ABSTRACT
Feral swine (Sus scrofa), an important prey species for the endangered Florida panther (Puma concolor coryi), is the natural host for pseudorabies virus (PRV). Prior to this study, PRV had been detected in just three panthers. To determine the effect of PRV on the panther population, we prospectively necropsied 199 panthers and retrospectively reviewed necropsy and laboratory findings, reexamined histology, and tested archived tissues using real-time PCR from 46 undiagnosed panther mortalities. Seven additional infections (two prospective, five retrospective) were detected for a total of 10 confirmed panther mortalities due to PRV. To further evaluate the effect of PRV, we categorized radio-collared (n=168) and uncollared panther mortalities (n=367) sampled from 1981 to 2018 based on the likelihood of PRV infection as confirmed, probable, suspected, possible, or unlikely/negative. Of 168 radio-collared panthers necropsied, PRV was the cause of death for between eight (confirmed; 4.8%) and 32 (combined confirmed, probable, suspected, and possible categories; 19.0%) panthers. The number of radio-collared panther mortalities due to PRV was estimated to be 15 (95% empirical limits: 12–19), representing 8.9% (confidence interval: 4.6–13.2%) of mortalities. Gross necropsy findings in 10 confirmed cases were nonspecific. Microscopic changes included slight to mild perivascular cuffing and gliosis (primarily in the brain stem), lymphoplasmacytic meningoencephalitis (cerebral cortex), and intranuclear inclusion bodies (adrenal medulla). The PRV glycoprotein C gene sequences from three positive panthers grouped with the sequence from a Florida feral swine. Our findings indicate that PRV may be an important and underdiagnosed cause of death in Florida panthers.
INTRODUCTION
The Florida panther (currently listed as an endangered subspecies, Puma concolor coryi) was part of a contiguous puma (also known as cougar or mountain lion) population across North America (Currier 1983; Onorato et al. 2010). The arrival of Europeans led to habitat loss and fragmentation, persecution, and depletion of prey; by the 1980s, only 20–30 panthers remained in Florida (McBride et al. 2008).
Europeans introduced swine (Sus scrofa) to Florida (Towne and Wentworth 1950), and feral swine can be found in every county in Florida and much of southern North America (Lewis et al. 2019). Swine are the definitive hosts of suid alphaherpesvirus 1, also known as pseudorabies virus (PRV), which causes pseudorabies (Aujeszky's disease) in swine and other mammals. This infection is endemic in feral swine in Florida, with reported seroprevalences of 35–79% (van der Leek et al. 1993; Carr et al. 2019). In domestic swine, clinical signs can include embryo resorption, abortion, or stillbirths in pregnant sows; fever and respiratory signs in juveniles and adults; and neurologic disease and death in neonates (Murphy et al. 1999). Clinical signs have not been reported in naturally infected feral swine in North America (Stalknecht and Howerth 2001); experimental infections in feral swine were largely subclinical (Hahn et al. 1997).
Infections with PRV in carnivores, ruminants, and rodents are typically fatal (Stalknecht and Howerth 2001). Reports of natural PRV infections in wildlife are rare and, among carnivores, include foxes (Alopex lagopus and Vulpes vulpes), wolves (Canis lupus), and raccoons (Procyon lotor; Sehl and Teifke 2020); reports in wild felids are limited to an Iberian lynx (Lynx pardinus; Masot et al. 2016) and a Florida panther (Glass et al. 1994). Glass et al. (1994) concluded that the effect of PRV on the panther population was unknown but could affect recovery. Only two additional panthers were diagnosed with PRV through 2012. One was an adult male that died acutely in 2010 of unknown causes. Encephalitis was not initially evident microscopically; attempts to isolate PRV from brain tissue were unsuccessful. Intranuclear inclusion bodies suggestive of herpesvirus infection (Jubb and Huxtable 1993; Sehl and Teifke 2020) were seen in the adrenal medulla, and subsequent virus isolation yielded PRV. These findings prompted our study, with the objectives being to describe the epizootiology of PRV in the panther population and determine the importance of PRV infection as a cause of death in Florida panthers.
MATERIALS AND METHODS
Samples and data (capture, necropsy, diagnostics, and telemetry) used in this study were collected from radio-collared and uncollared Florida panthers (denoted as FP and UCFP, respectively) and adult radio-collared Texas pumas (Puma concolor stanleyana) from 1981 through 2018 (Florida Fish and Wildlife Conservation Commission [FWC] 2019) in south Florida (Fig. 1). Eight female radio-collared Texas pumas were released into south Florida in 1995 to artificially restore historic gene flow within the once-contiguous North American population (Seal 1994), with three surviving Texas pumas being removed to captivity in 2002–03 (Onorato et al. 2010). Necropsy and PRV test results for five Texas pumas were included in this study.
The Florida panther population increased during the study. From 1981 to 1994, the minimum population was static at 20–30 individuals (McBride et al. 2008). The population grew steadily after genetic restoration, and by December 2015, it numbered 120–230 (McBride and McBride 2015; FWC 2019). Our study evaluated panther mortalities both retrospectively (1981–2012) and prospectively (2013–18).
Capture and necropsy
The FWC and the National Park Service captured panthers during 1981–2018 in southern Florida (south of 28°N) using trained hounds as described previously (McCown et al. 1990; McBride 2007). Panthers were immobilized using injectable anesthetics administered remotely and, once anesthetized, were examined, vaccinated, and sampled as described previously (Cunningham et al. 2008). See Supplementary Materials and Methods for a description of panther capture, radio-collaring, and handling. Uncollared panther carcasses were collected opportunistically (reported or otherwise found without the assistance of a radio-signal). Carcasses examined (168 radio-collared, 367 uncollared) ranged from fresh and unfrozen to skeletal remains. Necropsies and histologic examinations were performed as described previously (Cunningham et al. 2008). Sixty-five collared and uncollared panthers were evaluated for PRV using various testing modalities based on the availability and condition of tissues (see Supplementary Tables S1–S7). We conducted additional testing on archived samples from 46 (33 radio-collared and 13 uncollared) individuals with undiagnosed mortalities occurring 1981–2012. Of these, nine were reexamined histologically (we reexamined slides, made additional slides, and made slides from unexamined archived fixed tissues). The feral swine used for PRV genetic comparisons was a subadult female from Okaloosa County (Florida) captured and sampled as described previously (Hernández et al. 2018).
Diagnostics
Virus isolation: Virus isolation was performed using brain or adrenal gland tissue at the Southeastern Cooperative Wildlife Disease Study (Athens, Georgia, USA) and was attempted on samples from all individuals that died of unknown causes if suitable samples were available (12 Florida panthers, two Texas pumas; see Supplementary Table S2 and Supplementary Materials and Methods). Virus isolation results from the panther previously reported (Glass et al. 1994) were also included in our report.
Real-time PCR, genetic characterization, and phylogenetic analysis: Real-time PCR to detect PRV DNA in panthers/pumas was performed at the Michigan State University Veterinary Diagnostic Laboratory (Lansing, Michigan, USA) on fresh frozen and formalin-fixed, paraffin-embedded tissues as described in the Supplementary Materials and Methods. Tissues from selected panthers and a Florida feral swine were also tested by real-time PCR at the University of Florida Department of Wildlife Ecology and Conservation (Gainesville, Florida, USA; Supplementary Table S5) as described previously (Sayler et al. 2017). See the Supplementary Materials and Methods for a description of methods used for genetic characterization and phylogenetic analysis of PRV sequences detected in Florida panthers and a Florida feral swine.
Other diagnostics: Methodologies for immunohistochemistry, serology, and other diagnostics are described in the Supplementary Materials and Methods.
Statistical analysis
The methodology for testing the effects of covariates on PRV is described in full in the Supplementary Materials and Methods.
Estimated PRV proportional mortality: Mortalities of radio-collared Florida panther and Texas puma (n=168) and uncollared pumas and panthers (n=367) from 1981 to 2018 were assigned to one of five PRV categories based on various criteria. Case definition and estimated probability of PRV infection for each category are depicted in Table 1. These probabilities were subjective and assigned based on collection time and location, our ability to eliminate other causes of death, and the panther's association with other known positive cases (for example, two probable cases [UCFP55, 56] were offspring of a confirmed case [UCFP57] and died in the same location within days of each other).
We used SAS version 9.4 (Cary, North Carolina, USA) to perform probabilistic modeling to estimate the expected number of collared panthers infected with PRV. We assigned a fate (0 for not infected; 1 for infected) for each of the tested panthers drawn from a Bernoulli distribution with the relevant probability. We performed 50,000 random draws from the Bernoulli distribution and summed the total number of “infected” panthers for each draw. We then used the median value of the 50,000 random draws as our estimate of the number of infected panthers and the 2.5th and 97.5th percentiles as 95% empirical limits.
RESULTS
Numbers of panthers tested for PRV based on radio-collar status (collared and uncollared) and study period (retrospective 1981–2012 and prospective 2013–18) are depicted in Table 2. Other diagnostic data are provided in Supplementary Tables S1–S7.
PRV diagnostic testing in Florida panthers: Results for panthers testing positive for PRV testing are summarized in Table 3. We isolated PRV from two panthers (FP117, FP173) in this study; PRV had been isolated from FP29 previously (Glass et al. 1994). Tissues testing positive by real-time PCR for PRV DNA variously included brain stem, cerebrum, adrenal gland, salivary gland, kidney, liver, and mesenteric lymph node. Immunohistochemistry-positive tissues included brain stem, adrenal gland, spinal cord, and cranial nerves 9 and 10 (salivary gland and other organs were not tested; Table 3). Testing of frozen or fixed tissues archived >10 yr from two virus isolation–positive cases (FP29, FP117) using PCR was negative (Table 3).
Phylogenetic and genetic analyses: Glycoprotein C sequence data could not be obtained from all PRV-positive tissues, presumably due to degradation of viral DNA and the poor quality of sequenced amplicons. Clean bidirectional sequences were obtained from three panthers: FP117, FP156, and FP173 (GenBank accession nos. MW447841–MW447843) and a feral swine (GenBank accession no. MW447904), which shared a three-nucleotide off-frame deletion, similar to reference strains Auj/Dog/Fr82/060-1 (GenBank no. MN590225), dog/Italy/15608/2016 (GenBank no. MF040159), and Indiana S (GenBank no. D49436), shortening the threonine repeat from three to two. The maximum-likelihood analysis revealed that PRV from panthers and feral swine formed a well-supported clade (Fig. 2). A PRV isolate from a domestic pig in Indiana was the closest relative to the Florida PRV strains, with a two-nucleotide difference resulting in a substitution from proline to glycine at residue 80. The PRV sequences obtained from panthers in this study and the feral swine differed by 0–2 nucleotides (Fig. 3).
Serology: All panthers tested were negative for PRV antibodies (serum neutralization, n=46; latex agglutination, n=22; data not shown). Negative results included five confirmed PRV mortalities sampled between 1 mo and 3.5 yr before death. Blood collected postmortem from PRV-infected panthers was unsuitable for serology due to hemolysis.
Telemetry: All deaths of radio-collared panthers due to PRV were considered acute. Movements based on telemetry data were apparently normal 2–4 d prior to death and then became progressively restricted (Fig. 4). Carcasses were often found in open areas.
Gross necropsy: Panthers with confirmed PRV infection were in good to excellent body condition. Gross findings were nonspecific and included an infarct in the stomach (FP173), endocardial hemorrhages (FP173), and ptyalism (as evidenced by facial fur matted with saliva, FP118, FP173). Of the 10 confirmed PRV cases, eight had intact gastrointestinal tracts available for examination; three (FP29, FP117, UCFP57; 37.5%) contained feral swine hair, and the remaining five were empty. Self-mutilation was not observed.
Histology: Postmortem autolysis prevented meaningful histologic examination in all but three (FP117, FP118, FP173) of the 10 confirmed cases. The predominant microscopic changes were slight to mild mononuclear perivascular cuffing and gliosis primarily in the brain stem. Mild perivascular cuffing, gliosis, and lymphoplasmacytic meningoencephalitis were seen in the cerebral cortex of two panthers (FP117, FP173). Mild perivascular cuffing and clusters of lymphocytes were seen in the spine in FP117, and intranuclear inclusion bodies were observed in the adrenal medulla (FP173). Hemorrhage in the heart (FP173) was most severe in the subendocardial myocardium of the interventricular septum in the left ventricle, but it was also present within the myocardial interstitium and epicardium of the left ventricle. Fragmentation and hypereosinophilia of subendocardial cardiomyocytes consistent with degeneration and necrosis were rare and not present in other areas of the heart. Pulmonary congestion and edema were seen in all three panthers.
Estimated PRV proportional mortality: Pseudorabies virus infection was the confirmed cause of death in 8/168 (4.8%) radio-collared panthers necropsied during the combined periods. One radio-collared panther was categorized as probable PRV infection, seven were categorized as suspected, and 16 were categorized as possible (Table 2). Combining the categories (confirmed, probable, suspected, and possible) gave a maximum of 32 (19.0%) mortalities possibly due to PRV. The number of mortalities attributable to PRV was estimated at 15 (95% empirical limits: 12–19) radio-collared panthers, representing 8.9% (95% confidence interval: 4.6–13.2%) of mortalities. Location, age, sex, and individual heterozygosity were not associated with PRV infection (P>0.05). Although genotype was not included as a covariate, all confirmed PRV infections had some degree of genetic admixture; no cases were diagnosed in canonical (i.e., original, prior to genetic restoration) Florida panthers. Confirmed PRV occurred in panthers with and without existing feline immunodeficiency virus (FIV) infection. The geographic distribution of confirmed, probable, and suspected cases is depicted in Figure 1, and the number of cases is depicted by year in Figure 5.
DISCUSSION
We demonstrated that PRV is having a larger effect on the Florida panther population than previously documented. An estimated 9% of radio-collared panther mortalities were due to PRV, making it the third leading diagnosed cause of death, behind intraspecific aggression and vehicular collision (data not shown).
Previously, PRV had been diagnosed as the cause of death in only three (2.1%) radio-collared panther mortalities (Glass et al. 1994; FWC 2013). This apparent underdiagnosis was probably due to the inherent difficulty in identifying PRV infection in free-ranging felids (Roelke et al. 1993) combined with failure to collect and test the correct tissues, and poor sample quality. Infection with PRV in felids is often focal and unilateral (Dow and McFerran 1963), and the diagnosis may be missed if the sample does not contain adequate virus. Employing more sensitive diagnostics, such as real-time PCR, for severely autolyzed or fixed archived tissue samples facilitated retrospective diagnosis. In two confirmed PRV cases, however, PCR failed to detect PRV in frozen tissues from panthers confirmed by virus isolation to be positive >10 yr earlier. This low sensitivity in positive controls underscores the difficulty in diagnosing PRV from archived tissues. Therefore, a negative real-time PCR alone does not necessarily rule out infection. Among the radio-collared mortalities, the cause of death in 31 (18.4%) panthers remains undiagnosed. Assessing the true impact of PRV requires timely collection of carcasses and real-time or quantitative PCR testing of brain stem, especially the medulla oblongata just rostral to the obex (Hagemoser et al. 1980), cerebellum, spine, adrenal gland, tonsil, and salivary gland.
Feral swine carry PRV and are an important prey species. Feral swine consumption by panthers is credited with better physical condition, smaller female home ranges, and greater reproductive output (Maehr et al. 1990). Although not statistically significant, the distribution of PRV-infected panthers coincided with feral hog populations both geographically and temporally. Most panther PRV cases were detected north of I-75 (Fig. 1), where greater consumption of feral swine has been documented (Caudill et al. 2019). Confirmed cases were especially concentrated in and around game enclosures that contained high densities of feral swine. South of I-75, predation on feral swine is less common, and in Everglades National Park, feral swine represented an insignificant component of panther diet (Dalrymple and Bass 1996). Feral swine comprised a larger proportion (23%) of panther diet in southern Big Cypress National Preserve; this was still small compared to swine comprising an estimated 59% of the biomass consumed in the north (Maehr et al. 1990). Two PRV infections in panthers were documented south of I-75. One (FP156) occurred along the urban-wildland interface, where ephemeral populations of feral swine can be found. The other (FP108) occurred in 2002 before a decline in feral swine numbers (based on hunter harvest) in Big Cypress National Preserve beginning in 2003 (FWC 2017). Fewer additions to the population through release or escape, poor habitat quality, higher water levels (Leighty et al. 1954; Maehr et al. 1989), and an expanding predator population may have caused this decline.
The seroprevalence of PRV in feral swine in Florida varies but can reach 79% (Carr et al. 2019). Gresham et al. (2002) observed that PRV seroprevalence can increase over time in an isolated, nonhunted population. The prevalence of PRV infection can be elevated in fenced enclosures (Steinrigl et al. 2012), and PRV prevalence is higher among older swine (Pedersen et al. 2013).
In feral swine, most PRV infections are latent and unlikely to be transmitted to swine or secondary hosts (Hahn et al. 1997). Stressors such as concomitant disease, farrowing, and breeding activity can reactivate infection, leading to viral shedding and subsequent risk of transmission (Pomeranz et al. 2005). The frequency of PRV shedding in feral swine in Florida is approximately 7% (Hernández et al. 2018). Sporadic outbreaks of PRV among feral swine–hunting dogs in Florida may suggest periods of increased viral shedding among feral swine and subsequent exposure to secondary hosts (Cramer et al. 2011). Conversely, hunting with dogs also may increase PRV prevalence in feral swine. In Florida, PRV prevalence was approximately 20% greater where feral swine were hunted with dogs compared to areas where they were hunted without dogs (Carr et al. 2019).
Transmission to panthers probably occurs by ingestion and is supported by the apparent point source cause of death in what appeared to be a family group of three panthers. An adult female (UCFP57), confirmed to be infected with PRV, and two 18-mo-old offspring (UCFP55 and UCFP56) died within 5–6 d and <500 m of each other. Feral swine hair was found in UCFP57's stomach (the remainder of the gastrointestinal tract and that of the subadults had been scavenged). The subadults were too decomposed to be tested; no other cause of death was apparent, and they were classified as probable cases. These deaths occurred in an enclosure with a high density of feral swine.
Phylogenetic and genetic analyses of the glycoprotein C (UL44) gene of PRV from infected Florida panthers and a Florida feral swine support the idea that feral swine are the probable source of the virus in panthers (Fig. 2). Sequences from both species clustered together, indicating genetic relatedness, but sampling bias and a dearth of available sequencing data from feral swine PRV isolates preclude conclusions about species-specific genetic patterns or temporal relatedness. Full-length PRV genome sequencing and additional feral swine isolates would help to elucidate this relationship.
The incubation period in secondary hosts depends on route of exposure and amount and strain of virus (Hagemoser et al. 1980; Kirkpatrick et al. 1980; Platt et al. 1983). In domestic cats (Felis catus), depression, anorexia, and ptyalism began 2–4 d postinfection and progressed to restlessness, vocalizations, vomiting, coma, and death within 24–48 h (Dow and McFerran 1963; Horvath and Papp 1967; Hagemoser 1979). Intense pruritus and associated self-mutilation, as seen in infected dogs and ruminants (Dow and McFerran 1963; Murphy et al. 1999), occurs inconsistently in infected cats (Dow and McFerran 1963; Horvath and Papp 1967; Sabo et al. 1968; Hagemoser 1979). The incubation period in panthers is unknown but may resemble that seen in domestic cats. Acute onset of disease is inferred from telemetry data, where movement patterns of infected FP appeared normal until 2–4 d prior to death. Thereafter, movements became reduced, suggesting acute onset of clinical signs, and death occurred within 48 h. Observation of clinical signs was not possible in free-ranging panthers, and the presence of pruritus is unknown, although at necropsy, no signs of self-mutilation were seen. Ptyalism, however, was seen in two cases.
Gross and microscopic pathology also indicated acute death. Gross pathology was nonspecific, although endocardial hemorrhages like those observed in infected cats (Sehl and Teifke 2020) were seen in FP173. Microscopic changes were minimal. Evidence of inflammation in the central nervous system, albeit mild, was present in all PRV-infected panthers for which histology was available (in some cases, inflammation was detected upon reexamination of retrospective cases). Lesion distribution and severity resembled those seen in experimentally infected domestic cats and included mild mononuclear perivascular cuffing and microgliosis primarily in the medulla oblongata (Hagemoser et al. 1980). Unlike domestic cats, however, more pronounced, yet still mild, lesions were seen in the cerebrum in two panthers. In infected domestic cats, intranuclear inclusion bodies were observed occasionally in the brain stem (Dow and McFerran 1963; Sabo et al. 1968; Hara et al. 1991). Intranuclear inclusion bodies were seen in the adrenal medulla of one panther but were not observed in other tissues, including the brain or brain stem, of any infected panthers. Cardiac hemorrhage in FP173 was most severe in the subendocardial tissue, where it was accompanied by minimal myocardial necrosis. This differed slightly from cardiac hemorrhages described in dogs (Canis familiaris) with PRV infection, which is more widely distributed (Sehl and Teifke 2020), and in farmed American mink (Mustela vison), where hemorrhages were described only as epicardial (Liu et al. 2017). It was not clear if hemorrhages in FP173 were directly related to PRV infection, or if they represented a secondary lesion. Comparison of microscopic lesions among positive cases was difficult because tissues were not necessarily collected from the same location and organs of each animal. Further, because microscopic lesions in PRV-infected panthers were asymmetrically distributed and mild, absence of observed pathology did not rule out PRV infection.
In experimental studies, PRV infection in domestic cats is invariably fatal (Sabo et al. 1968; Maes and Pensaert 1987); however, the documentation of PRV antibodies in domestic cats (Weigel et al. 2000) and other species suggests that some individuals may survive exposure (Platt et al. 1983; Pirtle et al. 1986; Naidenko et al. 2013). An assumption in our study is that PRV was the cause of death in any panther in which PRV was found. This was based on the lack of PRV antibodies in live-captured panthers in this and previous studies (Roelke et al. 1993), histologic findings, and the clinical history associated with mortality (acute or unknown onset of clinical signs or death, and other causes of death ruled out). The virus was detected only in cases with a clinical history consistent with PRV, and PRV was not detected in negative controls.
Little information is available regarding risk factors for PRV infection in secondary hosts. In domestic pigs, young animals are more susceptible to infection than adults (Wittmann and Rziha 1989). Sex does not appear to play a role in susceptibility in domestic cats (Card et al. 1997). Comparison of risk factors for PRV infection in panthers was hampered by small sample size and unrelated changes in the population. With intensive management, the Florida panther population has changed in distribution, size, density, and genetics since the mid-1990s (Johnson et al. 2010). Not surprisingly, we found no differences in risk of PRV infection as a function of age class, sex, and host heterozygosity. Although most infections occurred north of I-75 (Fig. 1), the difference was not statistically significant. Coinfection with FIV was not evaluated as a risk factor, but PRV infections were documented in panthers with and without FIV coinfection.
In earlier studies, intraspecific aggression was the most common cause of death in radio-collared panthers, followed by vehicular collision (Maehr et al. 1991; Taylor et al. 2002). Infectious diseases accounted for only 4% of radio-collared mortalities reported by Taylor et al. (2002), although mortality due to infectious diseases appears to be increasing (Cunningham et al. 2008). Before this study, PRV had been diagnosed in only three radio-collared panthers; our retrospective and prospective analyses and testing yielded an additional seven confirmed infections (five radio-collared) for a total of eight radio-collared panther PRV cases. These represented approximately 4.8% of radio-collared mortalities, which is probably an underrepresentation. Therefore, we used various diagnostic criteria to categorize unknown mortality cases based on their likelihood of PRV infection. Combining all categories (confirmed, probable, suspected, and possible) gave 32 (19%) mortalities due to PRV, which is certainly an overestimate. Using a probabilistic approach, we estimated the number of radio-collared mortalities attributed to PRV to be 15 (8.9%). This estimate may be influenced by numerous factors, including vaccination against other infectious diseases, unequal capture effort across panther range, and variation in the distribution of both panthers and feral swine. Nevertheless, PRV infection is the third leading diagnosed cause of death in radio-collared Florida panthers and could affect population persistence in areas with high densities of feral swine. Hotspots for the disease, such as artificially high feral swine concentrations in fenced enclosures (which panthers can easily traverse), may act as a population sink, drawing panthers into newly unoccupied high-risk territory. During the study period, as many as eight radio-collared panthers whose death was, or could be, attributed to PRV died in or adjacent to (within 100 m of) high-fence enclosures containing dense populations of feral swine. This included six confirmed and probable cases in 2003. Although feral swine populations have declined in much of panther range, higher concentrations to the north could affect panther population expansion into central Florida (Roelke et al. 1993).
Management
Management to increase white-tailed deer (Odocoileus virginianus) density could decrease the risk of PRV to panthers by reducing their dependence on feral swine as a prey species. Additionally, feral swine management such as vaccination, harvest management targeting higher risk older swine, reduction of swine densities, and restriction of the use of dogs for hunting feral swine in panther range may decrease the prevalence of PRV or of viral shedding. These management actions would be most practical, and likely more effective, in closed feral swine populations. The goal of feral swine management would be to maintain the herd at optimal densities for preserving the prey source while reducing viral shedding. Glass et al. (1994) advised vaccinating panthers with an inactivated PRV vaccine; research is needed to identify a safe and effective vaccine suitable for use in panthers. A commercially available inactivated vaccine is not available in the US.
The PRV Eradication Program resulted in the elimination of PRV from domestic swine in the US in 2004 (Anderson et al. 2008). It remains endemic in feral swine populations, being found throughout the range of feral swine in the US (Pedersen et al. 2013) and worldwide (Muller et al. 2011). Therefore, PRV continues to be a threat to domestic animals and wildlife, especially those preying on or scavenging wild boar or feral swine (Capua et al. 1997). Further research is needed on the frequency of PRV shedding in feral swine and factors that may reactivate latent infections.
ACKNOWLEDGMENTS
We would like to acknowledge the initial work for this project accomplished by Melody Roelke and Carolyn Glass as well as biologists and veterinarians from the Florida Fish and Wildlife Conservation Commission (FWC) and the National Park Service. We also greatly appreciate the assistance of the Big Cypress Seminole Indian Reservation, Lee County Port Authority, Florida Forest Service, Immokalee Ranch, and others in the investigation of panther mortalities. The Livestock Protection Company, Inc., provided for the safe capture of panthers used in this study. The Florida Department of Health, C. A. Pound Human Identification Laboratory, University of Florida College of Veterinary Medicine, and Disney's Animal Kingdom assisted with diagnostics, and the National Veterinary Services Laboratories kindly provided PRV neutralization testing. We thank John Allen for graciously collecting feral swine samples on our behalf. Lisa Shender and Charles Dodd provided valuable edits to the manuscript. Lastly, we would like to thank the citizens of Florida who continue to support research and management of the Florida panther by FWC via the purchase of Protect the Panther license plates. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the US Fish and Wildlife Service.
SUPPLEMENTARY MATERIAL
Supplementary material for this article is online at http://dx.doi.org/10.7589/JWD-D-20-00119.