Dear editor and authors,

The paper by Campeny et al is an interesting investigation of REE potential in the Fuerteventura carbonatites. This is also interesting because oceanic carbonatites have never been thoroughly assessed for REE mineralisation before (as far as I know). There is also scientific value in that the rocks obviously contain antiskarns, a very timely research topic (currently unrecognised by the authors, see comments below). I recommend the paper to be revised.

line 25: Here, and after, please use significant figures correctly. A number like "10301.83" is meaningless because it implies that you know it to a precision of 0.01 ppm, and it's distinct to 10301.84 or 10301.32. Something like "about 10000 ppm" or even "about 1 wt%" should be ok, and please learn on correct usage of significant figures.

line 28: "enrichment" indicates a process. You just mean that the silicate rocks aren't as REE-rich as the carbonatites?

line 67: Not accurate. Light REE are coming from magmatic rocks, primarily carbonatites. Heavy REE are coming from soils and weathering products.

line 72: The 50% limit by Le Maitre is restrictive and misleading. The carbonatite community is moving away to define carbonatites as rocks that form from carbonate melts, regardless of carbonate mineral content. See Yaxley et al 2022 here http://www.annualreviews.org/eprint/JVDPC4NDH4BAFJNWDAVJ/full/10.1146/annurev-earth-032320-104243

line 73: "basic" is outdated - please use mafic, ultramafic.

line 86: Since you're specifically talking about REEs in carbonatites, the processes outlined by Anenburg et al 2021 are probably very relevant here https://doi.org/10.2138/gselements.17.5.327 and also Yaxley et al 2022 discusses this in detail.

line 87: This paragraph will benefit from reference to Weidendorfer's work, for exmaple https://doi.org/10.1130/G39621.1 and https://doi.org/10.1007/s00410-016-1249-5

line 92: Which "aforementioned approach"? You're doing mineralogy and geochemistry, but you hardly talked about any mineralogical and geochemical approach above.

line 123: "The FBC unit..." (other places in the places where you can safely switch word order and get rid of the "of the"...)

line 131: Partial fusion of what? The country rock?

line 139: Dikes of what? Carbonatites? Something else? Implied by your next sentence but worthwhile to clearly say this.

line 141: Can you mark on the figure exactly where the fenites and carbonatites are?

line 147: "Pyroxenite" cannot really be a magma type - it is a cumulate rock formed by crystallisation from a magma, with the magma migratin elsewhere. Alternatively, "pyroxenite" metasomatic zones around carbonatites are probably antiskarns - see for example https://doi.org/10.2475/03.2018.03 or https://doi.org/10.1016/j.chemgeo.2022.120888

line 157-184: If you paper is on REE, and these units have no REE, why are they in the paper? Consider removing this.

line 191: Remove commas from this sentence

line 192: Again, the Anenburg et al 2021 paper is probably relevant here

line 196: Not only that, sometimes weathering can make a deposit into an economic one, with the two best examples being Araxa and Mount Weld: https://doi.org/10.1093/petrology/egae007 and https://doi.org/10.1016/j.jsames.2023.104311

line 252: Check usage of word "basic"

line 256: "primarily aegirine-augite and biotite" - the "primarily" indicates there are other mafic minerals. What are they?

line 258: Described where? Here? If yes, then rephrase without "described" because it's implied...

line 295: What you're describing are precisely "antiskarns", a very hot topic of research these days. Concept first introduced by Anenburg & Mavrogenes (2018) which I referenced above, also see in depth discussion in Yaxley et al 2022. Vasyukova and Willy-Jones also talk about this (not sure if they use the name "antiskarn" though). Some other examples where similar textures and styles are observed: https://doi.org/10.1016/j.lithos.2023.107231 https://doi.org/10.1016/j.lithos.2023.107480 https://doi.org/10.1016/j.lithos.2022.106647. The britholites you're seeing are also very typical. Experimentally recreated here: https://doi.org/10.1126/sciadv.abb6570 (see experiment CbSi), and also observed in nature here http://hdl.handle.net/1885/154263 There's also a paper should be out in Contributions to Mineralogy and Petrology on the topic very soon with thermodynamic modelling. If it's not out by the time you do the revision, then contact me and I'll send it to you.

line 299: Nice! These britholite+allanite textures recreated in our Anenburg & Mavrogenes (2018) https://doi.org/10.2475/03.2018.03 very indicative of a late hydrothermal alteration of a primary carbonatite.

line 300: Correct spelling is sulfates, not sulphates (f- spelling endorsed by the UK Royal Society of Chemistry for example). Also IMA-approved spelling is baryte, not barite

line 320: Reduction in illite? This means you some to begin with, but this is the first time you're mentioning illite and chlorite

line 329: Why don't you just say which elements, instead of referring to a supp table?

line 333: Please use this for typical crustal values https://doi.org/10.1016/B978-0-08-095975-7.00301-6

line 337: See previous comment on significant figures. You can leave this precision for the supp tables (as long as uncertainty is reported with it), but not in the main text.

fig 8: Please add the Pm gap. Also better to use normalisation values from O'Neill 2016: https://doi.org/10.1093/petrology/egw047 CI in Table 1

line 352: You still haven't said which elements these are. What is "significant"?

line 355: How do you reconcile it with the fact that you had pyrochlore? Could this be an analytical artifact?

fig 9: Also interesting to see Zr depletion given that you had zirconium minerals in your rocks

fig 10: Y axis: These are ratios, right? Then this should be in log scale

line 387,395: Significant figures

line 388: Use Rudnick, not Balaram

line 409: Also discuss Mount Weld maybe? See Chandler reference above. Furthermore, all these large carbonatites are usually one big plug or intrusions. You have rocks which appear as various thin dikes and veins. Maybe a more comparable deposit would be Maoniuping - look it up on Google Scholar, lots of references for it.

line 422-427: I recommend removing this. The topic is much more complex than you put it, and not really in the scope of your manuscript.

line 434: Why is this noteworthy? Calcite is overwhelmingly the most common mineral in all carbonatites worldwide. This is akin to saying that cpx is noteworthy in basalt or quartz in granite...

line 503: This cannot be understated - Fuerteventura is an UNESCO biosphere reserve!

line 538: One final sentence could be useful here:

"Given the non-exceptional REE grade of Fuerteventura compared to other deposits, most REE being hosted in unexploitable and refractory britholite, irregularly distributed mineralisation with low overall tonnage, and Fuerteventura being a UNESCO biosphere reserve, we conclude that economic development of any REE resources on the island is extremely unlikely to occur." And something similar in the abstract as well. Just saying, because on a bigger scale you don't want to start any unnecessary hype that Fuerteventura is "the next big thing" because no good can come out of this.

-Michael Anenburg, ANU.

The authors present a summary of the geology of Fuerteventura and present mineralogical and geochemical data to evaluate the viability of some of the rocks and their weathered counterparts as potential sources of the REE. The paper is an interesting thought exercise, but based on the results, the authors should be absolutely clear that these rocks have 0% chance of ever being a REE mine – even on a small scale. The authors focus on the carbonatites as these have the highest grade – reaching 1 wt.% total rare earths in a single grab samples. While some comparison is made to REE grades in existing REE mines, such a comparison is disingenuous, and in many deposits 1% REE would barely make the cut-off grade. Moreover, the samples presented are relatively mineralogically complex, with the REE spread across three different mineral – with little indication of which mineral would be a focus. The carbonatite bodies are small and discontinuous, further rendering them an economic non-starter. Lastly, all the rocks studied appear to be located in a protected area – perhaps outside of the scope of the thought exercise, but an important point nonetheless, and one that should probably be made and emphasised in case a lay-reader might misinterpret the paper.

I was surprised by the lack of attention given to the weathered alkaline rocks. Considering these have a profile of a metre thick in places, could there be thicker weathering profiles elsewhere? Should the authors wish to take the thought exercise further, they may way to look into some of the samples as potential ion-adsorption type deposits. Interestingly, only a few of the samples actually seem to have weathered to clay, so I suspect that there will not be a large amount of easily leachable REE, but it may still be worth an inquiry.

Line 25: Check SF.

Lines 38-53, could be condensed to a few sentences.

Line 76, the example minerals you give here are fluorcarbonates.

Line 115: remove ‘ago’

Line 126: ‘associated to’ à ‘associated with’

Line 319: ‘carbonatite profiles’… on line 308-311, you mention that there is no weathering profile associated with the carbonatits, except for the development of calcrete veins. Perhaps you should be more specific on line 319, and say that the samples are of calcrete. How do the formation of these calcrete relate/differ from the calcretes mentioned on L180-184, described as forming from calcarentites, rather than an igneous precursor?

Line 407: It is disingenuous to take the average value of 2581 ppm REE across the whole complex, when the areas which actually define the resource are much higher concentration. No-one would take the felsic rocks from around the mountain pass area into a resource calculation, and the carbonatites have LREE contents over an order of magnitude higher than the Fuerteventura samples. A single sample of 1 wt% REE, while high for the Canary Islands, would barely make the cut-off grade for many carbonatite-hosted REE deposits. It is also somewhat disingenuous to compare grab samples (especially the highest grade grab samples) and compare these with resources from a select handful of other carbonatites. The values from most other carbonatites will reflect average grades over an area considered economically feasible to mine.

Line 414: check full stop after ‘Figure 11)’

Line 419-427: I wouldn’t make too much of this relatively flat HREE profile. The HREE are challenging to extract from carbonatites, and where these profiles are elevated, are commonly hosted in a different mineral to those which can be exploited commercially, and consequently lost during minerals processing. See https://doi.org/10.1016/j.mineng.2020.106617

Line 428-433: That the REE are split between three discreet phases only means they will be diluted further during processing.

Line 434-441: I don’t quite follow the logic here. What relevance does the presence of calcite have on the presence of REE-fluorcarbonates?

Lines 462-465: the examples given are all of carbonatites with significant weathering and the development of regolith up to (and over) 100 m thick. The carbonatites in Fuerteventrua have developed calcrete veins up to a few cm, locally. Why make the comparison?

Lines 443-472: There’s no consideration given here to ion-adsobtion type deposits which, given only the weathering profile above the syenite is developed more than a few cm, could perhaps be of consideration. Ion adsorption deposits require at least 50% of the REE in the weathering profile to be easily leachable using a medium pH reagent, such as ammonium sulfide. In these cases, the REE are loosely bound to clays developed on the weathering profile, and can be easily stripped from the clay and recovered. Ion adsorption type deposits have much lower cut-off grades where relatively cheap in-situ leaching can be applied, and low grade resources can be economic – especially where HREE contents are high. I am surprised that this avenue hasn’t been explored.

Lines 475-477: Based on the geochemical data, maybe, but based on the field observations, it is clear that the extremely small size of these bodies does not warrant any further investigation.

Line 477-478: Grade is not everything. Size and mineralogy are important too. A large, mineralogically amenable, low grade deposit can be much better than a small, mineralogically complex, high grade body.