• • 上一篇    下一篇

固定核空间法和最小凸多边形法估计藏狐家域的比较

刘晓庆,王小明,王正寰,刘群秀,马波   

  1. (华东师范大学生命科学学院,上海 200062
  • 出版日期:2010-05-27 发布日期:2010-03-02

Comparison of the Tibetan fox (Vulpes ferrilata) home range size using methods the fixed kernel estimation and the minimum convex polygon

LIU Xiaoqing,WANG Xiaoming,WANG Zhenghuan,LIU Qunxiu,MA Bo   

  1. School of Life Sciences,East China Normal University,Shanghai 200062,China
  • Online:2010-05-27 Published:2010-03-02

摘要: 最小凸多边形法(MCP)和固定核空间法(FKE)是目前最常用的家域计算方法,但受空间自相关性、偏远位点等问题的影响,两种方法均存在明显的局限性。本文根据2006 年和2007 年在四川省石渠县和青海省都兰县的7 只藏狐352 个活动位点数据,分析MCP 和FKE 家域估计的效果和存在的问题。结果显示:(1)利用概率百分比≤95% 时,MCP 计算结果和FKE 没有显著差异;(2)极端点对高百分比(85% ~ 100% ) 下MCP影响显著,而FKE 对极端点影响控制较好;(3)FKE 家域外形复杂,计算结果受平滑度系数设置影响显著。因此,研究领域行为时,应同时使用FKE 和95% MCP 两种方法。当数据分布较理想时,FKE 能够给出更为准确的面积估计,而MCP 则因其通用性,使得研究数据与其他研究的结果更具可比性。

关键词: 藏狐, 家域, 利用概率百分比, 最小凸多边形法, 固定核空间法

Abstract: The minimum convex polygon (MCP)and the fixed kernel estimator (FKE)are the two methods most used for
home range estimation. However,because of the problems like spatial data dependence and extreme data points,the usage of these two methods is limited. In this paper,we attempted to analyze and compare the results from these two methods, and discuss how to avoid the disadvantages of each method to make the home range estimation more accurate. We documented 352 locations of 7 adult Tibetan foxes (Vulpes ferrilata) in Shiqu County,Sichuan Province and Dulan County,Qinghai Province during 2006 and 2007. Both MCP and FKE were used to calculate the home range. We found:(1)when the utility probability percentage was set ≤95% ,the difference of the home range size calculated by these two methods had no significant difference;(2)although FKE was more robust than MCP,extreme data points influenced the calculation of both the two methods in higher utilization probability percentage (i. e.,85% - 100% ); (3)home range size calculated by FKE was influenced significantly by the setting of smoothing parameter h which could be determined arbitrarily and the least squares cross validation did not always provide the best evaluation of h. We recommended that both of the FKE and 95% MCP should be used in the same home range study. FKE can be the better home range estimator when the autocorrelation of data spatial distribution is not significant. However,95% MCP can be the only choice especially when the comparison of the results from different telemetry studies is needed.

Key words: Fixed kernel estimation, Home range, Minimum convex polygon, Tibetan fox (Vulpes ferrilata), Utility probability percentage