A study on the community analysis using three‐factorial data (time x site x species)

Abstract

Summary This study is to demonstrate that the analysis of biological communities using scaling unstandardized squared Euclidean distance is extremely effective to deal with data which are quite generally collected by 3 factors (time x site x species). The main part of this study has been explained in detail byWilliams andStephenson (1973). We have conducted some discussion introducing mathematic complementary explanation as well as expansive interpretation. The dissimilarity measure based on squared Euclidean distance is used to classify clusters from a dendrogram obtained by cluster analysis. The outline of the community structure can be known by comparing the values of mean variance per comparison and interaction. The table of mutual comparison among clusters introduces dynamic pattern expression for community. Contributions by species to time and site are capable of expressing a concrete role of the species. We apply the technique above to the demersal fish community in Shijiki Bay, Hirado Island, Nagasaki Pref. Number of individuals caught with Gochi trawls at 11 fixed stations in 1975–1984 are used as material for analysis. 84 species is accounted.