The success of coral reef ecosystems largely depends on mutualistic symbiosis between scleractinian corals and the dinoflagellate photosymbiont Symbiodinium spp. However, further investigation is needed to elucidate the flexibility of coral-algae associations in response to environmental changes. In this study, we applied a molecular method (high-throughput internal transcribed spacer 2 region of ribosomal RNA gene amplicon sequencing) to explore diversity and flexibility of Symbiodinium associated with Galaxea fascicularis, an ecologically important scleractinian coral species collected at five locations around Hainan Island, South China Sea. The results revealed a high diversity of Symbiodinium subclades with C2r and D17 being dominant in G. fascicularis. Clade D Symbiodinium occurred most frequently in habitats where the annual average sea surface temperatures are the highest, suggesting that temperature is an important factor in determining Symbiodinium D abundance in G. fascicularis. The distribution of coral-Symbiodinium associations are possibly mediated by trade-off mechanisms which change the relative abundance of Symbiodinium clades/subclades under different environmental conditions. These findings provide further evidence that reef-building corals such as G. fascicularis can shuffle their symbionts to cope with environmental changes, and have implications for our understanding of the ecology of flexible coral-algal symbiosis.

Fusion of embryos or larvae prior to metamorphosis is rarely known to date in colonial marine organisms. Here, we document for the first time that the embryos of the broadcast spawning coral could fuse during Platygyra daedalea blastulation and further develop into conjoined larvae, and the settlement of conjoined larvae immediately resulted in inborn juvenile colonies. Fusion of embryos might be an adaptive strategy to form pre-metamorphic chimeric larvae and larger recruits, thereby promoting early survival. However, future studies are needed to explore whether and to what extent fusion of coral embryos occurs in the field, and fully evaluate its implications. Amendments from Version 1According to the comments from the reviewers, we mainly made the following changes:1) We improved the text in many instances as suggested by Prof. Rinkevich. For instance, in Materials and Methods, we used "combined" instead of "mixed"; we replaced "each type of fusion" with "chimeric larvae". Moreover, to demonstrate the sectorial fusion, we described these conjoined larvae as multi-headed in the Discussion.2) We have corrected the wrong idea of the sexually produced embryos not being genetically distinct and rewritten this part in the Discussion.3) Prof. Rinkevich suggested that the results of Mizrahi et al., 2014 would change our statement of "for the first time". Our observation presented the embryonic chimeras as a result of fusion of embryos in a broadcast spawning coral, whereas Mizrahi et al., 2014 revealed that the larvae of brooding coral Tubastraea coccinea could metamorphose and aggregate in swimming groups. Thus, we did document for the first time the fusion between individuals at the embryonic stage in reef corals and the inborn colonies of multiple polyps upon settlement. Moreover, we followed the settlement and growth of these chimeric larvae. Overall, it is appropriate to state that we documented the embryonic chimeras and inborn colonies "for the first time" in reef corals. 4)We point out the absence of water turbulence in this study as suggested by Prof. Baird and state that it may be a factor triggering the fusion of embryos. We added the references of Mizrahi et al., 2014 and Gauthier et al., 2008 to further discuss the potential implications of this phenomenon.We are really grateful for the comments from the two reviewers and the efforts of the editorial office.