Abstract
A comparison is made of two methods of community analysis using the same chalk grassland data. The profound effect of quadrat size on the trend of association detected is demonstrated, and the effect on the association analysis based on summed chi-square values discussed with reference to a joint analysis of the same data using two quadrat sizes. The limitations and advantages of the methods are indicated and it is concluded the covariance analysis is of more value for detailed investigation of small areas whereas the association analysis is of use in the primary survey of an area despite the dependence of the results .on the size of sampling quadrat used.

This publication has 0 references indexed in Scilit: