotics are not used by the feed manufacturers. Vaccines and antibiotics for poultry diseases are administered by livestock extension workers and are given free of cost by the government. However, anecdotal reports indicate that poultry farmers in bordering regions have access to antibiotics across the border and use them as prophylactics.
A study conducted in 2007 in imported chicken carcasses in Bhutan reported 13% prevalence of Salmonella with Salmonella enteritidis as the most prevalent serotype (84.62%) compared to Salmonella typhimurium (15.38%) (Dahal et al., 2007). Based on the ndings of this study, new provision was added to the legislation which requires importers to mandatorily produce sanitary certi cates from the Export Inspection Council (EIC) in India certifying that chicken meat is safe and good quality. While a regulatory framework for the control and prevention of Salmonellosis in imported poultry meat into Bhutan is presently enforced, the safety of locally produced chicken meat remains a public health concern as no studies have been conducted to establish Salmonella prevalence in nationally produced poultry. Evidence on the epidemiology of Salmonella in domestically produced chicken meat in Bhutan, including data on antibiotic resistance patterns of Salmonella isolates, will help improve the understanding of occurrence and burden of Salmonellosis at food animal level and subsequently leverage the need for routine monitoring and surveillance in poultry.
This study was carried out to assess the prevalence of Salmonella and its serotypes in broiler chicken meat in high poultry production areas in Bhutan and investigate antibiotic sensitivity pro les of recovered Salmonella poultry serotypes. The ndings will help inform the national Salmonella control plan to avert the public health burden associated with Salmonella infection.

Farm and region level prevalence estimates of Salmonella contamination of broiler carcasses
Out of a total of 180 samples examined, the prevalence of Salmonella in broiler meat from the two major areas of broiler meat producer and supplier in Bhutan was 12.78% (95% CI, 10.98-14.58). The difference in prevalence of Salmonella spp. between Samphelling gewog and Darla gewog was 5.09 % (95% CL, 1.66-8.52). The prevalence of Salmonella spp. in Samphelling gewog 14.81% (95% CI, 8.2.7-21.5) was higher than in Darla gewog 9.72% (95% CI, 2.9-16.6). Similarly, between the farms sampled, Salmonella prevalence was the highest in farm C 19.44% (7 isolates) followed by farm A 13.89% (5 isolates) and farm B & D 11.1% (4 isolates) and farm E 8.3% (3 isolates) ( Table 1). Farm C had one of the lowest biosecurity scores (14), one of lowest population sizes (2,050 head of poultry), was relatively recent operation (4 years) and imports DOC from India. There was no apparent relationships between farmlevel seroprevalence and biosecurity indicators (Table 1). Among the 23 Salmonella isolates, two Salmonella serovars were identi ed. The S. typhi 73.9% (17 isolates) serovar was found more predominant than S. paratyphi B 26.1% (6 isolates). both the Salmonella serovars were prevalent in all farms in the study areas except for farm A (Table 1).

Antibiotic susceptibility pro le of Salmonella isolates
Our results demonstrate that the majority of Salmonella isolates were found sensitive to gentamycin 73.91% (17/23 isolates) and streptomycin 56.52% (13/23 isolates) ( patterns of multidrug resistance were established. About 95.65% (22/23 isolates) of the Salmonella isolates showed resistance to one or more antibiotics as shown in Fig. 1. Between serovars S. typhi and S. paratyphi B showed 94.1% (16/23 isolates) and 100% (isolates) resistance to one or more antibiotics respectively.   The study by (Wang et al., 2013) concluded that the variation of prevalence of Salmonella between farms could be attributed to the farm chicken breed.
Differences in the prevalence between the study farms can be attributed to differences in bird density as higher bird density has been noted in previous studies to be associated with increased odds of detection of positive Salmonella (Donado-Godoy et al., 2012). In Bhutan, commercial poultry (broiler and layer) are reared in deep litter production system and depending on the size of the farms, farmers usually have more than two poultry shed and each shed has the capacity of more than 1000 birds. All ve commercials' farms had the farm size between 2,000-2,500 birds and follow all-in all-out system.
The source of poultry could also pose a biosecurity risk for the introduction of Salmonella into a ock. In our study broiler day old chicks used by these In addition, to husbandry systems other factors could be implicated in the variation in prevalence observed in our study. As per (Wang, 2016), the difference between Salmonella prevalence in our study and others reported in the literature could be due to difference in the study design particularly number of samples collected, sample collection period, sample type, sampling procedures and method of detection used. Studies conducted in China, Greece and the USA described that prevalence and concentration of Salmonella in chicken was relatively higher in summer and spring than winter and autumn ( Our results also indicate a high proportion of Salmonella isolates were sensitive to gentamycin 73.91% (17 isolates) and streptomycin 56.52% (13 isolates) but high resistance to tetracyclines 95.65% (22 isolates . The emergence of high resistance of Salmonella to tetracycline, amoxicillin and trimethoprim antimicrobials in the study may be contributed to indiscriminate use of these antimicrobials in poultry production as growth promotion, prophylactic and therapeutic purposes (Nourouzi, Mirzaii, & Norouzi, 2004) and Bhutan is no exception. As per the drug distribution report of Bhutan, 2018, the tetracycline trimethoprim was largely distributed for the poultry use. The consistent reports of high resistance of Salmonella to tetracycline, trimethoprim and amoxicillin antimicrobials in several studies conducted globally raise serious public health concern since these antimicrobials are used as rst line drugs for human and animals. Our study also observed that 95.65% (22 isolates) of the Salmonella isolates showed resistance to one or more antibiotics with six antibiotic resistance patterns and 60.87% of multidrug resistance. A study conducted by (Dahal et al., 2007) in imported frozen chicken in Bhutan, also reported Salmonella resistance to one or more drugs antimicrobials namely; nalidixic acid (96.15%), amoxicillin (11.54%) and cephalexin (5.77%). A widespread multidrug resistance of Salmonella was reported in chicken eggs in Nepal , broiler chicken at slaughter house in China (Zhu et al., 2017) and broiler farm in Bangladesh (Mridha et al., 2020). Recognizing the role of meat in Salmonella transmission to humans added by reports of high prevalence of Salmonella and multidrug resistance from the two main producer and supplier of broiler meat areas in Bhutan, altogether, it raises serious public health concern which require immediate action.
The ndings of our study need to be interpreted in light of the fact that within the ve commercial broiler farms that ful lled the study's inclusion criteria, we collected samples from the same batch of birds in the farm to get the required sample size. While we employed a cluster sampling weight to adjust the sample size calculation, birds in the batch are usually raised in same production system.
In this study we present the rst estimates of Salmonella prevalence and antibiotic resistance pro les in nationally produced broiler meat in the dominant broiler production and supply areas in Bhutan. These ndings should serve as an important resource for national Salmonella control program managers to plan the design of farm-level surveillance to help reduce the human risk of Salmonellosis at the country level.

Study setting and sampling frame
A cross-sectional study was conducted from January to April 2016 to assess the prevalence of Salmonella and its serotypes in broiler chicken meat in high poultry production areas in Bhutan and to investigate their antibiotic susceptibility pro le. The two major areas that produce and supply commercial chicken broilers in the country -the gewogs of Samphelling and Darla located in the Chukha district -were selected for this investigation (DoL, 2015).
The study focused on sampling chicken broilers produced in commercial poultry enterprises. For the purposed of the study, commercial chicken broiler farms were de ned as farms which have more than 2,000 birds with a permanent shed. As per the sampling frame, there were a total of ve commercial broiler farms in both the study areas ful lling the commercial farm de nition. All ve commercial broiler farms ( Sampling methods and procedures. The sample size for the prevalence survey was calculated using the epi tool formula (Sergeant, 2009): where Z: is the value from standard normal distribution corresponding to desired con dence level (for 95% CI), P: is expected true proportion, e: is desired precision. [At 95% CI, α: 5%] and sample size was calculated as n = 180.
During the slaughter time poultry are pooled at the slaughter facility and are slaughtered by trained workers following hygienic practices. Depending on the market demand, the slaughter period of the bird will vary from farm to farm. Usually, the birds are slaughtered between 40-50 days of rearing from the day-old chick (DOC). Since all ve farms in the study follow all-in all-out system, only one production cycle of the broiler birds was considered and 36 samples each from one farm was collected in each visit. The schedule of sample collection was tied with the slaughter date of the broilers in a farm and were collected over 5 weeks period (i.e. one visit every week). Following a systematic random sampling method, 36 healthy broiler carcasses from each of the ve broiler commercial farms in Samphelling and Darla gewogs were selected after the evisceration and before packaging resulting in a total sample size of 180 samples. From the selected broiler carcasses, a breast muscle portion weighing approximately 30g was collected in a sterile plastic sachet by the trained laboratory technicians. The samples were stored at below 4°C temperature in cool box and were transported to the laboratory within 8 hours from the time of sample collection. During the sample collection time the additional information on farm size, farm biosecurity status, source of feed, water and DOC, and management practices was also collected. A nely cut pieces of 25 grams of chicken breast was mixed with 225 ml of buffered peptone water and was shaken for 2 minutes. The mixture was incubated at 37°C ± 1°C for 18 ± 2 hours. After the incubation period, a loop-full of material from the Rappaport-Vassiliadis soya (RVS) broth and Muller Kauffmann tetrathionate novobiocin (MKTTn) was transferred and streaked separately onto the Brilliant green Phenol Red Lactose Sucrose agar and Xylose lysine deoxycholate (XLD) agar separately and were incubated at 37°C for 24 hours. The plates were incubated in an inverted position and observed for the growth of typical Salmonella colonies.
Identi cation of Salmonella was performed on six biochemical con rmation tests namely, Triple sugar agar (TSI agar), Urea agar, L-lysine decarboxylation medium, Voges-Proskauer reaction, Methyl red reaction, and indole reaction. After biochemical con rmation, all Salmonella positive isolates were transferred into half-strength NA in 1.5-mL Eppendorf tubes and stored at 48C.

Antimicrobial susceptibility
Antibiotic sensitivity testing for identi ed Salmonella isolates was carried out using the disc diffusion method described in Clinical and Laboratory Standards Institute (Cockerill, Clinical, & Institute, 2012). A total of six antimicrobials used in the disc diffusion method including Ampicillin (10mcg/disc), Amoxycillin (30 mcg/disc), Tetracycline (30mcg/disc), Streptomycin (10mcg/disc), Trimethoprim (5 mcg/disc) and Gentamicin (10 mcg/disc). These antimicrobials were selected since they are commonly used in poultry industries in Bhutan. The results were interpreted using CLSI methods (Cockerill et al., 2012). The multidrug resistance for this study is de ned as resistance to three or more antimicrobial classes (Falagas,

Statistical analyses
Sample laboratory results and farm-level biosecurity data were entered into Microsoft Excel, version 2016. The data were then transferred into the WHONET o ine database. Prevalence of Salmonella spp. was calculated from the total number of samples tested and the prevalence of Salmonella serovars was calculated from the total number of positive Salmonella spp. isolated. Con dence intervals for a single proportion and difference in proportions were estimated using the statsmodels library in python3.
The sensitivity and resistance to different antimicrobials of Salmonella spp. positive isolates were presented as a proportion. The epidemiological data analyses of antibiogram patterns were managed using WHONET which is an analytical tool that facilitates understanding of the local epidemiology of microbial populations and selection of antimicrobial agents (Agarwal, Kapila, & Kumar, 2009).
Each farm-level biosecurity indicator was scored a value of "1" when the response represented an adequate biosecurity practice and "0" when the response was less adequate. A total farm-level biosecurity score was estimated as the sum of all responses to the questionnaire. Results were compared descriptively against estimates of prevalence.

Declarations
Ethics approval and consent to participate Not applicable Consent for publication Not applicable Availability of data and materials The authors con rm that all the raw data supporting the ndings of thisstudy are available within the article and its supplementary materials.

Figure 1
Percentage Resistance of Salmonella serovar recovered from raw broiler meat to six antimicrobial drugs