PARENT SESSION
Oral Session - Cross-Scale Landscape Analysis Chair(s): DiBari, John¹, ¹ Western Carolina University, Cullowhee, NC
Wednesday, March 31, 2004 10:00 AM - 12:00 PM Apollo Room 1

Influences of timber harvesting and fuelwood collection on giant panda habitat across a heterogeneous landscape. ^*BEARER, SCOTT L. ¹, JENKINS, CLINTON ¹, HUANG, JINYAN ² and LIU, JIANGUO ¹, ¹ Department of Fisheries and Wildlife, East Lansing, MI, USA² Center for Giant Panda Research and Conservation , Wenchuan County, Sichuan Province, China

ABSTRACT- Timber harvesting and fuelwood collection may influence wildlife habitats by converting parts of a contiguous forest to an earlier successional sere. We assessed the influence of harvesting and fuelwood collection on the giant panda habitats in Wolong Nature Reserve, China (approximately 200,000 ha in size). With field data on panda use (presence/absence), we developed a model of panda occurrence probability using logistic regression. We then applied this probability model to the entire reserve to test two hypotheses: 1) areas harvested for timber and fuelwood have lower proportions of highly suitable habitat than unharvested areas, and 2) highly suitable habitats in harvested areas have higher proportions of fragmentation. Buffers (ranging from 500-3000 m radii, based on panda movement and home range size) were created around each of 913 field sampling plots. We then calculated the characteristics of unsuitable, moderately suitable, and highly suitable habitat patches within each buffer. Results indicate harvested areas are composed of large, contiguous patches of unsuitable habitat, while unharvested areas have a less contiguous patch network of highly suitable habitats surrounded by moderately suitable habitats. Our research demonstrates the effects that timber harvesting and fuelwood collection may have on the wildlife habitat matrix. This will be useful for determining the consequences of logging bans that are being enforced in many countries throughout the world, including China. In addition, we present a method of determining fragmentation effects at different spatial scales by varying buffer size around field sample points.

KEY WORDS: adaptive sampling, Ailuropoda melanoleuca, logging, habitat use