Pathogen Surveillance and Detection of Ranavirus (Frog virus 3) in Translocated Gopher Tortoises (Gopherus polyphemus)

Gopher tortoises (Gopherus polyphemus) are the most widely relocated reptiles in the US (Seigel and Dodd 2000). They are federally protected in the western part of their range and a candidate for Federal conservation listing in the eastern portion (USFWS 1987, 2011). Gopher tortoises have experienced dramatic declines due to anthropogenic factors such as road mortality, illegal collection, and habitat loss and fragmentation, primarily from development (Enge et al. 2006). To reduce population-level impacts, they are often translocated from development sites (Florida Fish and Wildlife Conservation Commission 2012), although the translocation process carries the risk of inadvertent movement of pathogens (Kock et al. 2010).

Pathogens of historic and current conservation concern in gopher tortoises include Mycoplasma agassizii, Mycoplasma testudineum, and ranaviruses (including Frog virus 3–like ranavirus [FV3]). Infections due to ranavirus and mycoplasmas cause similar clinical signs (Diemer Berish et al. 2000; Jacobson et al. 2014; Marschang et al. 2016); diagnostic assays are needed for a definitive diagnosis. All three pathogens have been previously reported as being present during morbidity and mortality events in captive and wild chelonians (Diemer Berish et al. 2000; Jacobson et al. 2014; Duffus et al. 2015), yet reports of ranavirus infections in wild gopher tortoises are limited to two reports of tortoises with clinical disease (Westhouse et al. 1996; Johnson 2006). Although the epidemiology of ranaviruses is known in species such as the eastern box turtle (Terrapene carolina carolina; Adamovicz et al. 2018), in gopher tortoises it remains unknown.

Nokuse Plantation is a private preserve in Walton County, Florida, and the state's largest translocation recipient site (30°32′51.0″N, 86°00′59.0″W; Fig. 1). Prior to 2006, Nokuse had an estimated tortoise density of 0.004/ha because of past heavy collection and habitat conversion. From 2006 to 2019, over 5,000 gopher tortoises were translocated there from development sites. In response to a morbidity and mortality event in 2013–15 at Nokuse, we examined the prevalence of 12 pathogens including M. agassizii, M. testudineum, and FV3. We collected samples from affected tortoises, from other previously translocated tortoises (trapped), and from tortoises prior to release (prerelease).

During the event, several tortoises were found sick or dead in multiple enclosures across Nokuse. Tortoises found ill were transported to the Georgia Sea Turtle Center (Jekyll Island, Georgia, USA) for medical care, and were released at Nokuse after rehabilitation. As part of a complete mortality investigation including gross necropsy and histopathology, tissues were collected from tortoises suspected to have died of disease and clinically ill tortoises that died during rehabilitation. Gross necropsies and tissue collection were performed by one individual (T.M.N.) and stored at –80 C until analysis. Biological samples were collected from the following translocated tortoise groups from 2015 to 2017: rehabilitated, trapped, and prerelease. Some individuals were recaptured and thus sampled twice. Sampled tortoises came from 24 known counties.

Tortoises were identified by permanent marks, physically examined, weighed, and measured. Blood was collected from the brachial vein using a sodium heparinized syringe. Mucosal epithelial cells from the oral cavity and cloaca were collected with a single sterile cotton-tipped swab. Recaptured tortoises had only swabs collected. Ticks were collected if present. Time between release and pathogen surveillance varied between 0 and 5 yr posttranslocation.

We extracted DNA using Qiagen DNA Blood mini kit for tissues, blood, and swabs and Qiagen DNEasy kit for ticks (Valencia, California, USA) according to the manufacturer's protocol. Multiplex quantitative PCR was performed (Archer et al. 2017) to evaluate 12 pathogens (Anaplasma phagocytophilum, Borrelia burgdorferi, Ambystoma tigrinum virus, Bohle iridovirus, epizootic hematopoietic necrosis virus, FV3, M. agassizii, M. testudineum, Salmonella enteritidis, Salmonella typhimurium, Testudinid herpesvirus 2, and intranuclear coccidiosis). Samples positive with quantitative PCR were assayed using single-pathogen detection to verify presence and quantity (Archer et al. 2017). Samples were considered positive if all three replicates had a lower cycle threshold value than the lowest detected standard dilution (Allender et al. 2015). Ticks were screened only for FV3.

Samples tested included postmortem tissues (n=44), whole blood (n=78), and oral/cloacal swabs (n=143; Supplementary Material Table). Gastrointestinal tract (GIT; n=12), kidney (n=11), liver (n=11), spleen (n=5), and tongue (n=5) tissues from 13 dead tortoises were tested, selected for their probability of pathogen replication and quality of storage. Detection of FV3 occurred in 25% (3/12) of GIT samples from multiple years. Histopathology did not support an infectious process.

Blood samples were negative for all pathogens. Swabs were negative for all but three pathogens (Table 1). For individuals with positive pathogen detection, time between release and sampling varied between 0 and 2,914 d, with copy numbers being lowest for FV3 and highest for M. agassizii (Table 2). Samples were collected from 18 rehabilitated tortoises either at Georgia Sea Turtle Center or after rerelease at Nokuse, with two tortoises sampled twice. Detection of FV3 was negative, but 10% (2/20) and 40% (8/20) of swabs were positive for M. agassizii and M. testudineum respectively. Samples were collected from 68 trapped tortoises, with 20 tortoises sampled twice. Detection of FV3 was positive in 17% (15/88) of samples. Mycoplasma agassizii and M. testudineum were detected at 18% (16/88) and 10% (9/88), respectively. Swabs were collected from 35 prerelease tortoises in 2015 and 2017. Detection of FV3 was negative, but M. agassizii and M. testudineum were detected at 29% (10/35) and 3% (1/35) respectively.

Most tortoises (10/15) with FV3 DNA detected displayed clinical signs of disease: asymmetrical or eroded nares (4/15), pale mucous membranes (1/15), mild to moderate anemia (7/15), and lower body condition score (6/15). None presented with severe clinical signs. Five were asymptomatic. Seven of these tortoises were resampled in 2017, yet FV3 was not detected at that time.

Mycoplasma spp. are commonly detected in Florida gopher tortoises (Diemer Berish et al. 2010); thus, we expected to detect these pathogens. The prevalence of M. agassizii at Nokuse was 20%, similar to seroprevalence reported in other Florida populations (Diemer Berish et al. 2000). Although ranavirus has been linked to mortality in chelonians worldwide (Marschang et al. 2016), little is known about its epidemiology in gopher tortoises. To date, molecular diagnostics have been limited to individuals with clinical signs; therefore, current ranavirus prevalence across the state may be underestimated (Westhouse et al. 1996; Johnson 2006). Although it is unclear whether tortoises in this study were exposed pre- or posttranslocation, this report of ranavirus in translocated gopher tortoises includes affected individuals from multiple Florida counties, as well as asymptomatic tortoises with the potential to act as reservoirs. These detections underscore the need for detailed health assessments of both the individuals to be translocated and the recipient population.

Consistent with previous reports in chelonians, ranavirus was detected in oral–cloacal swabs and GIT tissue (Johnson et al. 2008). As swabs have shown good sensitivity and specificity for FV3 based on a challenge study (Allender et al. 2013), the lack of detection in blood suggests that tortoises were not viremic at time of sampling. This may suggest that the GIT may be an important infection site in chelonians, compared to the liver in amphibians, or that the virus can move through the GIT without an active infection, potentially suggesting that gopher tortoises may be less susceptible than other chelonians. Although it was detected in dead tortoises, low copy numbers and lack of an infectious process in histopathology suggest they did not die from FV3. Live tortoises with FV3 detected were negative upon resampling 1 yr later, suggesting that gopher tortoises may clear infections, that shedding is intermittent, or that detection was the result of the virus's transient passage during latency (Morales et al. 2010). This highlights the importance of repeated health screens for population monitoring as well as concurrent testing of other ectotherms in the habitat. In this study, ranavirus was detected from multiple years, suggesting that FV3 may remain in the environment or in another reservoir, although it was not detected in ticks, a potential vector (Johnson et al. 2007). Environmental (e.g., precipitation levels) and biotic factors likely contribute to prevalence differences among years, as ranavirus can persist for weeks in water (Nazir et al. 2012) and can be transmitted between taxa (Brenes et al. 2014). This may translate to increased infection potential for more sensitive species with overlapping distributions and potential close contact with gopher tortoises (e.g., box turtles, amphibians), although we saw no evidence of increased mortality of other species. Our results indicate that sustained surveillance of ranavirus of gopher tortoises, particularly before and during translocation, is needed.

This work was done under Florida Fish and Wildlife Conservation Commission scientific collection permit LSSC-15-00118 and the University of Georgia's Institutional Animal Care and Use Committee animal use protocol A2015 09-008-R1. Thank you to staff and AmeriCorps volunteers at the Georgia Sea Turtle Center for assistance with rehabilitated tortoises and sample collection.

Supplementary material for this article is online at http://dx.doi.org/10.7589/2019-02-053.