Low Occurrence of Salmonella spp. in Wild Animals in Bahia, Brazil—Population Assessment and Characterization in the Caatinga and Atlantic Forest Biomes

Simple Summary This study evaluated the possible role of wildlife in the Atlantic Forest and Caatinga biomes of Bahia, Brazil, as reservoirs of Salmonella. Very low frequencies (4/674 = 0.59%) of Salmonella infections and antibiotic resistance were observed. Thus, the findings of this study indicated that a wide variety of wildlife species do not carry Salmonella. This may be attributed to minimal human interference. Bacteria of potential public health concern were only detected in areas with high human interaction; therefore, we propose that Salmonella may be a good indicator of degradation in wildlife environments. Abstract Salmonella spp. are known to persist in the environment. Wild animals are believed to act as important reservoirs, with antimicrobial resistance frequently occurring in the environment. However, little is known about the role of the wildlife in Bahia as a reservoir for Salmonella in Brazil. This study aimed to isolate and characterize Salmonella spp. from wildlife in the Atlantic Forest and Caatinga biomes considering indicators such as the animal species, degree of anthropization, sampling area, and feeding habits. Convenience wildlife sampling and characterization were conducted, followed by microbiological and molecular identification of Salmonella isolates, serotyping, and antimicrobial susceptibility testing. A total of 674 fecal samples were collected from 12 municipalities during 2015–2021, and 4 were positive for the following Salmonella species: Salmonella enterica subspecies enterica serovar Agona (n = 1), Salmonella enterica subsp. enterica serogroup O:16 (n = 2), and Salmonella enterica subsp. enterica serovar Muenchen (n = 1). Antimicrobial susceptibility analysis revealed that one isolate was resistant to six antibiotics, including extended-spectrum penicillins and beta-lactamase inhibitors. These results indicated a low frequency of Salmonella spp. in the sampled forest fragments. The presence of Salmonella in wild animals increases the risk to public health and biodiversity and indicates that they can act as sentinels of environmental contamination or indicators of preservation.


Introduction
The current concept of health is extended to the idea of "One World, One Health" as it considers the tripartite relationship between humans, animals, and the environment [1].In this regard, wildlife is considered to play an important role as sentinels for the detection of infectious diseases related to the impact of anthropogenic actions and by taking an active role in transmitting pathogens to other animals and humans [2].Anthropogenic influences in forest areas occur due to demographic expansion related to economic activities, which generates waste that contaminates ecosystems and consequently alters the ecological relationships between species in different habitats [3].This combination of factors directly and indirectly promotes environmental and health changes in biomes and the agent/host/disease relationship [4].
Contact with wild fauna can increase the occurrence and/or incidence of pathogens both unidirectionally and bidirectionally [5], allowing organisms limited to certain areas and species to be transported to others [6].This relationship intensifies if commercial animal breeding sites are located close to forested areas, as this promotes the sharing of food, water, and shelter [7].Wild animals play an important role in the epidemiology of Salmonella spp.because they act as asymptomatic latent carriers of various serovars that can be isolated from a wide variety of wild animals [8] and survive in aquatic environments for long periods [2].Infectious pathogens of wildlife origin have recently gained attention and are considered important globally because of their role in livestock health and productivity [7].
Salmonella spp.can persist in the environment, facilitating infection of wild animals.Antimicrobial resistance frequently occurs in the environment, directly affecting populations of domestic and wild animals [9].The literature indicates that Salmonella enterica resistance has intensified because resistant organisms encode virulence and resistance genes [10] that are transported inter-and intraspecies via the horizontal transfer of integrons and/or plasmids [11].In addition, the zoonotic nature of Salmonella spp.has been observed in outbreaks in which strains from wild animals showed little genetic variance compared with serovars isolated from humans [12].Moreover, isolates from wild animals show phylogenetic similarities with those from domestic animals and food [13].
Thus, this study aimed to isolate and characterize Salmonella spp.from wildlife sampled from the Atlantic Forest and Caatinga biomes in Bahia, Brazil, considering indicators such as the animal species, degree of anthropization, sampling area, and feeding habits.This is essential because there is currently limited knowledge about the role of wildlife as a reservoir of this agent in the sampled regions.

Ethical Approval
The following components of this study were approved by the Ethics Committee

Study Area
Bahia is located in northeastern Brazil, covering a total area of 564,760,429 km 2 , with a population of approximately 14,136,417.This study was conducted in two biomes corre-sponding to 12 municipalities in the state of Bahia, Brazil: the Atlantic Forest (Belmonte, Ilhéus, Itambé, Itapetinga, Lajedo do Tabocal, Maracás, Mascote, Mata de São João, Morro do Chapéu, Una, and Uruçuca) and Caatinga (Brumado and Maracás) biomes (Figure 1).The areas were classified according to their biomes and vegetation types according to the Brazilian Institute of Geography and Statistics [14].A comprehensive assessment was conducted based on predefined indicators to evaluate the degree of anthropization in each study area.These included land-use intensity, vegetation cover, biodiversity, and the prevalence of human activities.These categories of land cover and land use were obtained using MapBiomas software v7.1 "https://mapbiomas.org/en(accessed on 28 August 2023)", from raster images of 30 m definition.Their percentages in each buffer were calculated in R software v4.1.1 using the "extract" function of the raster package.A buffer radius of 1 km was used around the collection sample points.The anthropization was considered as the sum of the percentages of the areas with pasture, other cultures, silviculture, and urbanization.

Study Area
Bahia is located in northeastern Brazil, covering a total area of 564,760,429 km 2 , a population of approximately 14,136,417.This study was conducted in two biomes responding to 12 municipalities in the state of Bahia, Brazil: the Atlantic Forest (Belm Ilhéus, Itambé, Itapetinga, Lajedo do Tabocal, Maracás, Mascote, Mata de São João, M do Chapéu, Una, and Uruçuca) and Caatinga (Brumado and Maracás) biomes (Figu The areas were classified according to their biomes and vegetation types according t Brazilian Institute of Geography and Statistics [14].A comprehensive assessment was ducted based on predefined indicators to evaluate the degree of anthropization in study area.These included land-use intensity, vegetation cover, biodiversity, and prevalence of human activities.These categories of land cover and land use were obta using MapBiomas software v7.1 "https://mapbiomas.org/en(accessed on 28 Au 2023)", from raster images of 30 m definition.Their percentages in each buffer wer culated in R software v4.1.1 using the "extract" function of the raster package.A b radius of 1 km was used around the collection sample points.The anthropization considered as the sum of the percentages of the areas with pasture, other cultures, culture, and urbanization.

Specimen Collection
The orders Didelphimorphia and Rodentia were captured with falling traps (bucket size of 60 L and arranged 10 m from each other), Sherman traps (31 × 8 × 9 cm), and Tomahawk traps (46 × 16 × 17 cm) arranged in transects and with bait comprising a mixture of sweet peanut, cornmeal, sardine, and banana.In 2015, the municipalities of Belmonte (June and July, n = 47), Mascote (August, n = 19), and Una (August, n = 15) were sampled for 10 consecutive nights, with a total of 13,020 traps/night.Sampling was conducted in the Itapetinga and Itambé municipalities for seven days each in August 2018 and February 2019, using 7742 traps/night.Similarly, in Lajedo do Tabocal and Maracás, sampling occurred for 10 days each in August 2019 and February 2020, using 6820 traps per night.
After collection for a previous research project, small mammals were euthanized by intramuscular injection of 100 mg/kg ketamine hydrochloride + 5 mg/kg xylazine hydrochloride (rodents) or 30 mg/kg ketamine hydrochloride + 2 mg/kg xylazine hy-drochloride (marsupials) in the gluteal region and deposited as witness material in the Alexandre Rodrigues Ferreira collection of mammals belonging to the State University of Santa Cruz (CMARF/UESC).The collected individuals were used for taxonomic and morphological characterization and demonstration of the sexual dimorphism of the species.However, pups and pregnant and lactating females were identified using numbered aluminum earrings and returned to the place of capture.
Samples were collected from golden-headed lion tamarins, Leontopithecus chrysomelas, in 2017 (March, May, June, and July) from Ilhéus (n = 12) and Una (n = 9).The collection method involved the use of Tomahawk traps measuring 48.3 cm long × 15.2 cm wide × 15.2 cm high, baited with bananas.These traps were deployed for approximately 11 h daily, accounting for 22 h per month at each location.After capture, L. chrysomelas were anesthetized with intramuscular ketamine 10 mg/kg and midazolam 0.3 mg/kg.
The Atlantic Forest endemic sloth, Bradypus torquatus, was sampled in 2021 (May and October) in Mata de São João (n = 11) using an active search.When the maned sloth was located, a trained climber accessed the treetops to catch the animals using their hands.During the physical restraint, the claws of B. torquatus were wrapped with Velcro tape to keep them closed.The animal was placed in a cotton bag and lowered to the forest floor using a rope.The animals were sedated with 4.0 mg/kg ketamine hydrochloride and 0.03 mg/kg medetomidine hydrochloride intramuscularly and 0.1 mg/kg atipamezole hydrochloride intravenously.
In the municipalities of Uruçuca, Ilhéus, and Una, wild birds were captured using mist nets, with each net active for approximately 10 h daily and inspected every 40 min.Ilhéus was sampled in 2016 (November) and 2017 (April and June) and Una in 2016 (November) and 2017 (March and August), with a sampling effort of 40,500 h•m 2 for five consecutive days.Sampling in May and June 2017 in Uruçuca resulted in a sampling effort of 32,400 h•m 2 for four consecutive days.
Manual collection of the Reptilia class was performed using an active search methodology in 2018 (September and October) in Brumado (n = 11), Ilhéus (n = 1), and Morro do Chapéu (n = 4) between 6 p.m. and 11 p.m. without standardization.
Fecal samples were collected using rectal swabs from 674 wild animals (Table S1) captured from forest remnants and adjacent pastures in the Atlantic Forest (n = 548) and Caatinga (n = 126).Wildlife genera and species were defined based on morphological and anatomical criteria.Upon collection, all swab samples were kept in 2 mL microtubes containing 1 mL of buffered peptone water (Liofilchem, Roseto degli Abruzzi, Italy) and promptly refrigerated to maintain their integrity during transportation to the laboratory for further processing.
Presumptive colonies of Salmonella spp.were grown in trypticase soy broth (HiMedia Laboratories Pvt.Ltd.) at 37 • C for 18-24 h before submitting to DNA extraction using a DNA PureLink™ Genomic DNA Mini Kit (Invitrogen, Carlsbad, CA, USA), according to manufacturer instructions, and were confirmed using polymerase chain reaction (PCR), as described below.The total PCR reaction was 25 µL and consisted of 1× PCR buffer (Invitrogen, Carlsbad, CA, USA); 1.25 mM MgCl 2 (Invitrogen); 200 µM dNTP; 10 pmol of each primer, ST11 (5 -AGCCAAACCATTGCTAAATTGGCGCA-3 ) and ST15 (5 TTTGCGACTATCAGGTTACCGTGG-3 ), which are specific for the Salmonella genus [16]; 1.25 U of Taq DNA polymerase (Invitrogen); and 2 µL of DNA (50 ng/µL) from the suspected colony of the pathogen.Sterile nuclease-free water was used to complete the reaction volume to 25 µL.S. enterica serovar Enteritidis PT1 was used as a positive control, and a DNA-free reaction was used as a negative control.The amplification occurred with an initial denaturation (5 min at 94 • C) followed by an amplification phase of 35 cycles (30 s at 94 • C, 30 s at 62 • C, and 1 min at 72 • C) and a final extension step (7 min at 72 • C) in a Proflex PCR system (Applied Biosystems; Life Technologies, Carlsbad, CA USA).PCR products were visualized on a 1% agarose gel stained with SYBR Green (Invitrogen) and examined under UV light [17].The PCR-positive samples were further serotyped using serogroup-and serovar-specific antisera obtained from the Enterobacteriaceae Laboratory of the Oswaldo Cruz Foundation in Rio de Janeiro, Brazil.

Statistical Analysis
Due to the limited number of positive samples, only descriptive statistical analyses were performed [20].In addition, the patterns of wild animals (diet, locomotion, species, biomes, and municipalities) were analyzed using descriptive and inferential statistics [20].

Results
Between 2015 and 2021, 674 samples were collected from wild animals in the Atlantic Forest and Caatinga.These samples were obtained from the 12 municipalities studied and were collected from the following animals: 114 from birds, 544 from mammals (317 rodents, 195 marsupials, 21 primates, and 11 sloths), and 16 from reptiles (Table S1).Different study areas showed varying land use and, consequently, different percetanges of anthropization.The forest environments studied showed lower percentages of anthropization (except for Itapetinga and Mascote) when compared with Caatinga (except for Brumado).It was not possible to observe if areas that had anthropization such as the cocoa plantations and agroforestry systems had any influence on biodiversity.Four Salmonella isolates (0.59%) were obtained from three mammals and one bird.These isolates belonged to three distinct serovars: S. enterica subsp.Enterica serovar Agona, Salmonella enterica subsp.enterica O:16, and Salmonella enterica subsp.enterica serovar Muenchen.All positive samples were found in areas of the Atlantic Forest (Table 1) and had percentages of anthropization of 51.8%, 47.0%, 79.2%, and 70.3% (Table S2).All negative Salmonella samples varied in their anthropization percentages (Table 2).The antimicrobial sensitivity profile of Salmonella enterica subsp.enterica serogroup O:16 isolated from a young female Didelphis albiventris showed resistance to 6 of the 15 antimicrobials tested (amoxicillin/clavulanic acid, ampicillin with sulbactam, cefoxitin, chloramphenicol, trimethoprim, and sulfamethoxazole).The other serovar isolated from an adult female did not exhibit antimicrobial resistance.S. Agona isolated from a female passerine showed intermediate resistance to amikacin and gentamicin.S. Muenchen from a male Cerradomys vivoi demonstrated intermediate resistance to amikacin.

Discussion
The results observed in this study may not be directly comparable with those of other studies because the actual prevalence of Salmonella spp.depends on the sampling location and methods applied [21].Thus, the low prevalence (0.59% in the 6-year convenience sampling) of Salmonella spp.may be attributed to cloaca/rectal swab sampling.However, in a previous Danish study on carcasses, the authors observed a low prevalence of Salmonella spp. in wild birds [22] and boars [9].Similarly, a low frequency of isolation has been reported in wild birds from a Swiss rehabilitation center (1.08%) [22] and in wild animals in northwestern Italy (4.30%), including canids, mustelids, wild birds, and ungulates [23].In addition, another study from western Italy reported a frequency of 2.87% in wild animals, such as magpies, foxes, crows, jays, brown hares, hedgehogs, starlings, swans, wild ducks, Porcupine's green woodpeckers, and Eurasian collared doves [7].No isolates of Salmonella spp.were obtained in a study on wild rodents (n = 237) in Poland [24].These results reinforce the fact that the prevalence of Salmonella spp. is usually lower in wild than in domesticated animals [25].
A low frequency of positive results (9.5%) was observed in the wild boar population in São Paulo [9].These results were attributed to the fact that as the number of human settlements in close vicinity increases, wild animals come into contact with other animal species, raw food, and high temperatures, all of which favor the occurrence of Salmonella spp.; therefore, they can serve as vectors for pathogen transmission [26].This behavior was observed in a previous study [27], where rodents from urban areas were more prone to carry pathogens.Therefore, as stated by Skov [21], the prevalence of Salmonella spp. in wildlife can vary greatly among studies, and this variation is not associated with sampling methods, study designs, or differences in prevalence between countries.
Although the areas sampled were exposed to human contact, the low occurrence of Salmonella spp. in this study may be attributed to the fact that the areas were highly conserved and animals from preserved areas showed a lower risk of contamination than those with human contact [26].In addition, the animals presented a mostly terrestrial and frugivorous nature, which may contribute to their occurrence because the behavior and feeding habits of wild animals are known to influence the likelihood of infection with Salmonella [27].Badgers, red foxes, raptors, and wild boars acquire Salmonella infections by scavenging contaminated carcasses or human leftovers [8].However, wild animals can be latent carriers and thus contribute to the low frequency of isolation because, during some periods, the microorganisms are located intracellularly and are not released into the environment [28].
All positive samples were obtained from the Atlantic Forest.This result agrees with that of another study on 36 wild mammals from the Atlantic Forest and Amazonia biomes in Brazil, which obtained only one isolate of Salmonella enterica subspecies diarizonae serotype 60:r:e,n,z15, which was isolated from a rodent in Amazonia [2].In another study of 207 wild rodents from farmland and nonfarmland areas in Sweden, only one Salmonella enterica serovar Typhimurium isolate from a subpopulation of 11 individuals was obtained from a chicken farm [29].In our study, the samples positive for Salmonella spp.from small mammals were from areas close to a cattle farm, which is a factor alongside the changes made to the forest area as it favors the habit of looking for shelter and food in peridomiciliary areas, thus facilitating the acquisition and dispersion of pathogenic microorganisms [30].Among the marsupials sampled in this study, two Didelphis albiventris tested positive for Salmonella enterica subsp.enterica O:16.Similar results were found when Didelphis albiventris in a rural area in Argentina (five poultry farms) and eight surrounding forest areas were sampled, and a 4% positivity rate for Salmonella enterica was found.However, the authors found no relationship between age and sex [31].Additionally, Casagrande [32] isolated Salmonella from 17.0% (18/106) of Didelphis aurita and 17.5% (7/40) of Didelphis albiventris individuals in their study, most of which were wild animals that lived in the forest around a human population.These authors also found no influence of sex or age on the frequency of Salmonella infections.
Resistance to environmental bacteria is a natural biological process, and its capacity for transmission and dissemination is enhanced by multiple drug resistance genes shed in the feces of livestock and humans [33].However, fragmentation promotes environmental stress in ecosystems, allowing the selective pressure of multiresistant agents on antibiotics [34].With interactions between humans, domestic animals, wild animals, and the environment, contact has intensified, increasing the chances of transmission of resistant strains [35].Therefore, free-ranging wild animals that have not been exposed to antibiotics may have high rates of drug resistance owing to environmental contamination [36].This resistance can also be acquired from consuming contaminated water or food and transported via direct contact with garbage and sewage residues [37].
The detection and analysis of microorganisms, mainly bacteria, can provide information on ecosystem exposure to biological agents [38].Based on our findings, we propose that the detection of pathogenic Salmonella spp.isolates be used for biomonitoring environmental quality in future wildlife studies.This approach facilitates the assessment of conservation and management practices.

Conclusions
In the present study, wild animals assessed in the mesoregions showed a low frequency (4/674 = 0.59%) of Salmonella, which is consistent with previous reports on wild animal populations worldwide.In addition, the isolates were found to be multiresistant, including resistance to extended-spectrum penicillins with beta-lactamase inhibitors.Thus, it is important to consider that specific behavior, diet, and the low percentage of anthropization may be the main factors affecting the studied species in the Atlantic Forest and Caatinga in Bahia, Brazil.Therefore, we propose that the detection of Salmonella spp.can be used for anthropization biomonitoring and as an indicator of environmental preservation in future wildlife studies.

Figure 1 .
Figure 1.Biome of areas where the wildlife was captured and sampled for Salmonella screening 2015 to 2021 in Bahia State, Brazil.

Figure 1 .
Figure 1.Biome of areas where the wildlife was captured and sampled for Salmonella screening from 2015 to 2021 in Bahia State, Brazil.
on Animal Use of the State University of Santa Cruz (CEUA-UESC) and the Oswaldo ; date of approval: March 2021); and the use of traps to capture wild reptiles in 2018 (CEUA-UESC protocol code: 034/16; date of approval: September 2016).

Table 1 .
Characterization of the captured and sampled wildlife positive for Salmonella spp.according to their respective location and environment.

Table 2 .
Characterization of the captured and sampled wildlife negative for Salmonella spp.according to their respective location, biome, and anthropization percentages by wildlife family and by municipality.

Table 3 .
Frequencies of captured wildlife species at each sample site in Bahia State, Brazil.