Skip to main content Skip to main navigation menu Skip to site footer
Type: Article
Published: 2019-07-15
Page range: 193–203
Abstract views: 205
PDF downloaded: 215

Phylogenetics of a small caddisfly genus (Thremmatidae: Oligophlebodes): comparison among multiple hypotheses from DNA barcode data

Dana Weaver, Biology Department, William Paterson University, 300 Pompton Rd., Wayne, NJ 07470
Joseph C. Spagna, Biology Department, William Paterson University, 300 Pompton Rd., Wayne, NJ 07470
Department of Environmental Science, Policy, and Management, University of California, Berkeley, CA 94720


Barcoding datasets can serve as a resource for species associations and delineations, but single-gene trees estimated by distance methods do not provide strong estimates of phylogeny. Using DNA data from the Barcode of Life Database (BOLD), we calculate a phylogenetic tree for Oligophlebodes (Trichoptera: Thremmatidae), a small genus of caddisfly endemic to the Western United States. Here we estimate a preliminary phylogeny for Oligophlebodes using Bayesian likelihood, and compare it to trees produced by distance and standard likelihood methods. Using the barcode region of the cytochrome oxidase 1 (COI) gene, we analyzed 44 individuals representing five species (and 2 unknowns) and a sister-genus outgroup (Neophylax) from locations ranging from Southern New Mexico northwest into British Columbia. Partitioned Bayesian likelihood analysis under the F81 (1st codon positions) and HKY80 + I + Γ (for both the 2nd and 3rd codon positions) model gave the consensus topology (Neophylax toshi, (O. sierra, (O. ruthae inc. spc. 1 & 2, O. sigma, (spc. 3 & 4, (O. ardis, O. minutus)))). Species identifications were supported by monophyly of most species-level taxa. However, confirmation of species identifications of unknowns was complicated by incomplete taxon sampling for spc. 1 & 2. Placement of spc. 3 & 4 may serve as support for taxonomic review of O. minutus. Compared with an existing published phylogeny of Oligophlebodes BOLD sequences constructed under RAxML, the Bayesian hypothesis had higher resolution at the basal node of Oligophlebodes. Because of their support values, both likelihood trees are recommended over the BOLD TaxonID tree (an unrooted neighbor joining tree using the Kimura 2-parameter model). The novel topology produced in the Bayesian tree supports further explorations by likelihood-based methods, including partitioned analyses, of our preliminary Oligophlebodes dataset that can be used as additional lines of evidence to support morphological work.