DIFFERENTIAL VARIATION OF MALES AND FEMALES IN POLISTES EXCLAMANS.

Abstract

The purpose of this study is to provide a biometrical approach to the question whether male Hymenoptera, produced from unfertilized eggs, are more or less variable in external morphological traits than the bisexually produced females. Polistes exclamans Viereck was chosen because sizable sibling groups (nest populations) including both sexes were readily available, the sexual dimorphism is not as marked as in some Hymenoptera, and it was possible to obtain male and female siblings which underwent development under comparable environmental conditions (i.e., at the same season of the year, and in the same size pupal cell). Unlike some other Polistes species, P. exclamans has only one foundress queen per spring nest. Unfortunately most of the early work that treated this question preceded modern genetics and hence was conducted with an erroneous theoretical background. Some of the authors (Casteel and Phillips, 1903; Kellogg and Bell, 1904; Kellogg, 1906) cite Weismann's theory of amphimixis as predicting that sexually produced individuals (females) should be more variable than parthenogenetically produced males. The collapse of this argument in the wake of the development of genetics was apparently responsible for the subsequent lack of interest in the problem: after the rash of papers early in the century, and Phillips' (1929) second study, no studies of the problem have appeared. Now the theoretical viewpoint has changed completely; it seems logical to argue that males, since there are no heterozygotes for any given locus, should be more variable than females, that is, more extremes should appear in the male population. Accordingly, in the section entitled Theoretical Genotypic Variances, following, theoretical expectations are developed which predict that the genotypic variance of haploid males should be greater than that of their diploid sisters. However, it is also possible to argue that the hemizygosity of the males would provide a lower dosage of a given gene than that of a homozygous female, so that the total variation of males and females might be approximately equal; or genes whose effects are limited to the female sex (Kerr, 1967) might be common in the population, which would lead to a similar result. The bearing of the results of this study on these possibilities is treated in the Discussion section; however, it is apparent that the answer to this problem has not lost its pertinence to the genetics and evolution of the Hymenoptera, or of arrhenotoky in general. The early papers also suffered from another major disadvantage in that they antedated the growth of statistics, and particularly multivariate statistics, which is particularly well suited to attack such a problem. Nevertheless their results are interesting and are briefly reviewed. Casteel and Phillips (1903) studied the comparative variability of the wings and appendages of worker and drone Apis mellifera L. They concluded that the drones were the more variable, both in linear measurements and in the occurrence of a much greater number of wing-vein abnormalities. In both this study and the more extensive one by Phillips (1929), the greater variability of the males is attrib1 Contribution number 1406 from the Department of Entomology, University of Kansas, Lawrence, Kansas. Done in partial fulfillment of the requirements for the degree of Master of Arts in Entomology. 2 Present address: Department of Entomology and Limnology, Cornell University, Ithaca, New York 14850.