ENDOPARASITES OF FORMOSAN BLACK BEARS (URSUS THIBETANUS FORMOSANUS) DURING ACORN SEASON IN YUSHAN NATIONAL PARK, TAIWAN

The Formosan black bear (Ursus thibetanus formosanus) is a subspecies of Asiatic black bear and the only member of the family Ursidae native to Taiwan. With an estimated population of 200–600 bears throughout the island (Lin 2013), the bears are listed in the International Union for Conservation of Nature Red List as vulnerable to extinction (International Union for Conservation of Nature 2019). Despite the bears having been listed as an endangered species under the Wildlife Conservation Law of Taiwan since 1989, they continue to be threatened by poaching, illegal trade, and habitat loss. Additionally, there is widespread acceptance that disease and health are major ecosystem factors, although it remains unclear how these factors affect the bear population on the island. To our knowledge, there have been no studies on the diseases or parasites of wild Formosan black bears.

Previous studies have investigated endoparasites in the Ursidae family (Bromlei 1973; Rogers and Rogers 1976; Jenness 1997). The extent of those investigation varied because of differences in sampling targets, diagnostic techniques, and study periods (Catalano et al. 2015; Figueroa 2015; Gawor et al. 2017), which were developed to overcome limitations such as small bear populations, large home ranges, harsh environments, and varied animal behaviors. For example, to increase access to biomaterials, targets for parasite investigation have included tissue samples or gastrointestinal contents from roadkill, bodies resulting from hunting activity and semi-captive animals, and feces that were collected in a targeted or opportunistic manner (Zhang et al. 2012; Aghazadeh et al. 2015; Borka-Vitalis et al. 2017; Sheikh et al. 2017). To enhance the diagnostic rate in some studies, parasites were identified by employing multiple standard parasitic examination or biomolecular methods (Gawor et al. 2017). Taking advantage of the natural behavior of bears allows researchers the best chance to collect parasites. For example, the prevalence of parasite infection varies strongly by season (Rogers and Rogers 1976), with the period before denning, when bears have been reported to shed parasites (Rogers and Rogers 1976) serving as a key period for investigation of the endoparasites of free-ranging bears.

In Taiwan, most bears do not hibernate (Hwang 2003). However, parasites can spread more easily when animal hosts occur at high density (Arneberg 2002), and we discovered a unique site in the Central Mountain Range of Taiwan (i.e., Dafan in Yushan National Park [YNP]), where bears congregate seasonally to feed on acorns (mainly Cyclobalanopsis glauca). These acorns represent a major food source for bears from October to February (Hwang 2003). Fluctuations in the acorn abundance affect how many bears congregate in the area, how long individuals stay, how far away they come from, and the prevailing sex ratio (Hwang et al. 2002). Therefore, this study focused on the gastrointestinal parasite profiles of Formosan black bears in Dafan, YNP, which is the only known high-density habitat for the species. Molecular genetic techniques are effective for the identification of individuals and sex determination among ursids (Boulanger and McLellan 2001; Pagès et al. 2009), and they have proven useful in conservation programs (Lukacs and Burnham 2005). Those techniques were used in this study for individual and sex identification after fecal collection, contributing subject-specific disease information.

With the growing recognition of parasitism and related health measures as major factors in conservation medicine (Aguirre et al. 2007; Smith et al. 2009), the objective of this study was to identify the gastrointestinal parasite profiles of Formosan black bears in YNP and to thereby contribute to enhancing associated conservation programs by facilitating disease monitoring.

Yushan National Park, the largest national park in Taiwan, comprises 1,055 km² of the Central Mountain Range. The study site, Dafen, where bears congregate in the fall to feed on acorns, is located on the eastern side of YNP (23°22′25″N, 121°05′21″E). It is 40 km (a 3-day hike) from the park entrance and comprises an oak-dominated habitat of about 10 km² at an elevation of 1,100–1,600 m. The climate is temperate, with a mean precipitation of 116±49.4 mm/yr. The highest precipitation occurs between June and September. The mean temperature was 15.2±1.9 C, with the highest values between July and August (27.3±1.6 C) and the lowest between January and December (7.2±1.5 C; Central Weather Bureau of Taiwan 2013).

Given the remoteness of the study site, fecal samples had to be prepared on site. During 2008–12, fecal samples were mainly collected along ridge lines and acorn survey transects in Dafen, with very few being collected from hiking trails. Bear scat samples were identified by size, shape, and color and the presence of nearby bear tracks. Freshness was determined by the color, moisture, and sharpness of features, and the samples were categorized into <2 d, 3–7 d, 1–2 wk, 2–4 wk, and >1 mo old. The upper part of the scat was collected with, and stored in, sealable plastic bags and then labeled with the collection date.

To avoid bias caused by repeat sampling, only one scat was sampled when fecal samples were <50 m apart and were considered to have been excreted at about the same time. Only scat samples considered to be <2 wk old were selected for analysis. We collected 705 scat samples but only chose the freshest 20 fecal samples to represent each month (from October to February) if >20 fecal samples were collected in a particular month. We used different storage methods depending on how old the scat samples appeared to be. For samples >1 wk old, to avoid breakdown of dead parasites, we mixed one part fecal sample with two parts 10% formalin and stored the mixture for later examination. For fresher samples, we used one of three methods: 1) storage without formalin fixation awaiting direct examination, 2) air drying of a wet mount on a slide with methanol fixation, and 3) preparation following the protocol for the Harada-Mori filter paper culture technique (Ash and Orihel 1987).

Before examination, each formalin-fixed sample was centrifuged at 3,000 × G for 10 min, and the supernatant was removed, ensuring at least 5 g of feces was available for examination. Fresh samples without formalin were examined directly.

Parasite eggs and oocysts were recovered from scat samples by the direct-smear method for the formalin-fixed samples and by the floatation method for the fresh samples. Regarding the direct-smear method, a wet mount slide was prepared with normal saline solution and 0.05 g of feces. Regarding the floatation method, saturated sodium chloride solution and 1.5–2 g feces were added to a 15-mL centrifuge tube, saturated sodium chloride solution was added until the surface of the fluid reached the point balanced with surface tension, and a slide was used to cover the top of the tube after letting it stand for 10–15 min (Price and Reed 1970; Soulsby 1982). Eggs and oocysts were observed under a microscope (Eclipse E100, Nikon, Tokyo, Japan) at 100× and 400× magnification and measured with an ocular micrometer calibrated using a stage micrometer. The species of each parasite was identified morphologically according to previous studies (Sprent 1968; Soulsby 1982; Samuel et al. 2001; Testini et al. 2011). For both preparation methods, two slides were prepared for examination.

Larvae were cultured when a fresh fecal sample contained parasite eggs that were not identifiable or for double confirmation when necessary. We identified each parasite based on the characteristic morphology of the rhabditiform and filariform larvae.

Harada-Mori filter paper culture technique: We used a modified version of the method proposed by Ash and Orihel (1987) for field research. Filter paper was folded at one end and cut into 6 × 3-cm slices. The filter paper was opened, fresh feces were smeared (0.1 cm thick) in the center third of the filter paper and the paper was refolded. It was placed vertically into a sealing plastic bag, and 3 mL of sterile water was added. The lower 1 cm of the filter paper was submerged in water, but the water was not in contact with the feces directly. After culturing at 25–30 C for 4 d, a sample of water was removed for larval identification, and the remaining water was stored at –20 C in a 5-mL freezer tube containing 2 mL of 70% ethanol for future examination. Thereafter, another 3 mL of sterile water was added to the bag containing the filter paper, and larvae were cultured again for 10 d, identified, and then stored as previously described.

Tile larval culture technique: A tile was soaked in sterile water for 30 min, and fresh feces was smeared on the tile (0.5 cm thick). The tile was placed in a glass Petri dish and sterile water was added to the height of the tile, without touching the feces. The dish was covered with another petri dish and stored at 25–30 C for 1 wk. The water was sampled for larval identification using a light microscope.

Modified carbol fuchsin staining: To conduct modified carbol fuchsin staining, fresh feces or a prepared fecal sample was smeared on a slide. The slide was then stained following a previously described protocol (Ortolani 2001).

The species of the first- and third-stage larvae were identified morphologically according to previous studies (Benbrook and Sloss 1976; Levine 1985; Bowman 2013).

To investigate the abundance of parasitic infection, 1 g of each feces sample was placed in a 15-mL centrifuge tube and mixed with 15 mL saturated sodium nitrate solution. The solution was then instilled into a pair of McMaster parasite egg–counting chambers, which enabled a known volume of fecal suspension (2×0.15 mL) to be examined microscopically. The solution was left to stand for 30 s before identifying and counting the eggs. The mean number of eggs per chamber (i.e., the mean abundance) was multiplied by 100 to obtain the eggs per gram (EPG; Chang 1996).

The mean intensity of the parasitic infection was defined as the mean EPG of an infected fecal sample, calculated as mean intensity=mean abundance×100/frequency of infection. We use the term frequency instead of prevalence because each fecal sample could not be assigned to a particular individual.

Genetic techniques, such as detection of microsatellite molecular markers, and sex determination using the amelogenin locus have been successfully applied in Formosan black bears (Chen 2014). We applied those genetic techniques to allow analysis of subject-specific data. Because of the nature of the scat-collection method, repeat sampling may have occurred; all scat samples underwent parasite examination on an individual basis, and the results were recorded on an individual basis. Any parasite eggs in the scat were considered to indicate infection.

All statistical analyses were performed with SPSS software (version 18.0, SPSS Inc., Chicago, Illinois, USA). Frequency of infection, mean EPG, mean intensity of parasitic infection, and richness of the parasite species were recorded (Müller-Graf et al. 1996). Frequency of infection was compared among parasite species using the chi-square test. Mean EPG and mean intensity of parasitic infection were compared among the 5 yr of acorn seasons using the Kruskal-Wallis test. Mean intensity of parasitic infection in 2010 was compared between sexes and among the scat samples of individually identified and unidentified bears using the Mann-Whitney U-test, whereas the frequency of infection in 2010 was compared using Fisher's exact test. Statistical significance was set at P<0.05.

We collected 705 fecal samples and examined the parasites in 220 samples (all of which underwent the floatation, direct-smear method, and modified carbol fuchsin staining; 154 samples underwent the Harada-Mori filter paper culture technique, and 10 samples underwent the tile larval culture technique). The frequency of infection was 77.3% (170/ 220). We found five kinds of nematodes through morphologic identification of eggs and larvae, comprising Baylisascaris transfuga (assisted by adult worm dissection; 65.5%), Strongyloides sp. (11.4%), hookworm (9.6%), Trichostrongylus sp. (3.6%), and Oesophagostomum sp. (3.6%). We also found an undetermined protozoan of the genus Cryptosporidium (2.7%), the cestode genus Taenia (0.5%), and two undetermined parasites (speculated to be Physaloptera sp. and Gongylonema sp.; 1.0%). Details regarding the parasite eggs found on fecal examination and the larvae identified from larval culture are presented in Figures 1 and 2.

The EPG was not determined for the two unknown helminths and the Taenia sp. (because of the low infection rates) or for the Cryptosporidium sp. (because of the special staining procedure that would have been required). The mean EPG across the 220 samples was 456.1±851.7 for B. transfuga and 25.6±159.3 for other helminths. The mean infection intensity was 790.2±997.4 (with 127 out of the 220 samples indicating infection) for B. transfuga and 403.6±515 (with 14 out of the 220 samples indicating infection) for other helminths.

The frequency of parasite infection was similar among acorn seasons (χ²=4.276, P=0.37). A total of nine parasite species were detected across the study period, and the richness of parasite species was greatest in 2010 (Table 1). There were no significant differences among acorn seasons in mean EPG (H=4.068, P=0.397) or mean intensity (H=2.045, P=0.728).

In 2010, 30 individual male bears (identified based on 44 scat samples) and 16 individual female bears (identified based on 24 scat samples) were identified (Chen 2014). There were no significant differences between males and females in the frequency of parasite infection (χ²=0.339, P=0.403) or B. transfuga infection (χ²=0.339, P=0.403) or between mean EPG for B. transfuga (T=377.5, P=0.981) or mean intensity of B. transfuga infection (T=205.5, P=0.645). There were also no significant differences between scat samples from individually identified and unidentified bears in the frequency of parasitic infection (χ²=0.044, P=0.548), mean EPG (T=2,689.5, P=0.795), or mean intensity of parasitic infection (T=1,180.5, P=0.983; Figs. 3, 4).

The frequency of endoparasitic infection in Formosan black bears was high (>70%). Other studies of bears have also indicated prevalences ≥50% (Table 2). In particular, a similarly high frequency of parasites to that in the current study was reported based on scat samples of Asiatic black bears in the Indian Himalayas (Sheikh et al. 2017), but the study did not report the number of individuals they sampled or whether the bears were concentrated in the area sampled. The high frequency of infection in our study may be explained by the bears having congregated seasonally at high density in a small area (i.e., Dafen). They came from all areas of the park, travelling up to 20 km to reach the concentrated food source (i.e., acorns; Hwang et al. 2010; Lin 2017). This unique food source is clearly important to the health of the bears but may also have the negative effect of promoting the spread of parasites. Studies in brown bears in Slovakia showed similar findings: the prevalence of B. transfuga infection was as high as 52.9–63.8% in separated or restricted areas (Major et al. 2009; Štrkolcová et al. 2018) and 14.3% in other areas (Goldová et al. 2003). The extensive overlap of male and female bear home ranges during acorn seasons (Hwang et al. 2010) may also explain why we found no differences between the sexes in the parasite infection frequency, B. transfuga infection frequency, mean EPG for B. transfuga, or mean intensity of B. transfuga infection. In contrast, previous research on American black bears (Ursus americanus) found significant differences between the sexes, with male bears exhibiting higher mean intensities of helminth infections than females had (Conti et al. 1983) and Baylisascaris eggs mainly being observed in males (67%; Johnson et al. 2013).

Baylisascaris spp. infection is common in wild animals, including ursids (Samuel et al. 2001; Schaul 2006). In fact, B. transfuga has been reported in all extant species of bear in the family Ursidae, and no non-bear definitive hosts are known (Schaul 2006; Sapp et al. 2017). The parasites can be transmitted to bears through ingestion of larvated eggs or infected mammalian paratenic hosts (Hoberg et al. 2018). Although Zhang et al. (2012) described Baylisascaris shroederi as a major cause of death in wild giant pandas (Ailuropoda melanoleuca), B. transfuga has rarely been reported to cause mortality in other hosts, including bears. However, infection can be dramatic in bears under stress, resulting in blockage of the intestinal tract by adult worms (Schaul 2006). Additionally, larva migrans syndrome caused by migration of B. transfuga larvae in the host has been reported in naturally infected Alaska moose (Alces gigas), experimentally infected mammals and birds, and Japanese macaques (Macaca fuscata) kept with American black bears in a Japanese zoo. Subjects were found to suffer from visceral, ocular, and even neural larva migrans, and the clinical manifestations varied from mild (or not apparent) to consistent weakness (Schaul 2006; Hoberg et al. 2018). However, the above-mentioned Japanese macaques were reported to have died of B. transfuga infection.

During 2008–12, B. transfuga was the most prevalent parasite identified in bears in this study, with a frequency >60% each year. The infection rate was relatively high (Table 2). This suggests intensive parasite exchange and, again, that may be due to bears congregating at high density in a small area. At the same time, the lack of significant differences among acorn seasons suggests a stable state of infection. Long-term observation of parasitic infection has been widely recognized as a useful approach for exploring diversity and changes in complex host-parasite systems, and the natural biological characteristics of this parasitism has important roles in structuring ecological communities (Hoberg et al. 2008). Moreover, there is growing recognition that long-term epidemiologic data sets are valuable for detecting the long-term cumulative effects of climate change, ontogenetic changes in host diets, niche shifts, and feeding specializations (Hoberg et al. 2008). The nature of the long-term investigation and relatively stable EPG and intensity numbers in Formosan black bears mean that the data are valuable in representing the balance between host and parasite infection and have the potential to be considered as a bioindicator for the health of the ecosystem and the wild black bear populations (Marcogliese 2005).

In this study, all scat samples were examined by the same researcher (T.-W.C.), avoiding the bias related to interobserver differences in the identification of the abundance of parasite species. The abundance of parasite species identified in the bear feces was greatest in 2010, the same year with the most-fruitful acorn season and the highest density of claw marks on trees (Hwang 2010). In contrast, 2009 was the poorest acorn season during the study period, with a low abundance of parasite species and with only eight scat samples collected because of the scarceness of the bears. This finding concurs with the epidemiologic model that suggested that temporary host population density is positively related to species richness in parasite communities (Arneberg 2002).

The second most common parasite identified was Strongyloides spp. That result is consistent with findings in American black bears (Schaul 2006). The high frequency of that parasite in Formosan black bears is of concern because several species of Strongyloides are known to be pathogenic to young animals (Conti et al. 1983). The identification of Trichostrongylus sp. and Oesophagostomum sp. in this study was unusual. Those species seem to be ubiquitous among herbivores and primates worldwide but have never been reported in ursids (Garcia 2007). Identification of heterogeneous nematodes has been reported in polar and grizzly bears in captivity (Canavan 1929; Stiles and Baker 1935), and food contamination was considered the source of infection. Formosan black bears coexist with macaques, wild boars (Sus scrofa), and herbivores in forests. Given that black bears are the top carnivores in the food chain and also opportunistic omnivores (Hwang et al. 2002), the presence of Trichostrongylus sp. and Oesophagostomum sp. in the feces may indicate cross-infection between sympatric species or ingestion of contaminated food items.

The two unknown parasites were thought to be Physaloptera sp. and Gongylonema sp. The source of Physaloptera infection is believed to be ingestion of infected beetles by the final host (Conti et al. 1983; Schaul 2006). This finding fits with the diet of Formosan black bears in the wild (Hwang et al. 2002). Gongylonema sp. has been reported at a low frequency of detection in American black bears (Kirkpatrick et al. 1986). Nevertheless, the sporadic reports in various populations suggest that black bears are natural hosts of Gongylonema sp. (Schaul 2006). Overall, our study identified multiple parasite species that have not previously been described in Asiatic black bears. Although preventive measures were applied during the study, larvae from contaminated environmental soil may have been another source of fecal parasites because we collected samples from dropped scat instead of directly from the gastrointestinal tract. Careful interpretation is warranted when encountering larvae commonly seen in other sympatric species but with a low identification rate in bears (Traversa et al. 2014).

Our finding that zoonotic parasites, including B. transfuga, Taenia sp., and Cryptosporidium sp., were present in the feces of Formosan black bears highlights the risk of zoonotic diseases among high-risk human populations. However, there is no unequivocal evidence of naturally occurring B. transfuga infection in humans (Bauer 2013). In contrast, Taenia sp. is known to cause cysticercosis in humans, with Taenia solium being identified as a leading cause of death from foodborne diseases in 2015 (World Health Organization 2015). Additionally, Cryptosporidium sp. causes acute and sometimes fatal diarrhea in humans and is considered as one of the most common foodborne and waterborne diseases, with worldwide spread (Rossle and Latif 2013). Risk factors for exposure to these zoonotic parasites include contaminated or inadequately treated drinking water, handling and consumption of locally caught and inadequately cooked game or fish, and direct contact with fecal material or soil (Schurer et al. 2013). Educational efforts regarding the zoonotic risks may have the added benefit of deterring illegal bear use and poaching.

Our study was subject to several limitations. First, the collection of fecal samples was opportunistic, so there was uncertainty regarding guaranteeing the freshness of all samples. Second, although efforts were made to preserve the samples, transportation to the laboratory was prolonged because of the remoteness of the study location. Delayed examination of the samples may have led to an underestimation of the abundance of parasite eggs (Foreyt 2001). Larvae collection may have been limited for the same reason; adult worm dissection and molecular analyses would improve species identification. In particular, the finding of Baylisascaris venezuelensis in South American Andean bears (Tremarctos ornatus; Mata et al. 2016), which represents a relatively isolated population of ursids, further supports the idea that B. transfuga represents an assemblage of species globally and highlights the need for further molecular and morphologic research to characterize possibly cryptic species (Sapp et al. 2017).

The authors thank Kwong-Chung Tung and Yi-Yang Lien for their technical support and instruction in parasitology, and we are grateful to many research assistants and volunteers who helped with fieldwork. We thank the Yushan National Park Headquarters and Taiwan Black Bear Conservation Association for providing research funding.