Evolution of pelagic direct development in the starfish Pteraster tesselatus (Asteroidea: Velatida)

Despite a diversity of larval forms, remarkably conservative features of asteroid development define a larval body plan that occurs throughout the class. However, recent work on the starfish Pteraster tesselatus has documented a highly derived pattern of development. Several features, including radial symmetry, parallel embryonic and adult axes of symmetry, absence of a preoral lobe, and formation of coeloms in the adult orientation from seven separate enterocoels, have not been reported in asteroids before. The complete absence of the larval body plan features that are found in other asteroids, indicates that P. tesselatus develops directly from the embryo to the juvenile and has a pelagic, nonfeeding (lecithotrophic), but nonlarval mode of development. I postulate that direct development evolved over an extended period in a lineage of brooding, deep-sea velatid (probably pterastcrid) ancestors of P. tesselatus. Selection for increased developmental efficiency (loss of nonfunctional larval features) in the brooded offspring, could explain the lack of larval settlement structures, the nonlarval arrangement of coeloms, the lack of a preoral lobe, the transverse orientation of the juvenile disc, and the lack of bilateral symmetry. The pattern of coclomogenesis could have been derived from that of other velatids (e.g. solasterids) by relatively simple changes in timing and orientation of entcroeoel formation. Rotation and posterior translation of the coelomic fate map of the archenteron prior to enlerocoel formation would produce the coelomic compartments in the adult orientation that characterizes direct development in P. tesselatus. These unusual developmental features lead to a radically different interpretation for the evolution of the pelagic ‘larva’ of P. tesselatus: (1) evolution of benthie brooding, (2) extreme simplification of development involving the loss of all larval features from the life cycle, and (3) subsequent re-evolution of pelagic development. In the case of P. tesselatus, where all larval structures were lost, there do not seem to be functional constraints preventing the re-evolution of pelagic development. Analysis of pelagic and benthie larvae, in other asteroids, suggests that major ecological transitions in life histories need not be associated with substantia] changes in morphology. The loss of pelagic development should have occurred repeatedly and should be readily reversible. These findings have interesting implications for the loss and evolution of pelagic dispersal in the life histories of marine benthie invertebrates.