SHIPS FIRING IN THE NIGHT? MESOZOIC RISE OF TELEOSTS ACCOMPANIED BY DIVERSITY LOSS AND TRAIT SHIFTS IN HOLOSTEAN FISHES

Teleosts are the dominant living group of aquatic vertebrates; they represent approximately half of all vertebrate species (29,000 sp.), assume a bewildering array of morphologies, and have come to occupy nearly every environment imaginable. In contrast, their holostean sister group (gars and bowfin) consist of a mere 8 species restricted to the freshwaters of North America. Despite the contrast today, classic palaeontological accounts point to a long 'golden age' of holostean diversity in the Mesozoic, asserting that they were dominant over teleosts. These accounts also argue for a rapid competitive replacement of the dominant holostean fauna with a dominant teleost fauna.

To interrogate these classic accounts, we assembled taxonomic, faunal, shape and body size data for Mesozoic neopterygians. Contrary to the traditional pattern, the period of clear holostean dominance is relatively brief, and the ascendancy of teleosts was a gradual process which was not accompanied by a rapid holostean decline.

Beyond diversity patterns, we examined the potential for interactions between holosteans and teleosts using functional traits and body size. Functional ordination space revealed that teleost analogues frequently encroached upon holostean functional capabilities late in the Mesozoic, granting some plausibility to the scenario of a competitive replacement. Mean body size trajectories revealed that while holosteans were initially relatively small in the Triassic, they increased in mean size in the Early Jurassic, leaving a void that was filled by the emerging teleosts by the Middle Jurassic. Holosteans then remain larger than teleosts for the rest of the Mesozoic, never returning to their Triassic mean size which remained occupied by teleost. This pattern whereby holosteans and teleost appear to segregate into different size classes is consistent with the expectations of niche partitioning, yet played out over a macroevolutionary timescale. The outcomes of our approaches suggest that fossil fishes present a promising system for investigating clade interactions on geological timescales.