Abstract:
I studied the development and homology of the neural complex and the claustrum in the ostariophysan subgroup Otophysi including for the first time representatives of all four major otophysan lineages.
I found that supradorsal cartilages, paired autogenous elements at the medial side of the neural arches, play a previously unrecognized ma-jor role during the ontogeny and evolution of the neural complex. In most otophysan taxa, supradorsals 3 and 4 fuse with supraneurals 2 and 3 (cypriniforms) or with supraneural 3 (characiforms, some siluriforms, gymnotiforms) during ontogeny to form the neural complex. In the major-ity of siluriforms the neural complex is exclusively formed by suprador-sals 3 and 4. In loricarioid siluriforms, a neural complex is missing, the fourth neural arch fuses to the occiput in development, and the first free vertebra is the fifth. Of the four otophysan subgroups, siluriforms are the most and cypriniforms the least diverse in the structure and development of their neural complex. Among characiforms I discovered striking het-erochronic shifts in the development of the components that form the neural complex. I review previous hypotheses about the homology of the different elements of the neural complex and discuss these in light of my new findings.
My investigation of the development of the claustrum revealed the following results. The claustrum of cypriniforms has a cartilaginous pre-cursor, which subsequently develops an extensive lamina of membrane bone. The membrane bone component of the claustrum and its close as-sociation with the atrium sinus imparis, a perilymphatic space of the We-berian apparatus, are both synapomorphies of cypriniforms. The characi-form claustrum is not preformed in cartilage and originates as a mem-brane bone ossification, a putative synapomorphy of that taxon. Among siluriforms, the claustrum is present only in more basal groups and origi-nates as an elongate cartilage that ossifies in a characteristic ventrodor-sal direction, possibly a synapomorphy of catfishes. Gymnotiforms lack a claustrum. Based on a review of all previous hypothesis of claustrum homology and my new findings, I conclude that the most plausible hy-pothesis is the one originally proposed by Bloch (1900) that claustra are homologues of the supradorsals of the first vertebra.