SALMONELLA IN WILD BOARS (SUS SCROFA): CHARACTERIZATION AND EPIDEMIOLOGY

The large study on Salmonella spp. in the population of wild boars from twelve hunting estates in the South-West Vojvodina, Serbia was conducted with the aim to investigate the prevalence of Salmonella spp. in wild boars and to trace Salmonella sources. The hunting estates had similar epidemiological characteristics, i.e. lowland regions with an intensive management system of wild boars. The prevalence of Salmonella in wild boars was determined and the examination of molecular similarities of strains isolated from wild boars and domestic animals (pigs and poultry from nearby farms) was performed. The total number of 425 wild boars (25.3% of total population), shot on of ﬁ cial hunts, were sampled (425 feces and 425 mesenteric lymph nodes samples) and examined by standard ISO protocols. Subtyping of the isolates was performed and compared by Pulsed-ﬁ eld gel electrophoresis (PFGE). The Salmonella prevalence in the fecal samples was 3.1% and in the lymph nodes was 0.2%. Salmonella Enteritidis was the most dominant serotype. A high molecular similarity was found between Salmonella isolates from wild boars and domestic animals. The proximity of communities and domestic animals, as well as improper removal of animal waste were identi ﬁ ed as important epidemiological factors which signi ﬁ cantly affect the epidemiology of Salmonella in wild boars from lowlands.


INTRODUCTION
There are many potential wildlife reservoirs of Salmonella (game meat species, birds, rodents, foxes) and it is recognized that wild boars carry enteric pathogens more frequently than the wild ruminants [1].Large differences in the prevalence of different Salmonella serotypes in game have been reported between particular regions of Europe, from 0-31.8% [2][3][4][5], but also between different regions in one country, 1.7-19.3%[6].In our previous fi ndings from Serbia a prevalence of 1.4% was reported [7].Apart from some data from Spain on the prevalence of Salmonella in individual hunting estates [8,9], there are no other detailed data on comparative analysis of the difference in prevalence between fenced and open hunting estates.Furthermore, information on source tracking of Salmonella infection is also lacking.
The aim of this study was to determine the prevalence and distribution of Salmonella in wild boar populations in open and fenced intensively managed hunting estates.This was performed by examining the extent of Salmonella carriage in wild boars and tracking the source of Salmonella infection by examination of molecular similarities of strains isolated from wild boars and domestic animals (pigs and poultry) from nearby farms.By limiting research to a specifi c geographic region (South-West Vojvodina) which is a lowland region, where hunting estates have intensive opened/fenced management, and domestic animals live in close proximity, the results may be applied to other, geographically similar hunting estates.

Study area and animals
The Vojvodina region is positioned in the north part of Serbia and with a very fl at terrain.The study was carried out in the period between October and December, during two consecutive hunting seasons (2013-2014) as part of a large project on pathogens occurring in wild boars.Research on the prevalence of Salmonella was supposed to continue but had to be stopped due to the outbreak of African swine fever [10].Animals originated from twelve different hunting estates in the region of South-West Vojvodina (10 large fenced and two open areas).Intensive management of wild boars for hunting purposes had been applied in all twelve hunting estates, which included: habitat management which facilitates the preservation of the natural ecosystem, supplementary feeding and intensive predator control.An important epidemiological fact is that all hunting areas are in close proximity to populated communities and farms (<5 km).The exact number of animals in the hunting area is a variable category, so the related data were collected from the hunting estate staff at the time of the hunt.The total number of wild boars in all twelve hunting estates in the course of the study was estimated to 1677 and in total 425 wild boars, shot on offi cial hunts, were sampled (25.3% of the total population).

The sampling procedures and microbiological procedure
The fecal samples from a total of 425 recently shot wild boars were collected in this study.The fecal samples were taken directly from the rectum (n=425) and the mesenteric lymph nodes (n=425) were aseptically excised during evisceration, 850 samples in total.All samples were transported in a cool box at 4 o C to the laboratory within 3 h.Isolation and serotyping of Salmonella was performed according to the ISO 6579:2002 (Annex D) [11] and the ISO/TR 6579-3:2014 [12] standards.Confi rmation was performed by API 20 Enterobacteriaceae (API 20E) strips.

Characterization of isolated Salmonella strains
In order to investigate the possible relation between Salmonella isolates from wild boar feces (n=13) and domestic animals (n=7) (Table 1), subtyping of the isolates was performed and compared by the Standardized Laboratory Protocol for Molecular Subtyping of Salmonella by PFGE method [13], (Figures 1 and 2).The macrorestriction of the genomic DNA was done with the SpeI and XbaI restriction enzymes.The macrorestriction of the Salmonella Braenderup H9812 strain was used as a molecular size standard.The obtained profi les were statistically analyzed by using Ward's linkage of correlation coeffi cients between PFGE patterns of different genotypes by using the SPSS cluster analysis software (IBM Corp. Released 2012.IBM SPSS Statistics for Windows, Version 21.0.Armonk, NY: IBM Corp.).

Prevalence of Salmonella spp. in wild boars
As shown in Table 2, the presence of Salmonella spp. was confi rmed in 13 wild boars' feces samples, with an overall prevalence of 3.1%.Salmonella was detected in eight hunting estates (66.7% of the total number), with the prevalence ranging from 1.7% up to 33.3% (in D and K estates).Salmonella was detected in only one mesenteric lymph node (0.2% prevalence).Salmonella was also detected in the feces of the same animal so this result did not affect overall prevalence.In four hunting estates Salmonella was not detected.S. Enteritidis was isolated in 1.9% fecal samples, S. Typhimurium in 0.9% and S. Infantis in 0.2% out of all examined samples.S. Enteritidis was isolated in one lymph node (0.2%).

Infl uence of different factors on Salmonella fi ndings
Salmonella was signifi cantly more commonly detected in animals from open hunting estates compared to fenced estates (χ2=28.11,p=0.00001).The prevalence of Salmonella, with statistically signifi cant higher fi ndings in female animals (χ2=8.25,p=0.02) and the prevalence of Salmonella in females heavier than 75 kg and older than 36 months was signifi cantly higher (χ2=8.07,p=0.018) compared to all other categories.

Salmonella characterization
The dendrogram carried out based on the Ward's linkage correlation coeffi cient obtained between the SpeI (Figures 1 and 2) and XbaI PFGE macrorestriction profi les demonstrates the existence of 8 groups of isolates.Two groups of identical isolates were obtained after digestion with the SpeI and Xbal restriction enzymes: group 358, 354 and 333; group 321, 326 and 320.The four groups of highly related isolates (277 and 412), (318, 269), (317, 246, 248 and 249) and (254, 412 and 277) were also obtained.

DISCUSSION Prevalence of Salmonella spp. in wild boars
A relatively low prevalence (3.1%), such as the one found in this study, was also reported in other published studies, as well as in our previous examinations [7].Navarro-Gonzalez et al. [4] determined the prevalence in the feces of 5% from wild boars in the suburban area of Barcelona.Somewhat higher values were 22% and 22.8% from fecal samples in Portugal and Italy, respectively [14,15].In Spain, seroprevalence in wild boars was much higher, 19.3-25.7%[6,8].Data about the seroprevalence are usually higher than data on the prevalence of isolated Salmonella.This has been confi rmed from the data from Italy where 66.5% of serum samples were positive for Salmonella by ELISA, while microbiological assessment showed that 10.8% of the wild boars had Salmonella in their feces [16].These differences are expected, as the presence of antibodies cannot be directly linked to the presence of the pathogen itself.The seroprevalence is a measure of historical exposure, which may or may not correlate to the microbiological burden at the time of sampling.Thus, serology helps to identify the infected population, but is not useful for determining the infection status of individual animals.Therefore, a combination of serological and bacteriological tests is important for monitoring infection in animal populations [16].
In this study Salmonella was detected in two third of sampled hunting estates.The prevalence of Salmonella was very variable, in four hunting estates the presence of Salmonella was not determined, six had a low prevalence (1-7%) and two hunting estates had a very high prevalence (33.3%).A variable prevalence between estates is in agreement with a previous study from Spain where Salmonella was isolated in 12% of the sampled hunting estates, with prevalence ranging from 5% to 33% between estates [9].Also seroprevalence is variable 28.2-51.8%[6].These data are in accordance with Gill [17], who also concluded that Salmonella has a frequency variation across populations.
Ortega et al. [6] speculated that Salmonella presence was more common in high density estates in Spain.In Vojvodina, a more noticeable presence of wild boars from open hunting estates was observed around waste sites, where usually a large amount of improperly removed carcasses of domestic animals were dumped (estate "K").This is directly correlated to statistically signifi cantly higher Salmonella prevalence in boars from open hunting estates compared to fenced estates, albeit open estates were usually associated with a low density population (<5 animals/km 2 ).
Based on the fi ndings from our study, the wild boars category mostly at risk of being infected with Salmonella were females heavier than 75 kg and older than 36 months.A higher prevalence determined in females was similar to the reports from Wahchek et al. [2] and Cano-Manuel et al. [8], although Ortega et al. [6] did not fi nd signifi cant differences in seroprevalence between male and female animals.While higher prevalence in adults is in contrast to the study by Closa-Sebastià et al. [18] where younger animals displayed a higher seroprevalence of Salmonella spp.antibodies.However, these authors also found higher seroprevalence in adults for other eleven viral and bacterial pathogens investigated and they concluded that adult wild boars have a higher probability of being exposed to pathogens.
High prevalence in older females fi ts well with the behavior of female animals, which was observed in the hunting estates in the course of this study.Older females have a great protein demand during the lactation season, they often roam in searching for food and cover large area often expressing scavenging behavior.They feed on corpses of deer, boars, wild birds, rats, domestic animals which contributes to the transmission of pathogens [19].

Salmonella epidemiology
PFGE subtyping of Salmonella Typhimurium from wild boars (isolate 320) and domestic pigs (isolate 326), showed that these isolates were identical.A high degree of similarity was also found in S. Infantis from wild boars (isolate 318) and domestic pigs (isolate 269).Although Salmonella Enteritidis is characterized by low genetic diversity, some similarities between isolates were found.S. Enteritidis isolate 254 found in wild boars (estate "K") is highly related with the ones isolated from poultry farms (isolates 277 and 412).S. Enteritidis, S. Infantis and S. Typhimurium frequently found in domestic animals [20][21][22].
Hunting estates with the highest prevalence (33.3%) of Salmonella were "K" (open estate) and "D" (fenced estate).Sampling of twelve animals in estate "D" was performed during one hunting day and included eight female animals (among which six were older than 36 months), meaning that a half of the shot animals were from the high risk group.A relatively high number of hunted animals that belong to the risk group explains in part the high incidence of Salmonella in the hunting estate "D".
In open hunting areas on mountains, usually lacking a relevant domestic pig population in the area, wild boars act as a true host for Salmonella, as well as some other food borne pathogens (Mycobacterium, Leptospira, Erisipelotrix), maintaining an active infection focus and pathogen circulation [8,19].However, in lowland geographical areas (like in the current study), the situation is different.The identical and highly similar PFGE profi les found in wild boars and domestic pigs and poultry indicate an existence of a molecular and possible epidemiological link.Wild boars may easily come into contact with domestic animals due to the lowland nature of the Vojvodina terrain.In our studies [23,24], it was concluded that some zoonotic pathogens (Trichinella spiralis) on the territory of Vojvodina circulate from domestic pigs to wild boars and vice versa.

CONCLUSION
This study provides valuable data on the presence of Salmonella spp. in the wild boar population from the lowland regions (Vojvodina).The overall Salmonella prevalence in wild boars from hunting estates in Vojvodina region was 3.1%, with a dominating serotype being S. Enteritidis.Females weighing more than 75 kg (older than 36 months) were in the highest risk category for Salmonella carriage.
A very variable prevalence was determined between hunting estates and was particularly high in open hunting estates which directly relates to the proximity of human settlements.High molecular similarity found between Salmonella isolates from wild boars and domestic animals indicates a molecular and possible epidemiological link between these animals in lowland regions.

Table 1 .
List of isolates examined by PFGE

Table 2 .
Prevalence of Salmonella in each hunting estate (HE) * "K" and "L" are open hunting estates and all other are fenced areas