Molecular phylogeny of the Puerto Rican Lepidocyrtus and Pseudosinella (Hexapoda: Collembola), a validation of Yoshii's "color pattern species"

Color pattern was one of the most important characters used to diagnose species in the genus Lepidocyrtus until the introduction of chaetotaxy to Collembola taxonomy. Chaetotaxy confirmed most species diagnoses based only on color patterns, but a number of populations with distinct pigmentation patterns have been found to be identical in all other morphological characters. The absence of individuals showing intermediate color patterns prompted Yoshii to suggest that, despite chaetotaxic identity, populations with distinct color forms represent valid species (implying reproductive isolation and therefore biological species) in what he designated as "color pattern species." In Puerto Rico Lepidocyrtus biphasis, L. dispar, and L. caprilesi show a remarkable variation in pigmentation and as a group include 11 different color forms. Here I present a phylogenetic analysis of the cytochrome oxidase I gene (COI) in 17 species of Lepidocyrtus and Pseudosinella, including 11 species and 10 color forms of Puerto Rican Lepidocyrtus, to test Yoshii's color pattern species concept. The analysis shows large genetic distances between species defined based on morphology alone (morphospecies) and between most color variants within morphospecies. The most often used calibration for the COI molecular clock (2.3% sequence divergence per million years) suggests that morphospecies diverged between 20 and 25 million years before present while color forms within morphospecies diverged between 8 and 19 million years ago. This indicates that changes in climate and sea levels during the Pleistocene were irrelevant to the speciation process in the Puerto Rican Lepidocyrtus. Examination of the genetic variation, phylogenetic relationships, and collection data in light of the biological and phylogenetic species concepts supports the hypothesis that most populations of morphospecies differing only in color pattern are distinct species, thus validating Yoshii's color pattern species concept. As a result it is suggested that morphological characters traditionally used in species diagnoses are very conservative indicators of genetic divergence, that the diversity of springtails has been greatly underestimated, and that studies concerned with identifying factors promoting speciation in Collembola could be misled if they do not include analysis of mitochondrial markers.