Menu

Mitochondrial lineages and DNA barcoding of closely related species in the mayfly genus Ephemerella (Ephemeroptera:Ephemerellidae)

1024 681 Stroud Water Research Center

Alexander, L.C., M. Delion, D.J. Hawthorne, W.O. Lamp, and D.H. Funk. 2009. Journal of the North American Benthological Society 28:584–595.

doi:10.1899/08-150.1

Abstract

We compared genetic lineages in the mayfly genus Ephemerella (Ephemeroptera:Ephemerellidae) identified from mitochondrial DNA (mtDNA) to current taxonomy in 9 morphological taxa, including 2 geographically widespread species, Ephemerella invaria ( = E. inconstans, E. rotunda, E. floripara) and Ephemerella dorothea ( = E. infrequens). Maximum likelihood and parsimony analyses of the mtDNA sequences placed E. inconstans and E. invaria in a well-supported clade; however, mean Kimura 2-parameter genetic distance between the lineages was high (5.2%) relative to distance within lineages (1.3%). The phylogenetic relationships of synonyms E. rotunda and E. floripara were not resolved, but estimates of mean genetic distance to E. invaria were high for both (8.5% and 11.6%, respectively). Populations of E. dorothea were grouped in 2 well-supported clades (12.9% mean divergence) that did not include the synonym E. infrequens (20.9% mean divergence, based on a single sample). A large genetic distance (18.6%) also was found between eastern and western populations of Ephemerella excrucians. Western samples of Ephemerella aurivillii were so genetically distant from all other lineages (32.2%) that doubt about its congeneric status is raised. mtDNA data have been useful for identifying genetic lineages in Ephemerella, but our results do not support use of cytochrome oxidase I (COI) as a DNA barcode to identify species in this genus because we also found evidence of incomplete mtDNA lineage sorting in this gene. Use of the barcoding gene rediscovered some old taxonomic problems in Ephemerella, a result that emphasizes the importance of completing empirical systematic description of species before using single-character systems for identification.