| An evolutionary hypothesis is presented for the phylogenetic relationships within the Melitaeini (Lepidoptera: Nymphalidae: Nymphalinae), based on cladistic analyses of morphological characters. Melitaeini is revised at the subtribal level, and the concepts of Euphydryiti and Phycioditi but not Melitaeiti are upheld as natural groupings. To create a natural classification scheme, Melitaeiti is restricted to a clade of Eurasian taxa, and three new subtribes are proposed for taxa formerly placed in Melitaeiti: Chlosyniti, Poladryiti, and Gnathotrichiti. Species level phylogenetic analyses are presented for all subtribes except the Melitaeiti and Phycioditi. Euphydryiti is revised at the generic level, and includes the genera Euphydryas Scudder, Hypodryas Higgins, and Eurodryas Higgins, with Occidryas Higgins synonymized with Euphydryas. Gnathotrichiti is revised at the generic level and includes a single monophyletic genus, Gnathotriche Felder and Felder, with Gnathotrusia Higgins placed in synonymy. Poladryiti is revised at the generic level, and includes three monophyletic genera: Atlantea Higgins, Higginsius Hemming, and Poladryas Bauer. Chlosyniti is revised in detail at the generic and species levels, and includes three monophyletic genera: Antillea Higgins, Microtia Bates, and Chlosyne Butler, with the genera Dymasia Higgins, Texola Higgins, Charidryas Scudder, Thessalia Scudder, and Anemaca placed into synonymy. Thirty-seven species taxa and seventeen subspecies taxa are recognized within the Chlosyniti. Detailed camera lucida drawings illustrate all genitalic characters and character states. The phylogenetic analysis of Poladryiti and Chlosyniti is used as a case study to investigate some issues of broader implication to systematic biology, including equally versus successively weighting characters, tree statistics, homoplasy, boot strap scores, and polymorphisms. Separate analyses of genitalic and pattern characters were conducted for the Cholsyniti/Poladryiti data matrix to investigate the proportion of groupings in conflict between independent data sets for different methods of analysis. The percentage of conflicting groupings was 0% for equal weighting of characters with parsimony, 36.8% for successive weighting and parsimony, 77% for a phenetic algorithm (UPGMA), and 100% between pairs of random trees. The absence of incongruence between independent data sets for the former analysis is argued as evidence supporting the effectiveness and theoretical validity of the phylogenetic methods used in this study. |
