期刊文献+

百花山国家级自然保护区褐马鸡栖息地利用分析 被引量:8

Modeling Habitat Factors and Suitability for the Brown Eared Pheasant(Crossoptilon mantchuricum) in Baihuashan National Nature Reserve, Beijing
原文传递
导出
摘要 栖息地破坏是生物多样性丧失的主要原因之一,对濒危物种的生存也构成严重威胁。褐马鸡(Crossoptilon mantchuricum)作为我国特有珍稀濒危物种,目前仅分布于山西、陕西、河北和北京地区。2013年10月至2014年6月,在北京百花山国家级自然保护区使用样线法进行野外调查,使用存在点与伪不存在点数据,结合海拔、坡度、坡向、距道路距离、距居民点距离和植被类型作为变量,运用CART、Random Forest、Tree Net三个机器学习模型对褐马鸡的上述6种栖息地因子重要性排序,结果显示,海拔是影响褐马鸡分布的最主要因子,其次是植被类型、距道路距离和距居民点距离。根据AUC值选择Tree Net模型为最适宜模型,划分褐马鸡的适宜栖息地类型,并计算适宜面积。本研究有助于了解重点栖息因子对北京地区褐马鸡的影响状况,利于褐马鸡保护管理对策的制定。 Habitat destruction is the main reason for the loss of biodiversity, and it is a serious threat to the survival of endangered species. The brown eared pheasant(Crossoptilon mantchuricum) is an endangered species endemic to China. We applied three machine learning models(CART, Random Forest, Tree Net) to analyze the factors influencing the habitat suitability of this bird in Baihuashan National Nature Reserve, Beijing, China. The Baihuashan Mountain belongs to the Taihang Mountain with the altitude mostly between 1 000﹣2 000 meters covered by different types of mountain forest and alpine meadows. During October 2013 to June 2014, we obtained 112 activity locations of the Brown Eared Pheasant, at 22 transect lines with the total length of 55.78 km(Fig. 1). We used six variables(altitude, slope gradient, slope aspect, vegetation type, distance to settlements, distance to roads) in our analyzed models. Our results show that altitude is the leading factor, followed by the factors of vegetation type and distances to the roads and settlements(Table 1). The Tree Net model, with the value of AUC(L/T, 0.960/0.937) being highest, is used to analyze the influencing patterns the six factors(Fig. 2, 3). The size of suitable habitat for the brown eared pheasant in the reserve is 67.98 km^2 of 217.43 km^2 the total size of the reserve according to the Tree Net model, the suitable area for the pheasant is only, only 31.27%. We expect our study can be useful as an example of successful model applying to the habitat management or population manipulation of the endangered species.
出处 《动物学杂志》 CAS CSCD 北大核心 2016年第3期363-372,共10页 Chinese Journal of Zoology
基金 林业公益性行业科研专项(No.201404422) 北京市教育委员会科学研究与研究生培养共建项目(No.2012GJ01)
关键词 百花山国家级自然保护区 褐马鸡 重要栖息因子 物种分布模型 Baihuashan National Nature Reserve Brown eared pheasant(Crossoptilon mantchuricum) Influencing factors Machine learning models
  • 相关文献

参考文献16

二级参考文献231

共引文献243

同被引文献144

引证文献8

二级引证文献36

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部