PROC. ENTOMOL. SOC. WASH. 90(2). 1988. pp. 133-154 COMPAR.\TIVE MORPHOLOGY OF THE BUTTERFLY FORELEG COXA AND TROCHANTER (LEPIDOPTERA) AND ITS SYSTEMATIC IMPLICATIONS Robert K. Robbins Department of Entomology. NHB STOP 127, National Museum of Natural History, Smithsonian Institution. Washington. D.C. 20560. Abstract.— I describe and illustrate five qualitatively distinct forms of butterfly foreleg coxa, trochanter, and basal femur, and give the distribution of each type by sex for the butterfly families. 1 code this variation in a character matrix of four characters with 9 character states, from which I derive a most parsimonious cladogram with four mono-phyletic groups: ( 1 ) Styginae (Riodinidae). (2) Hamearinae (in part) + Styginae + Riodini-nae + Euselasiinae (Riodinidae), (3) Lipteninae + Poritinae + Liphyrinae + Miletinae + Curetinae (Lycaenidae). and (4) Riodinidae + Libytheidae + Nymphalidae. The second and third groups have not been recognized previously as monophyletic. The fourth sup-ports previous results based on other leg characters, but is inconsistent with most published phylogenies to the butterfly families. Contrary to previous reports, the forecoxa of male Styx infernalis (Riodinidae: Styginae) extends beyond the trochanter, but this extension is smaller than in most other riodinids. 1 also show that the male forelegs of Curetis (Lycaenidae: Curetinae) and Riodinidae are qualitatively diflferent. a result that does not support the hypothesis that these two taxa are sister groups. Key Words: leg characters, butterfly, cladogram For more than 125 years, classification of male forecoxa does not extend spinelike be-the butterfly families has relied heavily on yond the trochanter and because it diflers foreleg characters, particularly those of the from riodinids in a few other structures, tarsus (Bates 1861, Ford 1945), but mor-Scott (1985) proposed that Ci//ff/5 (a genus phology of the male foreleg coxa and tro-that Ehrlich had considered to be a lycaenid) chanter has also been used in butterfly higher and Riodinidae (his Riodininae without classification (cf Horror et al. 1981 for an Styx) are sister groups because both have introduction to insect leg morphology), the male foreleg coxa extending beyond the Godman and Salvin (1879-1901) discov-trochanter. ered that the male forecoxa of riodinids ex-The few published figures of foreleg coxae tends beyond its articulation with the tro-and trochanters lack detail (e.g. Ehrlich chanter, and Stichel (1910-1911) and 1958a. b. Scott 1986). and I propose to solve Ehrlich ( 1958b) characterized the Riodini-this problem with the use of a scanning elec-dae (Ehrlich's Riodininae). in part, by this tron microscope (SEM). It is clearly impor-structure. Ehrlich also erected a new mono-tant that the morphology of these structures basic"subfamily" — of rank equal to the Ly-be well documented if they are to be used caenidae (his Lycaeninae) and Riodinidae— in constructing familial classifications of the for Styx infernalis Staudinger because its butterflies. The first purpose of this paper
BioStor
