Abstract

Extant amphibians consist of Salientia (frogs), Caudata (salamanders), and Gymnophiona
(caecilians). The mutual relationships of these groups are controversial, with either Batrachia
(Salientia + Caudata) or Procera (Gymnophiona + Caudata) as emerging clades in recent molecular
and morphological analyses. The monophyly of amphibians as a whole is supported by independent
data, but their origins and affinities with early tetrapods are debated. A new cladistic analysis
of early tetrapods retrieves Temnospondyli (the most species-rich group of early tetrapods) as the
closest relatives of crown group amphibians. One temnospondyl group, the Dissorophoidea, forms a
series of consecutive outgroups to crown amphibians. In particular, the Lower Permian amphibamid
Doleserpeton is the most derived plesion on the amphibian stem. The Albanerpetontidae, a group
of salamander-like tetrapods ranging from the Jurassic to the Pliocene, are placed as stem Gymnophiona.
The shortest trees support the Batrachia hypothesis. However, the Procera hypothesis
is not a significantly worse fit for the whole character set. Exhaustive treatment of characters and
taxa is the most appropriate way to disentangle contrasting phylogenetic signals in large matrices.
Tests of different crown topologies show that placement of amphibians within lepospondyls (e.g.
as sister taxon to Lysorophia) is not a significantly worse fit for the whole character set than a
close temnospondyl-lissamphibian relationship. However, the latter phylogenetic hypothesis best
captures the most coherent assembly of derived lissamphibian apomorphies.