Lungfish neural characters and their bearing on sarcopterygian phylogeny

Abstract

The phylogenetic affinity of lungfishes has been disputed since their discovery, and they have variously been considered the sister group of actinistians, the sister group of amphibians, or equally related to actinopterygians and crossopterygians. Previous discussions of these hypotheses have considered neural characters, but there has been no general survey of the nervous systems of sarcopterygians that examines the bearing of neural characters on these hypotheses in the context of a cladistic analysis. Such a survey of representatives of all living sarcopterygian groups reveals at least twenty-three characters that are possible apomorphies at some hierarchical level among sarcopterygians. Neural synapomorphies corroborate the phylogenetic hypotheses that actinistians, amphibians, and dipnoans are each monophyletic taxa. The hypothesis that Latimeria is the sister group of amphibians is the least corroborated, as only a single possible synapomorphy, presence of cervical and lumbar enlargements of the spinal cord, supports this hypothesis. The hypothesis that lungfishes are the sister group of amphibians is supported by two possible synapomorphies: loss of a saccus vasculosus and the presence of neurocranial endolymphatic sacs. The hypothesis that actinistians are the sister group of lungfishes is the most corroborated, based on five possible synapomorphies: presence of a superficial isthmal nucleus, a laminated dorsal thalamus with marked protrusion into the third ventricle, olfactory peduncles, evaginated cerebral hemispheres with pronounced septum ependymale, and electroreceptive rostral organs. However, all five characters may be plesiomorphic for bony fishes. The nervous systems of Latimeria and Neoceratodus are very similar to each other, as are the nervous systems of lepidosirenid lungfishes, caecilians, and salamanders. If Neoceratodus is the most plesiomorphic species of living lungfishes, then lepidosirenid apomorphies may have arisen by paedomorphosis. Our inability to examine the neural characters of a relevant outgroup (rhipidistians) may result in many sarcopterygian plesiomorphic characters being interpreted as apomorphic characters, due to the wide distribution of paedomorphic characters among living sarcopterygians and their possible resemblance to plesiomorphic characters present in living outgroups that can be examined.