The review well written by Simon M. Dittami et al gives a very interesting state of the art as well as the methods and perspectives about the concept of holobionts transposed to marine organisms. This review emerged from the brainstorming of the main participants of a workshop recently organized on this topic. This work also underlines the major limits which make the holobiont concept still puzzling and challenging to decipher even more for uncultivable species with numerous genes/molecules with still unknown functions. I think that some point of discussion should be addressed: do the authors think that unicellular and pluricellular organisms can be considered the same way in terms of plasticity/evolution and also in terms of methodologies? Also the authors do not mention the differences in between ecto and endosymbiosis and their respective impact on the genome evolution and adaptation, plasticity of both partners. The definition of endosymbiosis should be better detailed (intra, extracellular, strict…with some examples) and the ectosymbiosis should be defined. Do the authors think that the term “long lasting interactions” is finally well adapted to the holobiont concept if we take into consideration that along the life cycle of an individual (and its varying environment) the symbiotic community changes?

This manuscript, “the community perspective on the concept of marine holobionths: state-of-the-art, challenges, and future directions”, has been written by a working group created during the workshop on marine holobionths in March 2018. The manuscript builds on the evolution of the holobionth concept in the research landscape where new generation sequencing allowed unravelling the extent of long-term relationships between hosts and their diverse microbiota across the tree of life. The authors aimed at discussing the way concepts in research and marine holobionths evolved those last years to highlight the need of a better understanding of those complex systems, and detail the roadmap they forecast considering the technological and conceptual challenges this represents. To that extent, this is a very timely and interesting topic. The main objectives appear to be the promotion of an increased a focus/scale up of research at the level of the holobionts, and in fine of a better integration of manipulative experiments on models with large-scale inventories and analysis of interactions/co-occurrence. I found this perspective and opinion manuscript very interesting, but also very dense. I make several comments here below that I hope can be useful.

Main comments The team chose a “short” format that, because rather well documented, is also rather dense. Some of the sections may gain in being structured in subsections, or rephrased for the reader to follow better the lines of thoughts and connection among sentences and examples. This is particularly true for the sections on models and emerging methodologies. Section on “marine holobionts models” may be easier to follow, and possibly more accurate with a clear distinction between works on symbiosis (i.e. the observation and study of bilateral associations), versus their extension to the analysis of the complexity and function of a holobiont system as a unit (in a similar way Margulis is approaching this in its 1990 article). The detailed descriptions of symbiotic models may this way naturally reflect the reasoning that led to the development and increasing used of the holobionts concept. The part on evolution of holobionts seemed to me slightly disconnected from the rest of the MS mostly dealing with ecological approaches and consequences of the holobionts structure and function. As it stands, it is too short to render justice to the many aspects to consider and change in our approaches, to integrate the holobionts concept to evolutionary theories and studies (compared to subsequent sections mostly dealing with ecological studies). If remaining in the MS, this section may gain in being better integrated with the following ones (meaning those should also be expanded). Another option would be to keep the focus of the MS on ecology and skip that part. The part on marine holobionts functioning relates much more to the determinants and mechanisms of association and dislocation of partners, than to their respective functions in an ecosystem, maybe another title would be more appropriate? The way the two main axes of research are presented in figure 3 and discussed pages 9 and 10 is very interesting and offers a very clear view of the extent of the gap that needs to be bridged. Would it be interesting to consider the use of large-scale studies describing complex systems over large spatial and temporal scales to identify new models for experimental approaches in order to help bridging that gap? How could this be done on the basis of large networks describing complex systems of “Russian dolls” in nature? This may interestingly link to the section on emerging methodologies and the use of “-omics” techniques? In relation to this, the figure 3 is interesting but may be more complete and offer a better global view of the scope and conclusions of this article if completed to illlustrate the way new technologies will help bridging the identified gap. It may also include the illustration of the way such changes in paradigm and approaches can have a cascade of consequences on the applications detailed in the last section. Line 427 to 429, and later: A significant amount of literature has been published those last years on marine systems including not only corals or sponges but also macrophytes, showing the importance of taking into account the microbial compartment to understand the ecology and evolution of the host. How could management plans to fight against some invasive species, or conserve of declining ones similarly take into account and manipulate microbiota, reform cautions wisely discussed in the end of that section? Finally, this manuscript is mostly written on an ecological (rather than evolutionary) perspective. Following that line, it seems to me that an interesting add to this opinion article would be a discussion on the accuracy of the now exploding amount of habitat models and forecasts for habitat shifts (particularly under global change scenario) most often only taking into account the abiotic compartment and ignoring the biotic one.

Minor comments and suggestions: Line 120 As opposed to vertical transmission, I may suggest that the arisen transmission is just the acquisition of the associated microbiom from the environment (which process can be continuous and independent of the emergence of new generation?).

Line 159-161 and on: the authors offer brief and interesting history of the research lines that led to coin the term holobionts. This article from Margulis should be cited as 673-670 (not just 673). Despite the term holobionts is introduced in the legend of 2 figures, most discussion in this particular article relate to the difference between symbiosis and mutualism, analogy between symbiosis and sex, and symbiosis as a main source of novelty (compared to gradual changes; in the sense of Mereschkovsky). May be this is a semantic distinction, yet the way this is phrased in the 1990 article relates symbiosis to evolutionary novelty rather than describing it as the way “evolution mainly work…”. Is it possible that the description of the proposition of Margulis (lines 159-161) in the present manuscript would be more related to its book “Symbiosis as a Source of Evolutionary Innovation” (1991) that is most often cited as the origin of this term? Line 156: Is widespread (I is missing) Line 210-213: sentence difficult to understand Line 328 to 332 the sentence is unclear. Line 352 biology is a rather vague term here, and it may be worth here to detail the need for physiology and biochemistry?