Main content area

Unexpected structured intraspecific diversity of thioautotrophic bacterial gill endosymbionts within the Lucinidae (Mollusca: Bivalvia)

Brissac, Terry, Higuet, Dominique, Gros, Olivier, Merçot, Hervé
Marine biology 2016 v.163 no.8 pp. 176
Bivalvia, adults, antagonists, coevolution, cross infection, ecological function, endosymbionts, genes, genetic markers, host specificity, hosts, internal transcribed spacers, phylogeny, ribosomal RNA, seagrasses, sequence analysis, strain differences, surveys
In chemoautotrophic associations, sequence comparison of 16S rRNA has been the method of choice to study bacterial diversity in the context of host/symbiont coevolution. However, the relative low rate of evolution of 16S rRNA has been shown to result in a diminished capacity to discriminate between closely related bacterial strains or species. Within chemoautotrophic associations, as described in several studies, the use of other genetic markers may reveal previously unobserved strain diversity among gill endosymbionts. Herein, we conducted a survey of symbionts harbored by six species within the Lucinidae family using five genetic markers (dnaE, gyrB, aprA, cbbL genes and the 16S–23S internal transcribed spacer). Thus, within the already described SoLuc_1 bacterial species shared by six host species we observed an obvious bacterial strain diversity. This diversity is structured in function of the geographic location of the hosts and not in function of ecological parameters or host phylogeny. Interestingly at the local level (same seagrass bed environment), we observed an unexpected specificity in host species/bacterial strain association (i.e., all individuals of the same host species harbor the same symbiotic sequence type). This specificity of association implied that there was a control of symbiont strain acquisition by the host, which was confirmed by a cross-infection experiment of starved adults performed in our study. Based upon our results and other evidence from the literature, we hypothesize that this pattern may be due to a “capture/escape” type antagonist evolution of the two partners.