LOOSENING THE CLAMP:
AN EXPLORATORY GRAPHICAL APPROACH TO THE CLIMATE LEAF ANALYSIS MULTIVARIATE PROGRAM
ABSTRACT
The Climate Leaf Analysis Multivariate Program (CLAMP) is an established methodology for physiognomic analysis of dicot leaf floras. This paper uses a meta-analysis of four studies that provide CLAMP data on 245 floras from Asia, Africa, and North and South America to demonstrate the application of a new analytical methodology for the exploration of the relationship between leaf morphology and environment. This methodology involves the application of a generalized
"pairs" plot or scatter plot matrix (SPLOM), a form of graphical analysis for multivariate data. It is compared with the results from regression, hierarchical cluster analysis, principle components analysis, and canonical correspondence analysis. Analysis of the available data using pairs plots reveals extensive multiple covariation among explanatory leaf physiognomic variables and identifies sources of systematic error that eigenvector ordination methods tend to conceal. Pairs plots provide a supplementary method for analyzing complex multivariate data on leaf physiognomy and contribute to biological understanding of leaf–environment interactions. Because pairs plots allow more flexible investigation of multivariate data than existing eigenvector and regression-based approaches, they may prove useful not only for analyzing CLAMP data, but also in exploring multiple covariation in other complex paleontological data sets.
W.A. Green. Department of Geology and Geophysics, Yale University, P. O. Box 208109 Yale Station, New Haven, Connecticut 06520
USA
KEY WORDS: leaf physiognomy, Climate Leaf Analysis Multivariate Program (CLAMP), exploratory graphical analysis, pairs plots, Scatter Plot Matrix (SPLOM)
Copyright: Paleontological Society
September
2006
Submission: 10 June 2006 Acceptance: 4 August 2006
