The Impact Of Timber Harvest On Wapiti Winter Habitat Selection

This thesis presents a research on wapiti (Cervus elaphus) winter habitat selection in Wan Da Mountains by the methods of "used habitat verse unused habitat" and "used habitat verse available habitat". Based on geographic information system and remote sense technology, the habitat availability was assessed on range scale. The environment factors were classified into 4 types such as topography variables, vegetation variables, human disturbance factors and forest harvest variables.The results revealed that middle and high harvested habitat were heavily used by wapiti and were selected positively, while low harvested habitat was less used and negatively selected. There had been high food abundance and availability in deforested land during 7 years after timber harvest. Clear-cut and thinning were the major timber harvest ways in study area.For vegetation, wapiti bed-site habitat heavily used mixed coniferous and broadleaf stands, cleared area, bare land, brush land, broad-leaved scattered woodland and positively selected the later four type stands. Wapiti forage habitat heavily used mixed coniferous and broadleaf stands, cleared area, broad-leaved scattered woodland brush land, and positively selected the later three type stands. Area with low degree of closure was heavily and positively selected by wapiti as bed-site habitat, and area with low and middle degree of closure were positively selected as forage habitat.As to topography factors, the research results showed that elevation significantly impacted wapiti habitat selection, while slope and aspect were not the significant impact factors in wapiti habitat selection. Wapiti bed-site habitat positively selected area with 200~400m elevation, south and southeast aspect, 10~30° slope, and forage habitat positively selected area with100~300 m elevation, south, southeast and northeast aspect, 10~20 and 20~30 slope.Wapiti bed-site and forage habitat were mainly distributed in the distance range from 600 m to 1 km to open road, and there was a buffer zone around resident area with radius 2.74 km, where no wapiti activity point was observed.Based on elevation, NDVI, vegetation type and the distance to open road, this research established a Logistic model, which could better discriminate wapiti activity habitat. According to this model, this research conducted a calculation of raster layer overlap by GIS, which gave a map on habitat suitability. According to the overlap result, 41.18% area was suitable wapiti habitat in study area; 33.32% area was unsuitable wapiti habitat; 25.51% area was slight suitable wapiti habitat. The suitable habitat was fragmented, with 1596 patches, average area about 5.99 ha.