WINCHELL, Christopher J.*; MARTIN, Andrew P.; MALLATT, Jon: Phylogeny of Living Elasmobranchs Based on LSU and SSU rRNA-Gene Sequences

For over a century, debate on the higher-level phylogeny of elasmobranchs (sharks, skates, and rays), has focussed on affinities of the morphologically unique members of this group, such as skates and rays (batoids), Chlamydoselachus, Heterodontus, and Squatina. We compared the large-subunit (28S and 5.8S) and small-subunit (18S) ribosomal RNA genes of 22 elasmobranchs, 2 chimeras, and 2 bony fishes. We sequenced nearly all of these genes. We used maximum likelihood (ML), maximum parsimony (MP), and minimum evolution (ME) optimality criteria for tree building and non-parametric bootstrapping. Our best ML tree placed batoids as the sister group to sharks, found a monophyletic clade of orbitostylic sharks (i.e., Squatina was a squalomorph shark), showed galeomorph monophyly (except that Heterodontus was not included), and favored a lamniform + orectolobiform clade. However, some of these relationships were only weakly supported by non-parametric bootstrapping. With parametric bootstrapping, we rejected Shirai’s (1992, 1996) hypnosqualean hypothesis, Compagno’s (1973, 1977) hypothesis of four unresolved elasmobranch clades, and several other dominant hypotheses. We also estimated elasmobranch phylogeny under a Bayesian framework. A consensus of trees from the Bayesian posterior probability distribution placed Heterodontus as the sister to a batoid + galeoid group, and left the position of Chlamydoselachus essentially unresolved. Although this topology differed from our best ML tree, the two trees were indistinguishable by parametric bootstrapping. Our most significant finding is that the rRNA genes do not relate batoids to squalomorph sharks.