Molecules versus morphology: do five genetic loci tell the same story about checkered beetles (Coleoptera: Cleridae) as morphology?

A molecular phylogeny of the checkered beetle (Coleoptera: Cleridae) subfamilies was constructed utilizing mitochondrial ribosomal (16S), mitochondrial protein-coding (COI), nuclear protein-coding (EF1a, wingless) and nuclear ribosomal (28S) genetic loci. A classification recognizing eight subfamilies—Thaneroclerinae, Clerinae, Tillinae, Hydnocerinae, Enopliinae, Epiphloeinae, Korynetinae, and Tarsosteninae—was chosen for labeling the analyzed taxa (Opitz, 2002). 150 ingroup OTUs representing 45 checkered beetle genera were analyzed. Models were fit to molecular data set partitions using the program jModelTest v.0.1.1 (Posada, 2008). The results were compared to previous phylogenetic hypotheses regarding the subfamilies of Cleridae which vary in the number of recognized subfamilies (2-12). 1) The monophyly of the subfamilies will be individually tested using multiple criteria and both parsimony and likelihood-based models and 2) the results of the molecular analysis will be statistically compared to the most current morphology-based hypotheses to check for significant topological differences. To test for significant differences in topologies between the best unconstrained tree and constrained trees (produced using RAxML) we used PAUP* for the parsimony-based Templeton test (Wilcoxon signed-ranks test) (Templeton, 1983) and winning sites (sign) test (Prager & Wilson, 1988), and CONSEL (Shimodaira & Hasegawa, 2001) to do the likelihood-based approximately unbiased test (Shimodaira, 2002), Shimodaira-Hasegawa (Shimodaira & Hasegawa, 1999), and weighted Shimodaira–Hasegawa (Shimodaira, 2002) tests. These results will serve as a preliminary foundation for more detailed future analysis of and within the subfamilies.