Phylogeny and Life History Evolution of the Genus Chrysoritis within the Aphnaeini (Lepidoptera: Lycaenidae), Inferred from M itochondrial cytochrome oxidase I Sequences Douglas B. Rand,* ,1 Alan Heath,² Tamara Suderman,* and Naomi E. Pierce* * Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138; and ² Department of Zoology, University of Cape Town, Rondebosch 7700, South Africa Received October 28, 1999; revised March 27, 2000 Phylogenetic relationships among 26 South African species in the tribe Aphnaeini (Lepidoptera: Lycaeni- dae) were inferred from DNA characters of the mito- chondrial gene cytochrome oxidase I (COI), using max- imum-parsimony methods. The resulting phylogenetic estimate supports the systematic hypothesis made by Heath (1997, Metamorphosis, supplement 2), based on morphological characters, that at least three preexist- ing genera (Chrysoritis, Poecilmitis, and Oxychaeta) should be collapsed into the single monophyletic ge- nus Chrysoritis. Two of the species groups described by Heath within Chrysoritis are also monophyletic, while one is paraphyletic and thus unsupported by the molecular data. Strong node support and skewed tran- sition/transversion ratios suggest that two Chrysoritis clades contain synonymous species. Aphytophagy ap- pears as a derived feeding strategy. Evolutionary pat- terns of ant association indicate lability at the level of ant genus, while association with different ant sub- families may have played an ancestral and chemically mediated role in the diversification of South African aphnaeines. © 2000 Academic Press Key Words: taxonomy; Chrysoritis; Lycaenidae; tran- sition/transversion ratios; host plant association; ant association; myrmecophily. INTRODUCTION South African butterflies in the tribe Aphnaeini (Lepidoptera: Lycaenidae) provide a model system for systematic studies and comparative life history analy- sis. The group’s component species occupy several hab- itat types throughout South Africa, covering a wide range of altitudes and plant communities. Like most lycaenids, these aphnaeines exhibit specialized rela- tionships not only with their host plants but also with attendant ants. Most of these ant relationships appear to be mutualistic, with the butterfly larvae offering a nutritious secretion to the workers in return for pro- tection from predators and parasitoids (reviewed in Fiedler, 1991; Pierce, 1987). The larvae feed on host plants from several families, although a few aphnaeine species have relinquished herbivory in favor of exploit- ing ants. These larvae induce trophallaxis from their ant associates, surviving exclusively on mouth-to- mouth feedings (Cottrell, 1984; Pierce, 1995). Phylogenetic analysis offers a means of exploring such complex systems by generating evolutionary hy- potheses inferred from molecular sequences. Phyloge- netic estimates of species divergence can provide im- portant insights into questions regarding taxonomy, biogeography, and coevolution. Heath (1997) revised the taxonomy of the South African Aphnaeini based on morphological characters, primarily of wing patterns and genitalic features. Here, Heath’s systematic hy- potheses are compared with the phylogenetic estimates derived from molecular sequences of the mitochondrial gene cytochrome oxidase I (COI), which has proven effective in resolving insect relationships at taxonomic levels comparable to those within the Aphnaeini (Cro- zier and Crozier, 1993; Brower, 1994a,b, 1996; Pierce and Nash, 1999). The implications of the molecular phylogeny for life history evolution within this tribe are also examined. In his revision of the South African Aphnaeini, Heath (1997) made detailed morphological compari- sons in order to reorganize the tribe’s genus-level tax- onomy. The following conclusions are most relevant to this study: (i) The monotypic genus Argyrocupha was brought within the genus Trimenia, based on genitalic unifor- mity between the two taxa. (ii) Based on similarities of wing pattern, genitalic features, and larval morphology, four genera—Chry- soritis, Poecilmitis, Oxychaeta, and Bowkeria—were 1 To whom correspondence should be addressed at Museum of Comparative Zoology, 26 Oxford St., Cambridge, MA 02138. Fax: (617) 495-5667. E-mail: doug_rand@post.harvard.edu. Molecular Phylogenetics and Evolution Vol. 17, No. 1, October, pp. 85–96, 2000 doi:10.1006/mpev.2000.0820, available online at http://www.idealibrary.com on 1055-7903/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved. 85