I have read and assessed the paper by Johnson et al, in which they outline the application of CAUs to assess calcification rates in marine ecosystems. It is an excellent contribution that is well written, concisely-presented, and will certainly be very useful to a wide range of marine scientists. I have no major issues with the paper as it stands, and would be happy to recommend it for publication without any substantial revisions. However, I have a couple of suggestions that the authors may want to consider before publication of the manuscript:

1) While I understand that it is not necessary to show data that backs up the use of CAUs for this particular format at MEE, and even though I don't necessarily think that the tool requires a "proof-of-concept", I do think that it would be really helpful for the broad readership of the journal to see some exemplary data that can be gathered from CAU deployment. For instance, if the authors had data from two or three year-long deployments of a few units, either across different systems or bioregions (to highlight the fact that the CAUs can pick up variability in system- or region-specific ecosystem functions), then it would be wonderful to include this in the paper. In my opinion, it doesn't even have to be "new" data -- it can be pulled from previous work they have done, as long as it is referenced appropriately. This will offer readers a direct glimpse into just how useful such a standardized assessment can be, and (if they can provide data from different systems) emphasizes the universal utility of the approach across ecosystems and bioregions.

2) Since there are a couple of different data-types that can be gleaned from the CAUs (e.g., recruit numbers, net calcification, community composition), I would recommend that the authors include a bit more detail on this, maybe in a small table. Ideally, it would provide all the different data that can be extracted from the CAUs, the units that these data are expressed in, as well as perhaps the most relevant statistical approach to analyze these data. Of course there are many ways to skin a cat (no harm to our feline friends intended), but since the authors are presumably quite experienced with the data types coming out of the CAU deployments, a few recommendations would be quite helpful for researchers that want to employ them in their studies. It would also help standardize numbers and results in future publications that feature CAUs, which will be quite useful some 10 years down the road, when a diligent grad student might try to synthesize the results of these studies. None of the data should be particularly challenging to analyze, but may require some thought. For example, the gCaCO3cm-2yr-1 may frequently have a fairly odd data distribution, as it is (presumably) non-negative, continuous, and possibly not normal. This might necessitate either a log-transformation (for a normal ANOVA/linear model) or a generalized linear model with, for instance, a gamma-distribution. Similarly, numbers of recruits definitely won't be adequately analyzed under assumptions of normality, and should be treated appropriately using an exponential family distribution (being mindful of zero-inflation when counts are low). I don't mean to turn this paper into a statistical monograph by any means – nevertheless, end-users may greatly appreciate a bit of thought and guidance with regards to the data types that are gathered from the CAUs and how to treat them quantitatively.

Beyond these two suggestions, I have no other substantial comments. I'm providing a few minor editorial suggestions below, but overall, the authors made this a very easy paper to review.

L. 30: can be determined
L. 41: If you are able to provide some data, then this would be a good place to state something like: "we show that their application in different ecosystems distinctly identifies differences in ecological processes" or similar.
L. 57: "insights into the structure"?
L. 59: As an author on both papers, I feel like this statement is more prominently supported by Brandl et al. 2019. Not fussed either way though.
L. 123 (whole paragraph): I really appreciate the limitations the authors point. I would add another caveat to this, which is that, even thought he design is meant to expand the type of microhabitats available in a hard-bottom marine system, it is a crude imitation of three broad habitat types (exposed flat surfaces, a flat "crevice", and an overhang), that naturally falls well short of many natural benthic habitats. The authors are, I'm sure, well aware that there is a substantial number of papers that highlights the importance of fine-scale microhabitat complexity in governing the benthic community by mediating ecological dynamics such as recruitment (Nozawa et al. 2011; Edmunds et al. 2014) or predation and competition (Doropoulos et al 2016; Brandl & Bellwood 2016). I think it would be wise to acknowledge that the lack of this fine-scale complexity likely influences these dynamics and subsequently, community development and calcification rates.
L. 289: Would it be possible to provide guidance on the types of functional groups that you would suggest including?

Overall, I think the paper will be a great contribution that can support our efforts to standardize assessments of marine ecosystem function. If you have any questions about my comments, please feel free to reach out.

Best wishes,

Simon J. Brandl