植物生态学报 ›› 2006, Vol. 30 ›› Issue (6): 1030-1039.DOI: 10.17521/cjpe.2006.0132

• 论文 • 上一篇    下一篇

群落结构复杂性的测度方法研究进展

金森()   

  1. 东北林业大学林学院,哈尔滨 150040
  • 收稿日期:2006-03-27 接受日期:2006-06-09 出版日期:2006-03-27 发布日期:2006-11-30
  • 作者简介:E-mail: jinsen2005@126.com
  • 基金资助:
    国家自然科学基金项目(30571508)

A REVIEW ON METHODS FOR MEASURING COMMUNITY STRUCTURAL COMPLEXITY

JIN Sen()   

  1. College of Forest Science, Northeast Forestry University, Harbin 150040, China
  • Received:2006-03-27 Accepted:2006-06-09 Online:2006-03-27 Published:2006-11-30

摘要:

该文对群落结构复杂性的测度方法的研究进展状况进行了综述。根据测度方法建立的方法基础,将现有的方法分成3类:基于多样性的复杂性测度、基于计算复杂性的测度和基于几何学特征的复杂性测度。对每类测度方法进行了介绍,对其优缺点进行了评述。同时提出了未来研究中应给予重视的问题。结果表明,现有群落结构复杂性的测度方法普遍存在区分能力差的问题,对于基于多样性的结构复杂性测度,目前还缺乏确定各测度属性权重的客观方法;现有的一些基于计算复杂性的结构测度与多样性指标关系过于密切,还不完善,同时其生态学的意义还不明确,而另一些计算复杂性指标还缺乏实际检验。今后,如何建立既具有区分力、又与多样性在概念和数值上都有一定区别的群落结构的计算复杂性的测度方法、如何科学合理地确定复杂性测度中的属性权重以及如何建立结构复杂性的测度和功能过程之间的联系等都是需要深入和系统研究的。由于方法的相似性,有关群落结构复杂性的测度方法也可以应用到其它尺度上的结构复杂性的研究中。

关键词: 复杂性, 结构, 计算复杂性, 测度, 科尔莫哥洛夫复杂性, 霍夫曼编码, 生态复杂性

Abstract:

Ecological complexity has received increasing attention in recent years. Structural complexity is one of the most important parts of ecological complexity. This paper reviews the literature on the development of concepts and measures of community structure. Findings indicate that currently used methods can be classified into three groups: measures based on biodiversity, algorithmic complexity and geometrical properties. The many measures based on biodiversity are commonly used. The algorithmic method is new and has not been widely used. It represents the complexity of community structure by the difference of mean length of Huffman code of community attributes and 12th order of Rényi entropy. The methods based on geometrical properties are also commonly used, employing fractal dimension as the most important index. A common problem for all these methods is the difficulty of comparing community complexity with different complexity measures. Also, these methods weakly discriminate among different structures. For the methods based on biodiversity, it is difficult to objectively determine weights of the attributes used. For the algorithmic complexity based methods, the ecological meanings of the measures are still uncertain. Some of them are closely correlated with biodiversity indices, and the others need field testing. Future studies should focus on the following: 1) Determine algorithmic measures with high discriminant ability that differ from biodiversity indices both in concept and values. Because many algorithmic measures have been widely used in fields other than ecology, interesting results could be generated when applying these measures to description of complexity of community structure. 2) Objectively determine weights of attributes with ecological meaning used in complexity measures, a problem that exists in biodiversity and algorithmic measures that use multiple attributes. 3) Connect measures of structural complexity with functions and processes, a critical goal in the study of complexity of community structure.

Key words: Complexity, Structure, Algorithmic complexity, Measure, Kolmogorov complexity, Huffman code, Ecological complexity