庞泉沟国家自然保护区森林群落木本植物种间关系的分析

Interspecific Relationships Among Woody Plants of Forest Communities in Pangquangou National Nature Reserve at Mt.Guandi,Shanxi,China

应用基于二元数据的χ2检验、方差比率法和基于数量数据的Pearson相关系数检验及Spearman秩相关系数检验等数量方法,研究了庞泉沟国家级自然保护区(东经111°22′-111°33,′北纬37°45′-37°55′)森林群落(海拔1 600-2 430 m)21个优势种的种间关系。结果表明:研究区域内森林群落优势种群的种间关系较为松散(χ20.95N<W<χ20.05N),种的分布相对独立,森林群落总体处于稳定阶段;但局部地段优势种群间存在着一定程度的竞争。位于山体中下部的群丛组Ⅰ、群丛组Ⅱ的总体关联性存在一定程度的负关联,但未达到显著水平;位于山体上部的群丛组Ⅲ的总体关联性为无关联,种间关系较为松散。随着海拔的升高,三个群丛组的联结指数(VR)呈上升趋势,χ2检验显著率和Spearman秩相关检验的正负关联比有所增加,负显著率都明显下降,正显著率有升高的趋势,说明随着海拔的升高,种间关系表现为无关联,森林群落也趋于稳定。应用方差比率法与2χ检验、Pearson相关系数检验和Spearman秩相关系数检验较为清晰地揭示了庞泉沟自然保护区森林群落的种间关系。这四种方法可以互相弥补彼此的不足,结合使用效果更好。本研究结果支持随着群落演替进程的发展,群落结构及其种类组成将逐渐趋于完善和稳定,群落种群总体种间关系也将向着无关联发展的观点。

Based upon data obtained from field investigation, the interspecific relationships among 21 dominant woody species of forest communities in Pangquangou National Nature Reserve at Mr. Guandi were analyzed by using χ^2-test for 2 × 2 contingency table, variance ratio （VR） test, Pearson＇ s correlation coefficient test, and Spearman ＇ s rank correlation coefficient test. The results clearly show that the overall association of forest communities among all species tends to no correlation （χ^20.95N 〈 W 〈 χ^20.05N ） and the species distributions are relatively independent, however, there are some competitions among dominant species occurring at partial elevation. The overall associations of group Ⅰ and Ⅱ located at lower elevation （ 1 600 - 2 200 m） are some negative correlation and they are not significant （p 〉 0. 05）, but the index of group Ⅲ located at elevation （2 200 -2 430 m） is no correlation. With the altitudes increasing, the overall association indexes （ VR）, the ratio of the positive and negative association and the positive significant percentage of χ^2 test and Spearman rank correlation coefficients are all rising, while negative significant decreased. This means that interspecific relationships tend to null and forest communities will be more stable with the rising of elevation from 1 600 m to 2 430 m. The results suggest that it is better to investigate the interspecific correlation by combining variance ratio （VR） with test χ^2- test, Pearson＇s correlation coefficient test and Spearman＇s rank correlation coefficient test. They can make up the defects of each other＇ s. The results are in accordance with the idea that interspecific association tends to be null with the development of community succession, that is, when the forest community gets to the climax, they will have stable structure and species distribution are relative independent, moreover the interspecific association tends to be no correlation.