Search

Abstract Background Rapidly expanding human activities have profoundly changed the habitat use of both large carnivores and their prey, but whether and how human activities affect the interactions between them has received relatively less attention. In this study, we conducted a systematically designed camera-trapping survey on an endangered large carnivore (North Chinese leopard Panthera pardus japonensis) and its wild ungulate prey (Siberian roe deer Capreolus pygargus and wild boar Sus scrofa) in the Taihang Mountains of central North China. Using conditional two-species occupancy model based on data derived from the extensive sampling effort (15,654 camera-days at 102 camera sites), we examined the relationship of spatial use between leopards and each prey species under the effects of human presence, free-ranging cattle, roads and settlements. Results Humans and cattle had contrasting effects on the relationship of spatial use between leopard and roe deer, with higher and lower spatial segregation between them at human and cattle-frequented sites, respectively. Roads might create a shelter for wild boar from leopard predation, with less spatial segregation between them at sites close to the roads. Conclusions Our findings demonstrate that human activities are reshaping the spatial overlap between large carnivores and their prey, and have non-equivalent effects among different types of human activity. Such effects may further alter the strength of interspecific interactions between predator and prey, with far-reaching influences on the community and ecosystem that require more research.

Large mammals can perceive humans as predators and therefore adjust their behavior to achieve coexistence with humans. However, lack of research at sites with low hunting intensity limits our understanding of how behavioral responses of animals adapt to different predation risks by humans. At Heshun County in North China, where hunting has been banned for over three decades and only low-intensity poaching exists, we exposed two large ungulates (Siberian roe deer Capreolus pygarus and wild boar Sus scrofa) to the sounds of humans, an extant predator (leopard Panthera pardus) and a control (wind), and examined their flight responses and detection probabilities when hearing different type of sounds. Both species showed higher flight probabilities when hearing human vocalization than wind, and wild boar were even more likely to flee upon hearing human vocalization than leopard roar, suggesting the behavioral response to humans can equal or exceed that of large carnivores in these two ungulates even in an area without hunting practices. Recorded sounds had no effect on detection probability of both ungulates. Additionally, with repeated exposure to sounds, regardless of treatment, roe deer were less likely to flee and wild boars were more likely to be detected, indicating a habituation-type response to sound stimuli. We speculate that the immediate flight behavior rather than shifts in habitat use of the two species reflect the low hunting/poaching pressure at our study site and suggest further examination of physiological status and demographic dynamics of the study species to understand human influence on their long-term persistence.

Assessing a species’ threatened status and then developing specific conservation strategies accordingly rely heavily on knowing that species’ complete and accurate spatial distribution. In this study, we used the Asiatic black bear (Ursus thibetanus) in China to represent a large threatened species for which distribution information is limited and spatially biased. We grouped the two main sources of black bear occurrence data into two different resolutions: (1) coarse resolution data corresponding specific management units (e.g., nature reserves) that cover large areas, and (2) fine resolution data composed of longitude and latitude records that were bias in their geographic range. Our distribution mapping approach integrated those two data types to examine black bear spatial patterns across the country. We used both presence and absence data in the Random Forest algorithm to predict black bear distribution at coarse (30 km) and fine (3 km) resolutions, and then refined the coarse-scale prediction with the fine-scale prediction using a map fusion technique based on the Bayes theorem. We thus generated an integrated high-resolution range map that was both more accurate than the coarse-scale map and more representative of the black bear geographic range than was the fine-scale map. Our results showed that the total estimated range of Asiatic black bears in China was 462.3 × 103 km2, 77.50% less than the most recent IUCN range map and 70.90% less than the area of habitat (AOH) estimation. Using those results, we identified two island and six mainland management units in China and, based on the predicted habitat conditions, proposed specific conservation strategies for each unit. Our study results provide practical knowledge and pragmatic guidance for future conservation planning and action for this species, and our framework provides an example and template for range estimations of species with similar types of occurrence records.