【摘要】 利用４种计算距离系数的方法，通过聚类分析，将南充市近郊退化灌丛草坡群落划分为７个群丛．结果表明，ＲＥＤ和ＢＣ距离系数通过聚类分析所得结果较符合实际．在作树状聚合图时，为便于直观分析且在图中不出现交叉，采用了Ｃ对策处理更多还原

【Abstract】 Based on data from field survey, a cluster analysis of degraded shrub and grass communities in the suburbs of Nanchong city has been carried out with special reference to 4 kinds of distance coefficients, ie., DE, MED, RED and BC. The results show that the BC distance and RED coefficient are more suitable for clustering than residual parameters. To avoid intersection among different orders in the derived “TREE” diagrams, however, the C strategy is the best. Generally, 7 associations could be obtained by means of cluster analysis.更多还原