The gopher tortoise (Gopherus polyphemus) is a keystone species in the environment. It is listed as Threatened in southwest Alabama, Mississippi, and Louisiana by the US Fish and Wildlife Service (USFWS) and in Florida by the Florida Fish and Wildlife Conservation Commission (FWC), which recently uplisted it to Threatened status from a Species of Special Concern. Populations are also found in Georgia and the southeast corner of South Carolina.

Gopher tortoises are known for their enormous (4.5 m long × 2 m deep) burrows which also serve as a home to over 300 species of invertebrates along with about 80 vertebrate commensals, including frogs, juvenile gopher tortoises as well as other chelonian taxa, snakes, small mammals, and birds such as the Florida scrub jay and burrowing owl. Several of these commensals are legally protected and of high conservation value, such as indigo snakes and gopher frogs, adding to the intrinsic value of the gopher tortoise in the ecosystem. Tortoises live 40–60 y or more and dig many burrows within their home range and preferred habitat of longleaf pine forests.

According to FWC biologists, the number of tortoises has dropped by 60%–80% in the last century because of human-related habitat destruction (ABRPI 2008). In an effort to minimize the gopher tortoise's potential decline in Florida, in 1991 the FWC established incidental “take permits” for the state-listed reptile. This permit authorized the “taking (i.e., entombment or killing) of gopher tortoises incidental to development activities provided that adequate compensatory occupied tortoise habitat is preserved and managed in perpetuity.” Since then, more than 94,000 tortoises have been knowingly entombed; although, actual numbers are estimated at up to 900,000 lost from the population resulting from this state management practice (Cox 2007). In 2006, the FWC expanded the distance limits of the nonmandatory relocation sites to 100 miles north or south of the original property location to increase the number of available relocation sites. Still, this was not enough to encourage the conservation of the species, and they were uplisted to Threatened, leading to the July 2007 law mandating the relocation of all gopher tortoises to either on-site or off-site Relocation Parcels. Although take permits issued before or on 30 July 2007 would still be honored, public and private landowners were still encouraged to relocate all tortoises.

Because of their long life span and hardiness, these tortoises are ideal indicators of long-term environmental effects and disease prevalence monitoring (Gibbons 1988; Meyers-Schone and Walton 1994). These traits may also lead to dissemination of disease when relocation to naïve areas occurs. For example, there has been some concern about releasing wild turtles into the natural environment from rescue centers that have tested positive for Salmonella (Strohl et al. 2004). This is because Salmonella not only lives in the intestinal tract but can penetrate developing turtle eggs, leading to infection at hatching (Feeley and Treger 1969; Kaufmann et al. 1972). Salmonella can also be cultured from sites outside of the host (Bauwens et al. 2006; Hahn et al. 2007), and direct contact is probably not necessary for transmission (CDC 1990; Friedman et al. 1998; Winfield and Groisman 2003).

The prevalence of Salmonella in turtles and the significance of their role as natural carriers and potential reservoir hosts for the disease have public health, animal health, and ecological implications. Reptile-associated salmonellosis continues to be a human health problem throughout the world (Mermin et al. 2004; De Jong et al. 2005; CDC 2007). Every year in the United States, an average of 70,000 human cases of salmonellosis are attributed to contact with reptiles, which is about 6% of overall human salmonellosis cases (Wong et al. 2003).

Salmonella has been thoroughly studied in captive and pet reptiles, but data supporting the incidence of Salmonella in reptiles in the wild are scant (Wong et al. 2003). In wild birds, Salmonella infections are widespread, and persistent carriers are an important common source of infection to other animals, including humans (Petrak 1982; Scott 1988; Refsum et al. 2002; Tizard 2004). In reptiles, however, there is a continuing debate over the prevalence of wild turtles as reservoir hosts of Salmonella (Adesiyun et al. 1998; Richards et al. 2004).

With the cooperation of the privately owned Nokuse Plantation in Walton County, Florida, that provided ideal habitat for gopher tortoises, we were able to test for Salmonella in wild gopher tortoises relocated from St. Johns and Orange Counties, Florida. We used these data to help determine the prevalence of the bacteria in these relocated wild tortoise populations and the presence or absence of antibiotic resistance.

Methods

The gopher tortoises in this May and June 2007 study were salvaged from two separate relocation projects located in Orange (central) and St. Johns (northeastern) Counties Florida. The animals were excavated from their respective construction sites and brought to the Nokuse Plantation, a 48,000-acre private conservation area in Bruce, Florida, in the western panhandle, for relocation in natural longleaf pine habitat. The St. Johns County population was transported for five hours on the day of capture in 61 × 91 cm individual cages secured within an air-conditioned trailer. In contrast, the Orange County group was housed loose without water in a barn for at least one week before they were placed free to roam in the back of a covered pick-up truck for a six-hour drive. Upon arrival the tortoises were soaked in water for 2–3 h and then promptly tested.

The tortoises were sexed or recorded as juvenile if under 10 y (sex undetermined), weighed, and inspected for the presence or absence of overt clinical disease. A sterile six-inch Dacron-tipped plastic applicator (Solon Manufacturing Co, Solon, ME) was inserted into the cloaca, and fecal material was extracted. The swab was immediately placed in a chilled Selenite-F enrichment broth vial (VWR International, West Chester, PA), covered with Parafilm® (Pechiney Plastic Packaging, Chicago, IL), and placed in a cooler for a maximum of two h until refrigerated. The vials were shipped within 24 h to the Athens Diagnostic Laboratory at the University of Georgia College of Veterinary Medicine for testing. The Salmonella Genome Polymerase Chain Reaction (PCR) Test was performed on each individual vial. The PCR positive samples were then cultured for the organism on xylose lysine deoxycholate agar (XLD) and brilliant green with novobiocin (BGN) and then subcultured on blood agar. Salmonella grouping and antibiotic susceptibility disc diffusion tests run on Mueller Hinton agar (Remel, Lenexa, KS) to ampicillin, ceftazidime, chloramphenicol, amoxicillin/clavulanic acid, enrofloxacin, nitrofurantoin, orbifloxacin, tetracycline, ticarcillin, and trimethoprim/sulfamethoxazole were performed on all aerobically cultured Salmonella. The antibiotics chosen made up the antibiotic susceptibility reptile panel at Athens Diagnostic Laboratory.

The prevalence of Salmonella was calculated for both Orange and St. Johns County groups. A Chi-square test with p-value using Fisher's exact test was then performed to evaluate any significant difference in Salmonella prevalence between the two groups.

Results

A total of 60 G. polyphemus were tested (Table 1). The St. Johns County sample (n  =  27) included 13 males, 7 females, and 7 juveniles. The Orange County sample (n  =  33) included 12 males, 16 females, and 5 juveniles. The tortoises from Orange County were severely dehydrated upon arrival and, in combination with their deprivations before and during travel, were regarded as stressed. The external fluid therapy prior to sampling improved their condition significantly. All tortoises appeared clinically unremarkable with exceptions being gopher ticks, 3 animals with healed pulling (human interaction) injuries, one with grey feces, and one with serosanguinous transudate from the cloaca.

Table 1 Gopherus polyphemus from two Florida counties that were PCR tested for Salmonella.

View Fullscreen

From the 60 samples, 6 from clinically healthy tortoises were PCR positive for Salmonella. These included 2 females and 1 juvenile from Orange County and a female, male, and juvenile from St. Johns County. Two samples from each county produced bacterial colonies on aerobic culture after secondary enrichment. All samples were susceptible to the 10 antibiotics tested, and all were of group D1 Salmonella except one St. Johns County organism that was in group E2-F.

The prevalence of Salmonella was 10% overall in the gopher tortoises. It was 11% in St. Johns County animals and 9% in Orange County tortoises, which was not significantly different (Chi-square  =  0.07 with p-value  =  1.00).

Discussion

Gopher tortoises are long-living potential disease reservoirs for humans and other occupants of its habitat, as well as an ecologically important threatened species. This makes the gopher tortoise an ideal sentinel species for infectious diseases (Keymer 1978). Studies of Salmonella prevalence in this population are important for both wildlife and human health.

We found a 10% prevalence of Salmonella from gopher tortoises being relocated from Orange and St. Johns Counties to Walton County, Florida. These results are similar to studies that examined commercial freshwater turtles from the United States (Shane et al. 1990) and Belgium (Pasmans et al. 2002) as well as captive and household tortoises from Italy (Corrente et al. 2004) and wild gopher tortoises from Georgia (Lockhart et al. 2008). The prevalence seen here is not as high as in some other studies (Briones et al. 2004; Hahn et al. 2007; Hildago-Vila et al. 2007), and some factors have been proposed to explain the differences.

There has been debate over which method most reliably tests for the presence of Salmonella and that poor technique may be the cause of low prevalence results in some studies (Hahn et al. 2007). Classically, microbiologic culture has been used to detect the bacteria. The use of PCR with enrichment media has produced results that are more sensitive than the standard technique and highly specific for Salmonella detection (Mitchell and Shane 2001; Oliveira et al. 2003; Corrente et al. 2004; Mader 2006). In this study, a multiple technique protocol was used that included enrichment PCR and culture (Bauwens et al. 2006; Hahn et al. 2007).

Another potential reason for the variation in prevalence rates is the intrinsic nature of Salmonella spp. to be shed intermittently. If only one sample is taken, an underestimation of detection rates may result (Chiodini and Sundberg 1981; Cambre and McGuill 2000; Strohl et al. 2004). However, the results from studies using more than one sample show inconclusive evidence for the need to test multiple samples (Saelinger et al. 2006; Hildago-Vila et al. 2007). In this study, only one cloacal sample was obtained from each animal before the tortoises were released into their new environment.

Stress has been acknowledged to increase susceptibility to infection and increase shedding of the bacteria (Smith et al. 2002). Based on the two separate relocation efforts, our data suggest that gopher tortoises may shed the bacteria more or less equally under good or somewhat poorer environmental conditions.

The differences in prevalence of Salmonella-infected turtles have been documented throughout the world and in different natural environments. Geographical location and habitat variation may help explain differences in Salmonella prevalence, 0–53%, noted in various groups of wild turtle studies throughout the United States (Mitchell and McAvoy 1990; Harwood et al. 1999; Brenner et al. 2002; Richards et al. 2004; Saelinger et al. 2006; Hahn et al. 2007, Lockhart et al. 2008).

There has been a high level of antibiotic resistance found in Salmonella isolates from pet and captive reptiles. Prior to this study antibiotic susceptibility had not been examined in natural Salmonella isolates. There have been directed efforts to clear Salmonella from turtle populations, including eggs, through antibiotic treatment, but this primarily resulted in resistant strains of bacterial isolates (Mitchell and Shane 2001; Diaz et al. 2006). Without explicitly stating whether the reptiles had been exposed to previous treatment in other studies looking at captive reptilian antibiotic resistance, up to 48% of the bacteria isolated had multiple drug resistance (Seepersadsingh and Adesiyun 2003; Corrente et al. 2004). All Salmonella isolates tested in our study of two wild populations of tortoises were susceptible to 10 commonly used reptile antibiotics, including tetracycline, ampicillin, and sulfamethoxazole, which showed resistance in previous studies.

The gopher tortoise is a threatened, keystone species, and numerous efforts are underway to relocate and save this species from extirpation. Knowing that they can carry and shed Salmonella, especially during relocation efforts, make precautionary measures during these procedures very important. There may be a health and an environmental impact associated with relocating tortoises infected with Salmonella, which will in effect provide opportunity to spread to other species, including those that use gopher tortoise burrows, predatory species, and potentially humans. A quick screening method before relocating tortoises would be to test the soil, water, and possibly other burrow inhabitants for Salmonella spp. (Way et al. 1993; Pasmans et al. 2002; Hahn et al. 2007). With the uplisting of the gopher tortoise, and the law mandating relocation, special care should be taken when handling tortoises and selecting relocation sites.