A Comparative Study of Reproduction and Development in the Polychaete Family Terebellidae

The reproduction and development of four species of terebellid polychaetes from the west coast of North America were studied and compared with several other terebellid species to reveal the covariation of life history traits in the group, and assess any limitations on terebellid life history evolution that may be imposed by ancestry or body design. The four species in the present study span the range of reproductive and developmental modes known for the family Terebellidae. Eupolymnia crescentis and Neoamphitrite robusta are both free spawners that reproduce during discrete 3-month breeding periods. In E. crescentis, oogenesis takes from 5 to 8 months and spawning occurs from July to September, maximum oocyte diameter is 210 {mu}m, and fecundity reaches {approx}128,500 during a single breeding period. The E. crescentis larva develops near the bottom for about 7 days before settling as a five-setiger juvenile. Neoamphitrite robusta reproduces from April to July after a 12-month oogenic cycle; oocytes in this species measure up to 180 {mu}m, and fecundity reaches ~830,000. The two brooders in the study, Ramex californiensis and Thelepus crispus, brood their larvae in the maternal tube. T. crispus reproduces continuously for at least 6 months, and has up to 51,500 larvae in a single brood. The oocytes in this species (400 {mu}m) give rise to larvae that are brooded to the one-setiger stage and then emerge to undergo a one-day planktonic period before the larvae settle and become juveniles at eight setigers. Ramex californiensis reproduces continuously year round; larvae are brooded in cocoons that are laid sequentially in the tube, with up to 44 larvae in a single cocoon. Development from the 410 {mu}m oocytes is direct, and juveniles have 11 setigers. Unlike E. crescentis and N. robusta, in which oogenesis is synchronized within individuals to produce a peak of large oocytes during the discrete spawning period, R. californiensis and T. crispus females have a wide range of oocyte sizes throughout the year. Correlation analysis and analysis of variance of reproductive and developmental traits of these and several other terebellid species revealed some expected trends. For example, egg size varies according to the mode of reproduction (free spawning, extratubular brooding, or intratubular brooding), and is also correlated with juvenile size. However, egg size does not predict fecundity in terebellids when body size is held constant, and brooding is not restricted to small-bodied species. Indeed, the largest and smallest species in the study brood their larvae intratubularly, suggesting that allometric constraints may not be important in determining mode of reproduction in these polychaetes. The Terebellidae is a diverse family found in all marine habitats, yet all known terebellid larvae are non-feeding; this contrasts with the occurrence of both planktotrophy and lecithotrophy in other polychaete families, and leads to the proposal that larval development in terebellids has been constrained during the evolution of the lineage. The results of this study demonstrate that generalizations regarding complex relationships among life history traits are often inappropriate. The need for more comparative studies of marine invertebrate reproduction and development, and the integration of phylogenetic analyses into the study of life history evolution in marine invertebrates is highlighted.