Using Species Distribution Models to predict habitat distribution for pangolins（Manis pentadactyla pentadactyla）on the Coastal Mountain Range

• Main Authors: Chi-Yan Lai, 賴智恩
• Other Authors: Dr.Kurtis Jai-Chyi Pei
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/66578447761993721692

碩士 === 國立屏東科技大學 === 野生動物保育研究所 === 102 === This article analyzed 276 records of the pangolin (Manis pentadactyla pentadactyla) or its tracks, collected from April 2009 to September 2012 in the southern part (area south of the Provincial Highway 30) of the Coastal Mountain Range (CMR), Taitung County, Taiwan, to establish their habitat distribution model. These records correspond to 193 1-ha grids. Four wildlife distribution models, namely: MAXENT、DK-GARP、OM-GARP and Two-class SVM, were used. The AUC value of the four models were all above 0.7, which indicates the prediction is reliable. This article considered only those grids predicted “present” by at least 2 models as Potential Pangolin Present Grids (PPPGs). Result showed that the total coverage of the PPPGs was 277.6km2 (= 43.2% grids) in the study area; while only 21.8% grids in the Coastal Mountain Range Major Wildlife Habitat were PPPGs that suggested this protected area is not a suitable habitat for the pangolin. Results also showed that pangolins prefer 10-40° slopes, lower elevation (200-1000m), above 125 of high Whole Light Sky Space, high range of NDVI, slopes with western and north-western aspects. There were only 14.1% of grids in the northern part of the CMR were predicted as PPPGs. Overall, PPPGs covered 350.6km2 (= 27.2 %) of the Coastal Mountain Range. However, although the southern part of the CMR is presently the most important pangolins habitat, very few area within it are protected by the law. There should be more conservation actions, such as community-based conservation activities, in the non-protected areas. Also, the northern part of the CMR should provide more suitable habitats for the pangolin if active habitat management for the lower elevation forests can be done.