Molluscan populations of the Eden Estuary, Fife, and the use of numerical taxonomy methods to determine their distribution patterns

Abstract

The aim of this study is twofold, firstly to recognize life and death assemblages among subfossils by using standard criteria and secondly to test the applicability of Numerical Taxonomy methods in defining biotopes and biofacies in Palaeoecclogy. Among the criteria applied in distinguishing life and death assemblages, the following: dispersion of subfossils, faunal composition and diversity, population density, surface condition of shells and ratio of opposite valves, indicate that little postmortem transportation has taken place and that the subfossil assemblages represent accurately the life assemblages from which they were derived. Other criteria were either not applicable or when applied proved unreliable. In examining biotopes and biofacies, several sets of data are combined in alternative techniques. For biofacies the most ecologically meaningful results are obtained by the correlation coefficient with numerical data or by the Matching coefficient with binary data. Most cluster methods produce similar results with the above coefficients. For biotopes analysis, both similarity coefficient and cluster method are critical to the resultant classification. The most successful distribution patterns are derived by use of the Matching coefficient and W.P.G.A. with binary data or Squared Euclidean Distance and Ward's method with standardized-by-species numerical data. It is suggested that the grain size of the substrate is the main factor determining the distribution of Mollusca in the Eden estuary, since the "textural facies" defined by Eastwood (1977) on the basis of the grain-size distribution in the Eden, are consistent with the biotopes delineated in this study.