Animal‐borne video systems provide insight into the reproductive behavior of the Asian black bear

Abstract Previous studies on the mating system of the Asian black bear (Ursus thibetanus) have been limited to observations of captive populations and estimations of multiple paternities. Hence, the mating system of wild bears remains poorly understood. Animal‐borne camera systems (i.e., cameras mounted on animals) provide novel tools to study the behavior of elusive animals. Here, we used an animal‐borne video system to record the activities of wild bears during the mating season. Video camera collars were attached to four adult Asian black bears (male “A” and “B,” and female “A” and “B”) captured in Tokyo, central Japan, in May and June 2018. The collars were retrieved in July 2018, after which the video data were downloaded and analyzed in terms of bear activity and mating behavior. All the bears were found to interact with other uniquely identifiable bears for some of the time (range 9–22 days) during the deployment period (range 36–45 days), and multiple mating in males was documented. Both males and females exhibited different behaviors on social days (i.e., days when the bear interacted with conspecifics) compared with solitary days (i.e., days with no observed interactions with conspecifics). Compared with solitary days, the bears spent a lower proportion of time on foraging activities and higher proportion of time on resting activities on social days. Our results suggest that Asian black bears have a polygamous mating system, as both sexes consort and potentially mate with multiple partners during a given mating season. Furthermore, bears appeared to reduce their foraging activities on social days and engaged more in social interactions.

Sexual selection has resulted in pronounced sexual size dimorphism and the evolution of polygamous mating systems in a large number of mammals (Andersson & Iwasa, 1996). Knowledge about mating systems not only has ethological values, but also is important for species-specific conservation and management, as well as biodiversity conservation in general (Palomero et al., 2007).
However, a profound understanding of mating systems currently exists only for a relatively small number of animal species (Arnold & Duvall, 1994;Steyaert et al., 2020), as mating system research is challenging, especially for elusive species.
Among bears (Ursidae), mating behavior and mating systems are probably best documented for polar bears (Ursus maritimus), brown bears (U. arctos), and American black bears (U. americanus). Their mating activities occur in early summer, and they are considered as polygamous in the broadest sense, as both males and females mate with multiple members of the opposite sex during a given mating season (Spady et al., 2007;Steyaert et al., 2012). However, it is also known that bear populations and species exhibit variation in their mating systems, strategies, and behaviors, despite sharing several characteristics (e.g., sexual size dimorphism, polygamy) (Spady et al., 2007).
Knowledge about the mating behavior of the Asian black bear (U. thibetanus) in the wild is limited (Steyaert et al., 2020). As it is an elusive species, making direct observations in natural habitat is difficult, and a lack of intensive research limits our knowledge. However, adult aggregations and mating behavior have been observed in wild bears (Hashimoto & Anrui, 2010). Both sexes can mate various times and with various partners in captivity (Yamamoto et al., 1998), and multiple paternity litters can occur in the wild (Yamamoto et al., 2013).
This indicates that Asian black bears have a seasonal polygamous mating system. However, little information is available on the mating system of wild bears and no detailed observations of their mating behavior have been reported to date.
Biologging involves the attachment of sensors or tracking devices to animals and enables many aspects of the life histories of both terrestrial and marine species to be documented or measured (Hussey et al., 2015;Kays et al., 2015;Wilmers et al., 2015). Despite the huge technological advances in terms of biologging, determining the exact behavior of an animal at a given point in time remains difficult with traditional (high-resolution) global positioning system (GPS) tracking devices, accelerometers, or various physiological sensors, as ground-truthing data typically lack. However, recent advances in camera technologies have increased the potential for recording georeferenced animal behavior at any given point in time (Thompson et al., 2012). Such cameras are typically programmed to collect behavioral data for short time intervals ("clips"). For bears, these behaviors include stalking prey and predation outcome, sleeping, and mating, as well as search and handling time of food items (Bowersock et al., 2015;Brockman et al., 2017).
Here, we used this new camera technology to record the mating behavior of wild Asian black bears. We deployed collars with an animal-borne video camera on a small number of wild bears that were known to be sexually mature to shed light on this species' behavioral mating systems, which was for the primary objective of this study defined as interactions with other bears, especially of the opposite sex. By evaluating social interactions and mating behavior between the camera-equipped bears and unmarked individuals, we expected that 1) Asian black bears in the wild are indeed polygamous, which was shown in captive bears (Yamamoto et al., 1998). Given that, in brown bears, breeding bears devoted smaller proportion of their time to feeding, compared with nonbreeding bears (Fernández-Gil et al., 2006), we predicted that 2) Asian black bears may trade foraging activities for mating related behavior when together with a conspecific of the opposite sex. Here, we also evaluated these two general aspects of animal mating systems to demonstrate the potential of these new camera systems in studying the mating behavior of elusive species. We discuss limitations and strengths of this novel technology to evaluate the mating behavior of wild bears, and compare our results with those of previous studies.

| Study area
The study was conducted in Okutama Town (35°48′N, 139°5′E), which is approximately 100 km west of the Tokyo Metropolis. This region has heavy rainfall in summer and little snow in winter. In the period 2006-2017, the mean annual precipitation was 1,510 mm and the mean annual temperature was 12.4℃ (range: 0.6℃ in January to 24.2℃ in August) (Japan Meteorological Agency, 2020).
Conifer plantations consisting of Japanese cedar (Cryptomeria japonica), Japanese cypress (Chamaecyparis obtusa), and Japanese larch (Larix kaempferi) cover approximately 50% of the study area, whereas natural forests dominated by Castanea crenata and Quercus serrata in the lower mountain zone [400-500 m above sea level (a.s.l.)], Q. crispula, C. crenata, and Fagus crenata in the middle zone (500-1,500 m a.s.l.), and Abies homolepis and Tsuga diversifolia in the upper zone (1,500-1,800 m a.s.l.) cover about 40% of the area . The remaining area comprises infrastructure and agriculture land.

| Bear capture
Asian black bears were captured from May to June 2018 using barrel traps baited with honey. A total of five traps were set within 5 km of each other, and the trapped bears were immobilized with tiletamine hydrochloride and zolazepam hydrochloride (Virbac, Carros, France, 8 mg/1kg of body weight). Basic body measurements were taken, and premolar extraction was performed for age determination.

| Camera schedules
As Asian black bears in the study area are mostly diurnal , we programmed the collars to record a 10-s video clip of the bear every 15 min, with a duty cycle of 13 hr on (5:00-17:45; total 52 clips per day) and 11 hr off, based on sunset/sunrise times in June and July in the study area. Video recording started at 5:00 on the day after capture. Each camera had a battery life of about 18h, and video data were stored onboard the devices. Since captive bears mostly exhibit mating behavior between June and July (Yamamoto et al., 1998), the battery life and onboard memory capacity of the camera-collar technology were considered sufficient for the collection of frequent observations of mating behavior. The collars were dropped by remote control at the end of the programmed lives of the camera batteries. The collars were then recovered and returned to the manufacturer for downloading the video clips in the mp4 format.

| Behavioral data analysis
To avoid potential effects of handling and attempted remote collar drop-offs, only the data collected between 5:00 on the day after capture and 17:45 on the day before their collars drop-off were used in the analyses. Each video clip was first assessed for the presence of other bears in the recording. However, it is possible that the number of interactions with other bears may be undercounted if the other bear was not directly in front of the camera or the camera view was obscured by hair or soil. To address this uncertainty, any day during which a conspecific was captured on video at least once was considered to be a day in which the bear interacted with other bears.
Using this definition, the clips were classified into two groups: "social days" (i.e., days when the bear interacted with conspecifics) and "solitary days" (i.e., days with no observed interactions). Each video clip was then assigned a specific behavior: "traveling," "sleeping," "resting," "foraging," "sniffing," "others," and "unclear" (7 behaviors in total). For the ethogram used in this study, see Table 1. Clips from social days were further classified according to the behavior of the conspecific.
Whenever possible, videoed conspecifics were assigned a sex and uniquely identified by multiple researchers based on external characteristics such as external genitalia, facial scars, and the shape of the white patch on their breast. Individual recognition followed research, which has been used in brown bear studies (Clapham et al., 2012;Nawaz et al., 2008;Shimozuru et al., 2017).
Most Asian black bears have a characteristic white patch on their chest (e.g., large bib-like mark, small point-like mark, or slim line mark), the size and shape of which is characteristic of the individual. Identification using these chest marks is confirmed to be highly reliable in this species (Higashide et al., 2012). Bears that could not be identified were considered as "unknown individuals," while those for which the sex could not be clearly identified were classified "unknown sex." Bears for which the researchers could not agree on an identification were also considered as unknown individuals. If bear mating was recorded by a male bear's camera, the other bear was assumed to be female. After the behavioral classification, the daily proportion of each behavior was calculated for each social and solitary day and for each bear (Table 2).
We analyzed the video data using Microsoft Movies and TV and Windows Media Player.

| Statistical analysis
Permutational multivariate analysis of variance (PERMANOVA) was used to test whether bears exhibited behavioral changes in response to having a social day or a solitary day. The presence or absence (0, 1) of each behavior per bear was used as response variables in relation to bears having a social day or a solitary day. The Jaccard distance, which is the dissimilarity matrix quantified directly from count data, was calculated to be used in the PERMANOVA based on 1,000 permutations. Significance testing for differences in occurrence of each behavior by social days or solitary days was  (Fisher, 1935). The analysis was conducted using the "adonis" function in the "vegan" package (Oksanen et al., 2020) and "fisher.bintest" function in the "RVAideMemoire" package (Hervé 2020) of the statistical software R 3.5.3 (R Core Team 2020) with a significance level of 0.05 based on the adjusted p-value for multiple comparisons.

| Behavioral differences between social and solitary days
The ratio of social days to solitary days for each bear was as fol-

| Individual mating behavior
Male A interacted with two other unique bears during the observation period ( Table 3). One of those bears was a female (unmarked female 1), with whom male A interacted for two days and exhibited mating behavior on two days (16:30 on 30 May and 9:30 on 2 June).
Male A then interacted with a different bear of which the sex could not be identified (unknown sex 1) from 4 July to 8 July.
Male B interacted with three other bears during the observation period, two of which were female (females 2 and 3) ( Table 3) Female A interacted with two other bears during the observation period, of which one was a male (male 2) ( Table 3). Female A interacted with male 2 during three separate occasions that lasted multiple days, that is, 13 days (17-29 June), and two times for 3 days (6-8 July and 14-16 July).

TA B L E 3 Timelines of social interactions of bears in Okutama,
Japan, during mating season in 2018. "x" are indicated in "solitary days," and the bears that interacted with the collard bear are indicated in "social days" 30 Jun x F2 x M3 1 Jul x F2* x M3 2 Jul x M1***, F2* x M3 3 Jul x F2 x M3 (Continues) Female B interacted with three other bears for observation period, of which two were male (males 3 and 4) ( Table 3). The sex of the third bear could not be identified (unknown sex 3), and this bear was recorded from 8 to 9 June. Female B interacted with male 3 for 12 days (24 June -5 July) and later with male 4 for 5 days (6-10 July).
On 5 July, female B rested on a tree close to male 3 (6:00) before the two moved closer (7:00). After that, female B and male 3 traveled  Brown bears have a polygamous mating system in the broadest sense, as both sexes can associate and mate with various partners during a mating season (Steyaert et al., 2012). Our study shows that in Asian black bears, individuals of both sexes can interact (i.e.,

| D ISCUSS I ON
an indication for mating behavior in an otherwise solitary species) with more than one individual of the opposite sex during the mating season, and we documented multiple mating in males. Those observations suggest that Asian black bears have a polygamous mating system. However, it remains unclear whether females actually mated with the males they associated with, due to the camera orientation on the female's collar and the mating position of bears (males mounting females). Additional observations obtained from animal-borne video systems and from a larger number of individuals of both sexes can advance our knowledge about the mating system and mating strategies in wild Asian black bears.
In such species, male infanticide can be common as male reproductive strategy (i.e., sexually selected infanticide). When litter loss occurs during the mating season, the victimized mothers are released from lactation and can enter estrus and mate shortly (within days) after the loss (Steyaert et al., 2014). Hence, males can create mating opportunities and increase their reproductive success by killing entire litters of unrelated offspring during the mating season Bellemain, Zedrosser, et al., 2006;Steyaert et al., 2014;Swenson et al., 2001). Collecting infanticide cases in wild bears is extremely difficult, and until now, only anecdotical evidence of infanticide in Asian black bears exists. However, our camera collars recorded an adult male bear consuming a cub of the year, which might have been an infanticide victim. Unfortunately, our data cannot provide information about how the male obtained the bear cub because no clips of interactions between the adult male and the cub's mother were recorded. The camera collars can be helpful for detecting potential infanticide cases. Combined with field investigations and genetic tools Bellemain, Zedrosser, et al., 2006), this technology is promising for shedding light on the functional significance of infanticide in Asian black bears.
The energy balance is an essential factor to evaluate the nutritional state of bears (Robbins et al., 2004). The energy balance of Asian black bears has previously been suggested to follow a seasonal, bimodal pattern, dropping from spring to summer and increasing from summer to a peak in autumn (Furusaka et al., 2019). In particular, decreased food consumption by Asian black bears in summer (i.e., June and July) was associated with a large decrease in their energy balance. One factor that has been proposed to explain this is low availability of food resources during this time of the year (Fuchs et al., 2019;Furusaka et al., 2019). However, our results suggest that within this period, bears decreased foraging activities on social days compared with solitary days, which accounted for about half of the days in the mating season in one male and both females. This implies that not only reduced forage availability, but also mating-induced reduced foraging may explain a general decrease in their energy balance during this period of the year.
This decrease in foraging activity concurrent with mating activities has been suggested in brown bears (Fernández-Gil et al., 2006) and may support our observations. In addition, male brown bears

TA B L E 3 (Continued)
exclusive partnerships with specific females (Spady et al., 2007;Steyaert et al., 2012). Such mating behavior was also observed in the present study, suggesting that mate guarding is part of the mating system of Asian black bears. Mate guarding also suggested to constrain feeding activity in male primates (Alberts et al., 1996;Georgiev et al., 2014;Matsubara, 2003). Therefore, several behaviors associated with mating seem to induce decreased foraging activity for male bears during the mating season. Although behavioral changes related to sexual selection are mostly expected to occur in males (Clutton-Brock & McAuliffe, 2009), the same or similar behavioral changes within both sexes (i.e., the decrease in foraging and increase in resting when being social) were observed in the present study. However, our study only involved four focal study units, which implies that the interpretation of our behavioral observations is suggestive and has to be viewed with caution.
Male-male competition for access to receptive females based on the outcomes of aggressive behavior and fights has been shown in brown bears (Steyaert et al., 2020 Therefore, these collars hold promise as a tool to help better understanding the ecology of Asian black bears and other large and elusive mammals.

| CON CLUS ION
Very little information about the mating system of the Asian black bear in the wild currently exists, in part because it is an elusive species, making systematic observations in the wild difficult.
Animal-borne video systems can be a powerful tool to shed light on the behavior of such elusive species. Despite a very low sample size (two males and two females), our study advanced our current knowledge on the mating system of the Asian black bear. First, our observations suggest that wild Asian black bears are polygamous, as all bears interacted with more than one individual of the opposite sex, and we documented multiple mating in males during the mating season. Second, we documented that mating behavior affected activity budgets, with bears spending less time on feeding and more time on resting on social days. Such changes may add to the cost of reproduction the species experiences. In addition, we recorded contest competition among males for females and a potential infanticide case. Infanticide can be an important part of the mating system of polygamous species (sexually selected infanticide) and is extremely difficult to witness in the wild. Improving our knowledge on the mating system of the Asian black bear will contribute to improved management and conservation of the species.

ACK N OWLED G EM ENTS
We thank the member of the Asian black bear team for their assis-