Search

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.

Livestock grazing has become the most prevalent human disturbance in protected areas across the range of giant panda (Ailuropoda melanoleuca). Previous studies have documented the impacts of livestock grazing on habitat and food resources of giant panda, however, little is known about how free-ranging livestock influences other sympatric species. In this study, we investigated the presence of livestock on the habitat use and activity patterns of blood pheasant (Ithaginis cruentusbe), a representative species of phasianids in China, of which more than 50% are under threats due to habitat loss and fragmentation, illegal harvest and human disturbance. We combined camera-trap and sign survey data collected within Wanglang National Nature Reserve, where the livestock population has increased by nine-fold in the past decade, and used both an occupancy modeling framework and kernel density estimation to understand blood pheasant’s recent distribution, changes in occurrences, and diel activity patterns in the presence of free-ranging livestock in different seasons (i.e., breeding [April to July], non-breeding [August to October] and winter [November to March]). For diel activity, blood pheasant overlapped highly with livestock regardless of the season, as both are primarily diurnal species. With regards to distribution, we detected a significant positive correlation between the presence of blood pheasants and livestock (P = 0.02) in breeding season which was against our expectation; no significant relationship in non-breeding season; and a significant negative correlation in winter. Besides, livestock had a significant negative correlation with blood pheasant extinction probability, and no significant impact on colonization probability. Aided by results from two-season occupancy modeling, we argued that the spatial co-occurrence in breeding season wasn’t due to causality, but similar habitat preferences between them. Our study confirms that free-ranging livestock do have the potential to impact blood pheasant due to temporal and spatial overlap, but the consequences of this overlap were unclear.