Allopatry and parapatry in hairstreak butterflies (Lycaenidae, Eumaeini) with uniform male sexual structures

Abstract

An unusual characteristic of the Electrostrymon guzanta species complex is that its male primary and secondary sexual structures are uniform. Eumaeine taxonomy is often based on variation of these structures, and the taxonomy of this group is unresolved. Evolutionary changes in sexual structures are often hypothesized to promote reproductive isolation among closely related sympatric species, and the incidence of sympatry in this complex would be predicted to be low. To resolve the taxonomy and to assess the sympatry prediction, individuals of this species complex were partitioned into wing phenotypes based upon 21 qualitative characters, which were coded. We also measured 6 quantitative characters, which were analyzed with logistic regression. To determine the distinctiveness of phenotypes, the coded characters were analyzed with maximum parsimony because it provides the most efficient summary of the data. It also provides a hypothesis of character state evolution. Based on these results, the taxonomy of the E. guzanta species complex, including its nomenclature, was updated. Electrostrymon thurman Thompson & Robbins, new species, is named for high elevation individuals of the species complex from Costa Rica and western Panama (type locality). Chimeric holotypes of Kisutam simplisis K. Johnson & Kroenlein, 1993 and Angulopis strymonotis K. Johnson & Kroenlein, 1993 are restricted under ICZN article 73.1.5, and the latter name is synonymized with Electrostrymon joya (Dognin, 1895), new synonym. As predicted, the incidence of sympatry in the E. guzanta species complex is low. Three of the four species in the complex-E. guzanta, E. thurman, and E. perisus-are allopatric. The distribution of the fourth member of the species complex, E. denarius, overlaps that of E. guzanta and that of E. thurman, but it occurs at lower elevations (statistically significant). The theory of hybridization parapatry proposes areas of distributional overlap where there is reproductive interference. Data on reproductive interference are yet needed to determine whether this distributional overlap accords with the prediction of parapatry.