Previously titled: Prevalence and diversity of Salmonella isolated from layer farms in central Ecuador

Background: Given the considerable role played by Salmonella in the incidence of food contamination, around the world, surveillance of this infection is prioritized by both food producers and health care authorities. Data remains insufficient concerning the prevalence of Salmonella in poultry systems in Ecuador and in Latin America in general. Methods: In this study, we evaluated the presence and diversity of Salmonella serovars in samples taken from 21 layer farms and backyard layers in central Ecuador during August-November 2017. Salmonella was isolated following standardized methods (ISO 6579) and the serovar determination was carried out by PCR. Results: A significant presence of Salmonella was detected in the 21 farms evaluated, with a frequency of 76% (95% confidence interval (CI): 53-92) in environmental surfaces, 33% (95%CI: 15-57) in pooled cloacal swabs from layer hens, 33% (95% CI: 13–59) on feed samples, and 10% (95%CI: 1-30) in backyard layer feces from traditional local markets. The dominant serovars detected were S. Infantis and S. Typhimurium. Conclusions: This study forms a basis for further surveillance of Salmonella serovars in layer farms in central Ecuador.


Amendments from Version 1 Introduction
The genus Salmonella is considered a leading cause of foodborne illnesses around the world (WHO, 2017). These bacteria are among the most significant agents of food and water poisoning in the United States and Europe (Bäumler et al., 2000;Callejón et al., 2015;Varma et al., 2005). Globally, it is estimated that 93.8 million cases of Salmonella-related gastroenteritis occur annually, resulting in 155,000 deaths (Majowicz et al., 2010). In Ecuador, typhoid, together with paratyphoid fever, causes around 1,500 hospitalizations per year, while non-typhoidal salmonellosis leads to more than 2,000 hospitalizations over the same period. These infections have accounted for approximately 25% of total reported gastrointestinal illnesses in recent years (MSP, 2018). Salmonella represents a complex and diverse genus, but only a small number of serovars are involved in human infections (Issenhuth-Jeanjean et al., 2014;Thiennimitr et al., 2012).
Salmonella detection and the investigation of foodborne outbreaks is of maximum importance to public health, which has resulted in the establishment of epidemiological surveillance programs in many developed countries (EFSA, 2013). These measures provide information about endemic Salmonellaserovar patterns, outbreaks, temporal trends and the monitoring of control actions (CDC, 2018). The principal reservoir of S. enterica is the intestinal tract of livestock, representing one of the main sources of infection for humans (Antunes et al., 2016). Despite this epidemiological and economic importance, in South America there is a paucity of information concerning Salmonella in poultry systems (Alexandre et al., 2000;Donado-Godoy et al., 2012). Research conducted in 2016 pertaining to broiler chicken farming in Ecuador indicated the presence of Salmonella serotypes S. Infantis , S. Enteritidis and S. Corvallis (Vinueza-Burgos et al., 2016). However, among layer hens, no information about the prevalence or diversity of Salmonella has been reported in this country nowadays. The purpose of this study was to estimate the presence and diversity of Salmonella bacteria present in 21 layer farms in central Ecuador.

Samples
Samples were collected between August-November 2017 from different layer farms in central Ecuador (Tungurahua and Cotopaxi provinces), which accounts for around 60% of egg production in the country. A total of 21 farms (>1,000 birds) in Latacunga, Cevallos, Quero and Ambato (all Ecuador) were sampled, based on their willingness to provide verbal consent for this study (verbal consent was obtained over written consent owing to the farmers' reluctance to sign their names, as they perceived this could be used to identify them). Further details for each site can be found in Dataset 1 (Calero-Cáceres, 2019a)). One laying hen house per farm was selected. The following samples were collected in each house: 21 pooled cloacal swabs (10 cloacal swabs per pool); 21 manure drag swabs (environmental swabs, 1 per business); 21 caecum content samples (1 layer per farm). To evaluate the potential risk from contaminated feed, 18 composite samples were taken from farmyards (18 of the 21 farms consented verbally to having these samples taken). Additionally, 21 fecal samples from backyard layers were sampled in traditional local markets. All samples were transported in an icebox at 3-5°C within 2 hours of collection for bacterial isolation. The experiment was performed under supervision of the ethical committee of the Faculty of Agricultural Sciences, Universidad Técnica de Ambato.

Detection of Salmonella
Salmonella was isolated following standardized methods (ISO 6579) (ISO, 2017). The samples were pre-enriched in buffered peptone water (Oxoid, Basingstoke, England) and then incubated at 37±1°C for 18 h ± 2 h. Next, Rappaport Vassiliadis Soy Broth (RVS Broth) (Merck Millipore, Darmstadt, Germany) was used a selective medium, being inoculated with the pre-enriched culture and incubated at 41.5±1°C for 24±3 h. One loopful of the selective enrichment medium was streaked onto xylose lysine deoxycholate agar (XLD agar) (Becton Dickinson GmbH, Heidelberg, Germany) and incubated at 37±1°C for 24±3 h. Presumptive Salmonella isolates (identified as red/yellow colonies with a black center) were purified in Mac Conkey agar (Merck, Darmstadt, Germany) and incubated at 37±1°C for 24 h. Isolates were Gram stained and the following biochemical tests were performed for confirmation the genus Salmonella: a catalase test using 30% hydrogen peroxide (Merck Millipore, Darmstadt, Germany); triple sugar iron agar test (TSI) (Becton Dickinson GmbH, Heidelberg, Germany), Simmons citrate agar test (Merck, Darmstadt, Germany), Christensen urea agar test (Britania Lab., Buenos Aires, Argentina), and indole reaction using tryptone water (Merck, Darmstadt, Germany) and Kovac's reagent (Sigma Aldrich, St. Louis, USA). One isolate per positive sample was selected and cryopreserved using overnight growth in LB broth (Sigma Aldrich, St. Louis, USA) supplemented with 30% glycerol (Merck Millipore, Darmstadt, Germany) and maintained at -80°C until analysis.
Serovar determination by PCR PCR assays were performed to identify the genes under specific conditions (Table 1). One virulence gene related with fimbrial cluster (bcfC) was evaluated as target of S. enterica specie (Zhu et al., 2015). For serovars and biotypes: A modification methylase gene that are specific of S. Infantis (M.SinI) (Ranjbar et al., 2017). 23S rRNA gene associated to S. Typhi The reaction mixture contained: 12.5 µl of DreamTaq Green PCR Master Mix (Thermo Fisher Scientific, Massachusetts, USA), 0.5 µl of each primer (30 µM stock), 9 µl of nuclease-free water (Thermo Fisher Scientific, Massachusetts, USA), and 2.5 µl of crude DNA were used. PCRs were performed with an Applied Biosystems SimplyAmp Thermal Cycler (Thermo Fisher Scientific, Massachusetts, USA). A total of 10 µl of each PCR product were analyzed by agarose gel electrophoresis and stained using Sybr® Safe DNA Gel Stain (Invitrogen, Carlsbad, USA).

Results
Assessing the presence of Salmonella Overall, 31 out of 34 isolates which showed phenotypic characteristics in accordance with Salmonella were confirmed by PCR as S. enterica (Table 2)   The notable presence of Salmonella in poultry feed suggests that this may constitute a potential reservoir of this bacteria for poultry systems in the evaluated region, and consequently, a potential route of infection and colonization of poultry and subsequent entry into the food chain (Fink-Gremmels, 2012; Jones, 2011). In recent research conducted in an integrated poultry farm in Ecuador, 4.1% (8/194) of samples were positive for Salmonella, particularly in animal-based feed (Villagómez Estrada et al., 2017). Therefore, the accurate detection of this pathogen in feed is necessary in order to identify the critical points of contamination (facilities, raw materials, transport, storage, producers), that one may apply effective measures to reduce the risk of transmission.
The finding of only S. Infantis in poultry feed may be attributed to the high level of persistence of this serovar over time in poultry feed, which in turn results in its considerable presence in Ecuadorian broiler chicken (Andino et al., 2014;Vinueza-Burgos et al., 2016). Although the origin of this serovar in feed it is not well defined, studies have pointed to crosscontamination with feces, persistent contamination of storage bins and surfaces, and poor ingredient selection as main causes of feed contamination with Salmonella (Davies & Wray, 1997;Huss et al., 2018;Laban et al., 2014). Genomic tools, such as multilocus sequence typing, BOX-PCR, (GTG)5-PCR or whole-genome sequencing may help to identify the origin of feed contamination in a larger study.
Drag swabs revealed the presence of two different serovars in the sampled poultry farms: Infantis and Typhimurium. These non-typhoidal S. enterica serovars are commonly associated with poultry systems and are linked to outbreaks of foodborne illness (Anderson et al., 2016;Pui et al., 2011). Serovars vary in their persistence over time and geographic distribution around the world (Hendriksen et al., 2011), making further studies desirable in order to evaluate the variations in serovar persistence in the locations sampled in this research.
In backyard layer feces sampled at local markets, S. Infantis alone was detected (2/2), but further evaluation of Salmonella serovars in backyard flocks is recommended in order to improve surveillance of this potential source of salmonellosis (Behravesh et al., 2014). In pooled cloacal swabs, the serovars detected were Infantis (2/7), Typhimurium (2/7) and unidentified serovars (3/7). Infantis and Typhimurium serovars are commonly detected in poultry around the world (Foley & Lynne, 2008;Foley et al., 2011). Complementary analysis by serotyping the unclassified serovars is necessary in order to identify Salmonella bacteria not covered by the panel.
Data remains insufficient concerning the prevalence of Salmonella in poultry systems in Ecuador and in Latin America in general. The coordination of similar future studies may provide a starting point for surveillance of zoonotic bacteria within a defined public health area, leading to an improvement in policies and safe practices in the food industry. Such studies may reduce the risk of infection and establish protocols for corrective measures to be implemented in key upstream points of the chain as indicated by the data.
At the same time, the intervention of public health authorities may be required in order to ensure the participation of a fully representative range of poultry businesses in future research. This study depended on the voluntary consent of farm owners to providing samples, which may have led to selection bias. In order to establish a thorough program of surveillance, all potential sources of Salmonella infection in the region need to be made accessible to researchers.

Eloy Gonzalez-Gustavson
Facultad de Medicina Veterinaria, Universidad Nacional Mayor de San Marcos, Lima, Peru The scientific names in the manuscript need to be reviewed.
The author use incorrectly the terms prevalence and incidence as synonyms.
Incidence is defined as the presence of new cases of an illness over a period of time. The manuscript does not correspond to an incidence study.
I would like to see more information in the manuscript about the design to consider the results as prevalence. I have not found details about the representativeness of the sampling (sample size to get a prevalence) or if they were obtained randomly. If the method does not satisfy the requirement to be considered as prevalence, I recommend replacing the term prevalence by frequency, presence or proportion.
The confidence interval in one case was incorrectly estimated. "Feces from backyard layers showed the presence of Salmonella in 2 of 21 samples (10%, 95%CI: 0-22)" This values have to be calculated based on binomial distribution. Actually in the manuscript, the confidence interval was estimated based on normal or t distribution and 0 was included as lower interval.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes

Introduction
In the fourth line you would change the word "poisoning" by contamination At the end correct one "and" that is in italics When you say "among the layer hens, no information about prevalence or diversity of Salmonella has been reported", you could specify or support if that lack of information is only in Ecuador or at a South American or global level. In the same way, it could important to include some epidemiological data of salmonelosis in Ecuador.

Methodology
I think that is not necessary to do DNA extraction to perform PCR, because Salmonella an is gram negative bacteria and during the start of the PCR where the temperature rises to 90 ° C it is possible to destroy cellular wall and release the DNA. However, it does not affect the results, I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias • You can publish traditional articles, null/negative results, case reports, data notes and more • The peer review process is transparent and collaborative • Your article is indexed in PubMed after passing peer review • Dedicated customer support at every stage • For pre-submission enquiries, contact research@f1000.com