Assessing the biotic distance between groups of biological populations by means of replicated counts

Abstract

The problem of assessing the difference in species composition between groups of biological populations is considered when the vectors of relative abundances are estimated through replicated counts. A Monte Carlo study was employed to investigate the performance of classical MANOVA and Smith's permutation test. The use of MANOVA represents a promising strategy provided that an adequate number of populations is assigned to those groups in which a greater variability of abundance estimates arises. Copyright © 2002 John Wiley & Sons, Ltd.