Abstract
American black bears (Ursus americanus) are common, widely distributed, and broad-ranging omnivorous mammals in northern California forests. Bears may be susceptible to pathogens infecting both domestic animals and humans. Monitoring bear populations, particularly in changing ecosystems, is important to understanding ecological features that could affect bear population health and influence the likelihood that bears may cause adverse impacts on humans. In all, 321 bears were captured between May, 2001, and October, 2003, and blood samples were collected and tested for multiple zoonotic and vector-borne diseases. We found a PCR prevalence of 10% for Anaplasma phagocytophilum, and a seroprevalence of 28% for Toxoplasma gondii, 26% for Borrelia burgdorferi, 26% for A. phagocytophilum, 8% for Trichinella spiralis, 8% for Francisella tularensis and 1% for Yersinia pestis. In addition, we tested bears for pathogens of domestic dogs and found a seroprevalence of 15% for canine distemper virus and 0.6% for canine parvovirus. Our findings show that black bears can become infected with pathogens that are an important public health concern, as well as pathogens that can affect both domestic animals and other wildlife species.
Introduction
T
Through vectors, scavenging, and occasional predation, black bears may be exposed to pathogens, including agents of anaplasmosis, borreliosis, toxoplasmosis, trichinosis, plague, tularemia, brucellosis, bartonellosis, and rickettsiosis (Anderson and May 1979, Hill and Dubey 2002, Rah et al. 2005). Thus, bears may be a source of human disease, especially trichinosis and toxoplasmosis, through hunting and consumption of undercooked meat (Nutter et al. 1998). Black bears may be sympatric with wild and domestic canids, potentially sharing pathogens such as canine distemper virus (CDV) and canine parvovirus (CPV). Additionally, black bears have become adapted to living in close proximity to humans, which can result in more bear–human interaction (i.e., eating garbage, car strikes) and risk of disease transmission.
Bears may also be a source of infection to protected wildlife species, such as fishers (Pekania pennanti), which have experienced considerable reduction and fragmentation of their historic range and are candidates for Endangered Species Act protection (Gabriel et al. 2006, Davis et al. 2007). Infection with CDV, CPV, and Toxoplasma gondii in particular could be a severe problem for fishers. The Hoopa Valley Indian Reservation (HVIR) in the Klamath Mountains of northern California encompasses 339 km2 of intensely managed forest, a dense black bear population, and an important population of fishers (Matthews et al. 2008). Since logging of Douglas fir (Pseudotsuga menziesii) trees began on the reservation in the 1940s, the intensity of logging has increased considerably, and now black bears utilize managed and disturbed habitat more than would be expected by chance alone (Costello 1992, Costello and Sage 1994). Here we aim to determine the presence and prevalence for multiple infectious diseases in black bears on the HVIR that may impact wildlife conservation, public health, and domestic animal health.
Methods
Study area
The HVIR in the Klamath Mountains of northern California (40.0390N, −123.4093W) is approximately 366 km2 in area and is bisected by the Trinity River (Fig. 1). Elevations vary from 76 to 1170 meters with approximately 339 km2 forested with Douglas fir, tanoak (Lithocarpus densiflorus), madrone (Arbutus menziesii), Oregon white oak (Quercus garryana), and California black oak (Q. kelloggii). Common shrubs are evergreen huckleberry (Vaccinium ovatum), tobacco brush (Ceanothus velutinus), salal (Gaultheria shallon), and poison oak (Toxicodendron diversilobum). Nonforested areas include residential areas, prairies, large rock outcrops, and fields.

Map of study area and capture locations for black bears captured on the Hoopa Valley Indian Reservation, Humboldt County, California between May, 2001, and September, 2003.
Trapping and sample collection
Bears were captured as part of ongoing monitoring and management of the population using culvert traps baited with salmon between May, 2001, and October, 2003. Traps remained at each location for 6–10 days. Locations were selected on the basis of signs of bears, evidence of tree damage, shade, level ground, and a good location to process bears (Fig. 1). Traps were checked in the early morning and late afternoon. Captured bears were immobilized by a pole syringe using 4 mg/kg of ketamine hydrochloride and 2 mg/kg xylazine hydrochloride or 4 mg/kg tiletamine/zolazepam. Once anesthetized, bears were tattooed on the inside of the upper lip and tagged with colored and numbered ear tags in both ears. Blood was collected from the femoral vein into tubes containing EDTA and no anticoagulant. An upper first premolar was extracted to determine age by cementum annuli (Matson's Laboratory LLC, Milltown, MT). Blood samples were frozen at −20°C or colder until analyzed.
Laboratory analysis
Sample volume and financial resources constrained laboratory tests. All tests were performed on all bears captured in 2001 at the initiation of the study. Subsequent testing was performed as funding and samples were available. Samples from recaptured bears were not tested. DNA was extracted from blood using a DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA) following the blood spin-column protocol. PCR was performed for Borrelia burgdorferi sensu stricto, Anaplasma phagocytophilum, and Rickettsia spp. using real-time PCR (RT-PCR) (Leutenegger et al. 1999, Stenos et al. 2005, Drazenovich et al. 2006). Samples were considered positive if the cycle threshold was less than 40 and there was a characteristic amplification curve. Positive and negative controls were included in each run.
Indirect immunofluorescent assays (IFA) were performed for antibodies to Ehrlichia canis, A. phagocytophilum, R. rickettsii, B. burgdorferi, CDV, and CPV. Serum was diluted in phosphate-buffered saline (PBS) as reported in Table 1 and applied to commercial slides (VMRD, Pullman, WA). Slides were incubated at 37°C with moisture for 30 min, washed three times with PBS, and incubated for 30 min with fluorescein-labeled goat anti-dog immunoglobulin G heavy and light chain (Kirkegaard & Perry Laboratories, Gaithersburg, MD) diluted in PBS at 1:100. Slides were washed three additional times and counterstained with Eriochrome Black. Positive control serum from dogs was included in each batch. Samples were considered positive if they had strong fluorescence detected that was compatible with the morphology of the antigen on the slide; cutoffs are reported in Table 1.
Confidence intervals were calculated by inverting the score test.
CI, confidence interval; RT-PCR, TaqMan real-time PCR; CT, cycle threshold; Sero, serology; IFA, immunofluorescent assay; LAT, latex agglutination test; ELISA, enzyme-linked immunosorbent assay; OD, optical density; PHA, passive hemagglutination; HI, hemagglutination inhibition; SAT, slide agglutination test; NA, not applicable.
T. gondii antibodies were assessed using latex agglutination (LAT) (Eiken Chemical Co, Tokyo, Japan). A titer of ≥1:64 was considered positive. A modified enzyme-linked immunosorbent assay (ELISA) was used to test for antibodies to Trichinella spiralis (Chomel et al. 1998). The cutoff point for a positive test was set at an optical density of 0.3. Antibodies to Yersinia pestis were tested using passive hemagglutination (PHA) (World Health Organization 1970, Wolff and Hudson 1974); a titer of ≥1:32 was considered positive. Positive serum on PHA was tested with hemagglutination inhibition (HI). If the HI titer was below the PHA titer by less than two-fold, the PHA test was considered nonspecific and the sample was scored as negative (Hoar et al. 2003). Slide agglutination was used to evaluate antibodies to Francisella tularensis (Becton Dickinson, Sparks, MD). Any titer ≥1:20 after two consecutive tests was considered a positive result (Chomel et al. 1998). Testing for Brucella abortus antibodies was performed using the buffered acidified card antigen test (United States Department of Agriculture, Ames, IA) (Angus and Barton 1984).
Blood samples were cultured on heart infusion agar containing 5% rabbit blood and incubated in 5% CO2 at 35°C for up to 4 weeks for Bartonella spp. (Chomel et al. 1995a). Morphologic characteristics and the time of growth on the plate were used for the initial identification of the isolates (Koehler et al. 1997).
Statistical analysis
Data were maintained in a spreadsheet (Excel, Microsoft, Redmond, WA). Statistical analyses were performed using R software (R Development Core Team,
Results
A total of 321 black bears were captured, of which 153 (48%) were female and 168 (52%) were male (Table 2). Ages ranged from 1 to 20 years, with a median of 4 and mean of 5 years (age was not determined for 13 of the bears). The age distribution was skewed toward older bears. Thirty-five bears were recaptured during the study period.
Approximately 10% (30/288) of bears were actively infected (positive by PCR) with A. phagocytophilum, whereas none were PCR positive for Borrelia spp. (n=80) or Rickettsia spp. (n=164) (Table 1). Pathogens with the highest seroprevalence were T. gondii (67/239, 28%), B. burgdorferi (54/208, 26%), A. phagocytophilum (54/210, 26%), and CDV (24/157, 15%). Approximately 8% were seropositive for F. tularensis (6/80), 8% for T. spiralis (18/239), 5% for Rickettsia spp. (10/209), 1% for E. canis (2/158), 1% for Y. pestis (1/80), and 1% for CPV (1/157). No bears tested positive for B. abortus (n=239) or Bartonella spp. (n=80).
There was not a statistically significant difference in prevalence between males and females for exposure to pathogens. However adults were less likely to be A. phagocytophilum PCR positive than subadults (OR=0.2, p<0.001), and more likely to be T. gondii or T. spiralis seropositive (OR=2.8, p<0.001 and OR=4.3, p=0.012, respectively) (Table 3). No other significant associations were found.
p values and ORs are based on the Fisher exact test.
Neg, negative; Pos, positive; OR, odds ratio.
Discussion
Bears may serve both as a sentinel for diseases and a source of infection for animals and people. Monitoring bear populations, particularly in changing ecosystems, is important to understand ecological features that affect bears, bear population health, and bear–human interactions. Bear density may influence bear pathogen prevalence. In North America, black bear density ranges from a low of 0.06 bears/km2 in Arizona (LeCount 1987) to 0.99 bears/km2 in New Jersey (Carr et al. 2005). An estimate in 1998 on HVIR was 0.18 and 1.33 bears/km2 at sites separated from each other by the Trinity River (Matthews et al. 2008). Besides pathogen circulation, increased bear density can increase bear–human conflicts (Maine Department of Inland Fisheries and Wildlife, accessed September, 2014).
We found evidence of active infection with A. phagocytophilum in 10% of bears, with younger bears more frequently PCR positive than adults, consistent with a common pathogen that creates lasting immunity. This Ixodes pacificus tick-transmitted pathogen is common among other wildlife in the western United States, where the reservoirs are sciurids (Richter et al. 1996, Nieto and Foley 2008). There is also high risk of anaplasmosis in dogs (Canis lupus familiaris) and people in Humboldt County, on the basis of seropositivity in dogs and prevalence in I. pacificus (Madigan et al. 1990, Foley et al. 1999, Carrade et al. 2011, Foley and Piovia-Scott 2014). There are concentrations of A. phagocytophilum—seropositive dogs in the northern coast ranges of California, Sierra Nevada foothills, Siskiyou and Cascade Mountain areas of Oregon, and Puget Sound area of Washington (Carrade et al. 2011). Canine seroprevalence is patchy, about 19% in far northern Humboldt County (Bowman et al. 2009) and 51% in southern Humboldt County (Foley et al. 2001, Foley et al. 2007). Whether bears suffer clinical anaplasmosis is not known.
A similar geographical distribution to anaplasmosis is expected for borreliosis, considering that the two diseases share the same tick vector and reservoirs in California (Brown and Lane 1992, Lane et al. 2005). B. burgdorferi infection can cause Lyme disease with arthritis and sometimes cardiac and neurological disease in people and dogs (Lane et al. 2001, Foley et al. 2007). The Centers for Disease Control and Prevention ranks human risk as moderate in Humboldt County (Centers for Disease Control and Prevention, accessed November, 2013). The 26% B. burgdorferi seroprevalence among bears exceeds the 17% prevalence in bears, even in Wisconsin, where Lyme borreliosis is highly endemic in wildlife and humans (Kazmierczak et al. 1988). In contrast, dogs may have even higher seroprevalence, such as greater than 50% from hyperenzootic areas in the eastern and midwestern United States (Bowman et al. 2009), although no higher than 14% in California (Carrade et al. 2011). Seroprevalence was similar among age classes of bears in the present study. Immature I. pacificus ticks tend not to feed on bears, making bears an unlikely source of infection for people. However, bears may still serve as valuable sentinels for the presence of the pathogen in this landscape. No bears were B. burgdorferi PCR-positive, although the sensitivity of this test is low because the spirochetes are in blood only briefly early in the course of infection (Aguero-Rosenfeld et al. 2005). Sensitivity and specificity of serologic tests for Borrelia are also low; there is no way to differentiate between recovered and chronic infections, and there is considerable cross-reaction with other Borrelia spp. present in other mammalian species in Humboldt County (Eisen, et al. 2009).
Spotted fever group rickettsioses are emerging worldwide (Parola et al. 2005, Jones et al. 2008). For example, a cluster of 16 human cases of Rocky Mountain spotted fever in Arizona was associated with an epidemic in dogs and infestation of the brown dog tick, Rhipicephalus sanguineus (Demma et al. 2006). In Humboldt County, bears would likely be exposed to rickettsiae via Dermacentor spp. ticks, which host pathogenic and nonpathogenic rickettsiae (Furman and Loomis 1984, Shapiro et al. 2010). Rickettsia spp. have been detected by PCR from ticks of black bears in Florida, Georgia, and Louisiana (Yabsley et al. 2009, Leydet and Liang 2013). We found 5% seroprevalence, but could not identify the Rickettsia species because of lack of positive PCR results, precluding DNA sequencing. This finding is not surprising as rickettsiae are often associated with endothelium, so RT-PCR on blood has low sensitivity.
E. canis is sporadically detected in the western United States, with seropositive dogs identified primarily on the California border with Mexico and the southern Central Valley (Carrade et al. 2011). The pathogen is transmitted by R. sanguineus, which thrives on dogs in peridomestic settings, including shelters and some multiple-dog facilities, but is rare in Humboldt County (Foley, unpublished data). The seroprevalence of ehrlichiosis in dogs is low in Humboldt County, therefore it would be unlikely to detect ehrlichiosis in bears on the HVIR, consistent with the seroprevalence <2%.
We did detect a F. tularensis seroprevalence of almost 8%. Tularemia is a potentially fatal multisystemic disease (Hayes 2005, Foley and Nieto 2010). Humans can be exposed through contact with an infected animal and arthropod vectors such as ticks, biting flies, mosquitoes, and fleas; infection can also be water borne (Centers for Disease Control and Prevention, accessed May, 2014). Very little is known about the true rate of infection among animals and vectors in California, because the pathogen often occurs at a very low enzootic level, but it can occasionally emerge (Foley and Nieto 2010). For example, an outbreak was observed in an outdoor-housed primate colony in Davis in 2010, and cases were reported in a child in Sacramento and two adults in Napa County in that year as well (Vector-Borne Disease Section California Department of Public Health 2010, Sammak et al. 2013). There are no available data from surveillance in Humboldt County. Exposure has been reported in bears previously however, with black bears in Idaho having a F. tularensis seroprevalence as high as 19% (Binninger et al. 1980). These two datasets suggest that bears could be useful as sentinels for this pathogen, which receives so little attention during public health surveillance.
In contrast, we found little evidence of exposure to Y. pestis, the agent of plague. The pathogen occurs in California most commonly in chipmunks, ground squirrels, and other rodents, where infection can be subclinical or contribute to mass die-offs (Nelson 1980). Carnivores may be infected through fleabite or ingestion of infected rodents (Clover et al. 1989). Earlier studies of black bears in California found plague antibody prevalence from 15% to 36% (Ruppanner et al. 1982, Smith et al. 1984, Clover et al. 1989). The Humboldt County area is not considered endemic for plague, but seropositive carnivores have been detected previously (Smith et al. 1984, Chomel et al. 1994).
Among directly transmitted zoonoses, there was widespread exposure to T. gondii as has been shown before (Mortenson 1998). Cats (Felis catus) and bobcats (Lynx rufus) are definitive hosts for T. gondii and are common in the HVIR (Montoya and Liesenfeld 2004). Bears may be exposed through infected felid feces or ingestion of T. gondii organisms encysted within muscle tissues of prey (Riemann et al. 1978). Most authors report the highest prevalence in older male bears, possibly because of male feeding preference and larger home ranges (Binninger et al. 1980, Ruppanner et al. 1982, Briscoe et al. 1993, Chomel et al. 1995b, Zarnke et al. 1997a). We found adults to be three times more likely to be seropositive than subadults, but there was no difference between sexes.
Trichinosis is caused by the helminth T. spiralis, which encysts in muscle and can be transmitted through ingestion of undercooked contaminated pork, bear, and other wildlife (Zarnke et al. 1997b). Estimates of prevalence in bears range from 13% in Idaho and California to 27.5% in Alaska (Binninger et al. 1980, Ruppanner et al. 1982), compared with 8% in our study. Adults were four times more likely to be infected than subadults. Trichinosis tends to be more common in carnivores from remote regions versus those in more human-accessible areas (Worley et al. 1974, Schad et al. 1986). None of the bears we tested were positive for B. abortus, consistent with studies in California, Idaho, and Alaska (Binninger et al. 1980, Drew et al. 1992, Chomel et al. 1998). Similarly, we did not detect infection with Bartonella spp. (Chomel et al. 2006), although B. rochalimae and B. vinsonii subsp. berkhoffii have been isolated from gray foxes on HVIR (Gabriel et al. 2009).
We were also concerned that bears could be exposed to pathogens typical in dogs. We found 15% CDV antibody prevalence, compared to 36% in polar bears (Ursus maritimus) from Alaska and Russia (Follmann et al. 1996), 8% in grizzly bears (Ursus horribilis) from Alaska (Chomel et al. 1998), and 8% in black bears from Florida (Dunbar et al. 1998). Typically exposure rates of bears gradually increase over age (Chomel et al. 1998, Mortenson 1998). CDV can be spread by dogs, mustelids, coyotes (Canis latrans), and other carnivores that might be sympatric with bears. CDV can be immunosuppressive and cause fatal respiratory, gastrointestinal, and neurological disease in canids, procyonids, felids, and mustelids (Deem et al. 2000, Greene and Vandevelde 2012). Lack of antibodies in bears at HVIR would suggest that the population is vulnerable (Delahay and Frolich 2000, Courtenay et al. 2001, Hanni et al. 2003), which is a concern due to a local dog population that is largely unvaccinated (Stephenson, unpublished data).
Canine parvovirus may cause subclinical disease or infected animals can suffer diarrhea, vomiting, dehydration, and death (Barker and Parrish 2001). Infrequent CPV vaccination of dogs could also put bears, canids, and mustelids, such as fishers, at risk. Although <1% of bears had CPV antibodies, fishers at HVIR have a seroprevalence of 42% (Gabriel et al. 2006). If dogs were the source of infection, bears and fishers that overlap residential areas may be at increased risk, as was shown for gray foxes near developed sites in Marin County (Riley et al. 2004). Spatial analysis of a large sample of serologic results from carnivores in HVIR will eventually allow us to evaluate such relationships.
The assays we performed are validated in humans and domestic animals, but they are not validated for bears, which may reduce sensitivity and specificity. Secondary conjugates used for serology were anti-dog, which could decrease sensitivity due to lack of binding to bear antibody. Furthermore, serology detects exposure but cannot distinguish active infection and may be negative in acute infections. Another concern is the sampling strategy employed, where we were able to test all bears in the first year but were limited in our ability to test in subsequent years, which could reduce our ability to detect diseases that typically occur in outbreaks, such as plague, canine distemper, and tularemia. However, we documented infection or exposure to zoonotic, vector-borne, and canine diseases, including borreliosis, anaplasmosis, tularemia, toxoplasmosis, trichinellosis, distemper, and parvovirus. These data suggest threats to humans, dogs, and the vulnerable fisher. Future research should aim at understanding the effects these pathogens have on bears and the risk to humans from contact with bears and exposure to shared vectors. The intensive management of this forest can affect both bears and disease dynamics. Community members may contact wildlife and disease vectors through occupational and recreational activities, making understanding of disease dynamics essential for public health.
Footnotes
Acknowledgments
We thank Alexa Kownacki, Joy Worth, Joey Tse, Anna Olivero, Rick Kasten, Niki Drazenovich, and the Hoopa Valley Indian Tribe. Funding was provided by the United States Department of the Interior Bureau of Indian Affairs and the University of California Davis Center for Vector Borne Disease.
Author Disclosure Statement
No competing financial interests exist.
