Host use and diversification in symbiotic polychaetes

Abstract

I studied development and larval biology, host use, and gamete interactions in three obligately symbiotic polychaetes in the genus Arctonoe (Family Polynoidae). The three polychaetes are ectosymbionts of other marine invertebrates, and within the Puget Sound region, their host ranges do not overlap. Except for A. fragilis, which feeds on host tissue, they are probably commensals. The three species have similar ontogenies, with long-lived planktotrophic larvae; benthic juveniles seek out hosts soon after metamorphosis.Surveys of allozyme variation confirmed that the three Arctonoe species were genetically distinct. Populations of A. pulchra on two host species were also strongly differentiated, and should be considered distinct species. Populations of A. vittata associated with three host species were weakly differentiated in one of two years. Thus, some "host generalists" are actually comprised of isolated lineages of host specialists. I used transplant experiments to examine the role of physiological incompatibilities in restricting symbiont host range. A. pulchra and A. vittata died when paired with the seastar Evasterias troschelii, the principal host of A. fragilis. In other combinations of symbionts and hosts tested, all symbionts survived. While these incompatibilities can partially explain observed patterns of host use in Arctonoe, other processes must also be operating.Gamete incompatibilities among closely-related species may be a consequence of divergence at gamete recognition loci as a by-product of reproductive isolation, selection against hybrids, or a process of sexual selection at polymorphic gamete recognition loci. The first two hypotheses predict that gamete incompatibility appears after reproductive isolation has arisen, and the third that incompatibility appears simultaneously with isolation. Gametes of Arctonoe spp. are compatible in all crosses, over a range of concentrations and contact times, despite estimated divergence times of 1-3 million years before present. These data are consistent with the first two hypotheses but allow rejection of the third.Another common pattern in interspecific gamete interactions is asymmetric gamete compatibility. I use a simple genetic model to show that such asymmetries are predictable, but temporary, consequences of divergence between isolated populations.