Ins. Soc. 41:153-163 (1994) 1015-1621/94/020153-11 $1.50 + 0.20/0 9 1994 Birkh/iuser Verlag, Basel Multivariate morphometries and allometry in a polymorphic ant J. A. F. Diniz-Filho 1, C.J. Von Zuben 2, H. G. Fowler a,, M.N. Schlindwein a and O. C. Bueno 1 1 Departamento de Biologia, Instituto de Bioci~ncias, Universidade Estadual Paulista (UNESP), C.P. 199, 13506-900 Rio Claro, SP, Brazil 2 Departamento de Parasitologia, Instituto de Biologia, Universidade de Campinas (UNICAMP), C.P. 6109, 13081 Campinas, SP, Brazil 3 Departamento de Eeologia, Instituto de BiociOneias, Universidade Estadual Paulista (UNESP), C.P. 199, 13506-900 Rio Claro, SP, Brazil Key words: Caste, ant, morphometrics, principal component analysis, jackknife estimates. Summary In this paper we describe how morphological castes can be distinguished using multivariate statistical methods combined with jackknife estimators of the allometric coefficients. Data from the polymorphic ant, Camponotus rufipes, produced two distinct patterns of allometric variation, and thus two morphological castes. Morphometric analysis distinguished different allometric patterns within the two castes, with overall variability being greater in the major workers. Caste-specific scaling variabilities were associated with the relative importance of first principal component. The static multivariate allometric coefficients for each of 10 measured characters were different between castes, but their relative magnitudes within castes were similar. Multivariate statistical analysis of worker polymorphism in ants is a more complete descriptor of shape variation than, and provides statistical and conceptual advantages over, the standard bivariate techniques commonly used. Introduction Polymorphic worker castes are well represented among the ants (Oster and Wilson, 1978). Wilson (1953) recognized four major categories of worker polymorphism: 1) monophasic allometry, characterized by a nonlinear growth, with the allometric coefficient, k, different from 1.0; 2) diphasic allometry, in which the slope of the plot changes and is interpreted as two regressions crossing at an intermediate point; 3) triphasic allometry, characterized by two breakpoints of the slope; and 4) complete dimorphism, in which there are two very different sizes and morphologies among workers. Crucial to identifying the number of distinct morphologies within a species * Author for correspondence.