生物多样性 ›› 2016, Vol. 24 ›› Issue (1): 20-29.  DOI: 10.17520/biods.2015177

• 研究报告: 动物多样性 • 上一篇    下一篇

小兴安岭南麓马鹿冬季适宜生境评价

吴文1,2, 李月辉1,*(), 胡远满1, 陈龙1,2, 李悦3, 李泽鸣1, 聂志文1,2, 陈探1,2   

  1. 1 中国科学院沈阳应用生态研究所森林与土壤生态国家重点实验室, 沈阳 110016
    2 中国科学院大学, 北京 100049
    3 辽宁大学环境学院, 沈阳 110036;
  • 收稿日期:2015-06-22 接受日期:2015-10-26 出版日期:2016-01-20 发布日期:2016-06-12
  • 通讯作者: 李月辉
  • 基金资助:
    基金项目: 国家自然科学基金(41271201, 41371198和41201185)

Suitable winter habitat for Cervus elaphus on the southern slope of the Lesser Xing’an Mountains

Wen Wu1,2, Yuehui Li1,*(), Yuanman Hu1, Long Chen1,2, Yue Li3, Zeming Li1, Zhiwen Nie1,2, Tan Chen1,2   

  1. 1 State Key Laboratory of Forest and Soil Ecology, Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016
    2 University of Chinese Academy of Sciences, Beijing 100049
    3 College of Environmental Sciences, Liaoning University, Shenyang 110036
  • Received:2015-06-22 Accepted:2015-10-26 Online:2016-01-20 Published:2016-06-12

摘要:

通过2014-2015年两次冬季野外调查, 将收集的79处马鹿(Cervus elaphus)出现信息作为分布点数据, 选取地形、景观类型、植被特征和人类干扰4类19种因子作为环境变量, 利用最大熵(maximum entropy, MaxEnt)模型, 分析了小兴安岭铁力林业局辖区马鹿种群冬季潜在适宜生境分布特征和主要环境因子对马鹿分布的影响。结果显示: 模型预测精度较高, 训练集与验证集的平均AUC(area under the curve, 受试工作者曲线下面积)值分别为0.949和0.958; Jackknife检验结果表明: 景观类型因子对马鹿生境选择的影响较大; 坡向、距大路距离、距混交林距离、距灌草地距离和距农田距离是影响马鹿生境分布的主要环境因子, 其综合贡献值依次为27.8%、23.9%、19.5%、15.3%和10.4%; 距小路距离对马鹿分布影响较小。我们依据MaxEnt模型最大约登指数, 找到最佳中断点0.22作为阈值将马鹿冬季栖息地划分为适宜和不适宜两个等级, 其面积分别为663.49 km2和1,378.85 km2, 分别占研究区总面积的32%和68%。马鹿的适宜生境主要分布在铁力林业局辖区的东部山地和中部林地等区域; 南部地区接近铁力市区, 人类活动频繁, 不适宜马鹿栖息。对马鹿种群的保护管理措施提出3点建议: 控制人为干扰; 构建多样性景观; 优先保护马鹿的潜在适宜生境分布区。

关键词: 物种分布模型, 最大熵模型, 马鹿, 生境适宜性, 人类干扰, 小兴安岭

Abstract

Abstract: In the Tieli Forestry Bureau of the Lesser Xing’an Mountains in Northeast China, we recorded a total of 79 GPS coordinates of presence traces for red deer (Cervus elaphus) during winter field investigations in November 2014 and January 2015. We grouped 19 environmental predictor variables into four classes including terrain, landscape type, vegetation structure and human disturbance, we used the MaxEnt model to predict potential suitable winter habitat distribution for red deer and determine the contribution of each environment factor to the distribution characteristics. The accuracy of our prediction models was accessed by the area under the curve (AUC) values of a receiver operating characteristic (ROC) curve, and validation showed that the results had high average AUCs: 0.949 for training data and 0.958 for testing data. The results of a Jackknife test indicated that landscape type had the strongest influence on winter habitat suitability of red deer among the 4 class variables. Aspect, distance to road, distance to mixed forest, distance to bush-grass land, and distance to farmland appeared to be the most important parameters with contributed to the variance 27.8%, 23.9%, 19.5%, 15.3%, and 10.4%, respectively. Distance to path had an insignificant influence on red deer distribution. The Youden Index was applied to determine the threshold value for habitat classification in the MaxEnt modeling. The simulated habitat was divided into two classes including suitable (≥ threshold value) and unsuitable (< threshold value) habitat using a threshold value of 0.22. Our results showed that the area of suitable habitat was 663.49 km2 and the unsuitable habitat was 1,378.85 km2, accounting for 32% and 68% of the total area of Tieli Forestry Bureau, respectively. Areas with a high suitability of deer habitat were located mostly in the eastern mountainous region and the central forest area. The southern plain area was not suitable habitat for red deer because they were in close proximity to Tieli City. Results from our study suggest forest managers need to control human disturbance, build a diversified landscape, and protect predicted suitable habitat areas in order to better preserve the deer population.

Key words: species distribution models (SDMs), maximum entropy model, Cervus elaphus;, habitat suitability, human disturbance;, Lesser Xing’an Mountains