We are grateful to the 3 reviewers who have helped us to improve the paper. All their suggestions have been considered and the article modified accordingly. Our comments to the suggestions are written in italic bellow.

Reviewer 1

I will suggest adding a term referring to symbolism, since the important symbolic contribution of this set is highlighted, as well as the term social network, the main subject of the work. Key words have been added

From introduction, the works cited in this section do not include Whallon (2006), which the authors mention at the end of the paper, as refer to the relationship between ornament and mobility Reference has been added

The methodology is correct and adequate. In this section on microscope analysis, a detailed list of experimental work is given. Maybe the paper of Avezuela et al (2011), which presents a detailed experimentation on the gastropod Littorina obtusata, could also be taken into account. The reference has been added

Regarding Tritia gibbosula, we can note as on the French Mediterranean coast, in southern Europe it is also present in the Gravettian levels of Cova Beneito (Muro, Alicante)(Soler, 2019) and in Solutrean and Magdalenian levels of Cova del Parpalló (Gandia, Valencia) (Soler, 2015 and Soler y Aura, e.p.). Further east, in Croatia, it appears in Palaeolithic and Mesolithic levels of Vela Spila site (Boric and Cristini, 2018), evidence for long-term regional and diachronic differences and similarities in types of body adornment among foragers of the European area. Ref added but not for Coval del Parpallo, N. gibbosolus is only mentioned for layer II attributed to the Final Solutrean.

Table 2 shows the list of pieces from the three studies carried out at different times in level II of this site. We note that Taborin (1992) describes at least 3 species that do not appear in the current study. We have added to the text ‘’The long history of curation of the collection may explain discrepancies between previously published material and the material currently present in the collection.’’ All the material currently present in the collection is shown in figures 2 and 3.

From morphometric, technological and use-wear analysis we would note the accurate work do in this part, especially in the perforation’s description. The Kubicka (2017) paper could help in the study to discriminate predation hols because it is a review of animal predation perforations and their implications for the archaeological record that may provide some more concrete input to this work.
We have added the Kubicka 2017 reference in the method section. However, the authors focus on the perforation location to help discriminate between human-made versus predator holes on shells. In our own study, we use a combination of criteria based on taphonomic data and experimental data including hole location but also morphology of the hole, edges, presence of use-wear or tool marks (cutting, percussion…). The location of the hole alone shows a large degree of variability, as shown in the Kubicka paper, and is not a criterion that can be used on its own for identifying perforation origin on shells.

Regarding the use of ochre, it’s only mentioned in a Glycimeris described as a container. None of the ochre shells show traces of dye? It is quite common that when traces of dye appear, the do not appear on a single piece. In case there are no traces of dye in any of the other pieces, we consider that this should be mentioned. It will be interesting to know the composition and origin of this ochre.

No other evidence of ochre is observed on the material. The residue in the Glycymeris is of an important volume, very thick and first results (SEM, Raman, Xrf) show a complex composition. This piece is still under study and is not part of the personal ornament material. It will be published in a separate paper and its analysis required additional time. We have added to the text ‘’No trace of this red compound was found on the rest of the shell collection, and we suggest this Glycymeris specimen be considered functionally distinct from the rest of the assemblage; the composition of this red compound of the subject of ongoing analysis.‘’

In line 339 it is said that in 213 remains, no anthropogenic action can be determinate due it post-depositional alterations. If they are so important, perhaps they should be described in detail. It would be useful to have this information in an annex.
This is not what we say. It is written :
Of the 217 Tritia reticulata shells, 24 are perforated (Figure 8A). Perforations are mainly observed on the dorsal side of the last whorl, but several small perforations are also observed on the ventral side, and sometimes close to the apex on either the ventral or dorsal side. Post-depositional alterations and recent exfoliation present on many of the perforations precludes identifying their origin, and so taphonomic processes cannot be completely excluded (Gorzelak et al., 2013).
Our stat is that 24 specimens show a modification (anthropogenic or taphonomic). Other individuals are not perforated at all.

From discussion
The authors propose us two hypotheses to the raw material Rochefield accumulation, but more hypotheses are possible; we do not know what the ultimate meaning of these pieces is. The second hypothesis is clearly unacceptable because of the support proves to be adequate in on other perforated pieces.
We agree, the second hypothesis is not sustainable for the reason the reviewer provides, and so we support the accumulation hypothesis. In the text we also reject alimentary use of the shells due to distance from coast. On the other hand, T. reticulata and Antalis are widely known and accepted as personal ornaments across the European Paleolithic, and so we propose the interpretation of the unmodified shells as ornament blanks is the most parsimonious one.

From conclusions, in line 652 the quotation from White, 1997 does not appear in the final bibliography. We should add the ref or remove the citation Reference added

Finally, it is remarkable the attention paid to the procurement strategies of these objects. Nevertheless, we would need more evidences to know that Rochereil is a central place in the manufacturing process, control and diffusion of these pieces throughout the area. the sentence has been rephrased

Also the number of Tritia unperforated is remarkable, considering the number of pieces of that size that we needed for a single necklace, a cap or a garment, the interpretation of storage for the whole area is, at the very least, an idea for discussion. we have changed storage for accumulation everywhere in the text.

In short, it cannot be ruled out that they could be, for example, elements designs for a single necklace. We agree, but whether those shells were dedicated to an individual arrangement or were individually distributed does not change our discussion regarding procurement strategy, functional hypothesis and scheduling of the manufacturing process. Our aim is to stay parsimonious in our demonstration and to avoid any speculation regarding the precise social function of the shells within the Magdalenian communities, as our data mostly document raw material economy and society mobility and network of contacts.

The proposal of the term embedded catching to explain a specific aspect of mobility is interesting to talk about connected societies based on movement and information exchange. The work of Romano, Lozano and Fernández-López de Pablo, 2021 is an interesting work on the reconstruction of prehistoric social networks and cultural transmission. The Rochefiel data suggest such interaction and exchange, although we certainly would not hypothesize the number of undrilled pieces as an accumulation and control of subsequent distribution as the only possible hypothesis.
Cache, control, embeded catching are concepts we have removed from the new version of the manuscript

The consideration of Rochefield as an aggregation site, as Conkey described, need more justification, the record of this site is not comparable to Isturitz or Altamira.
We do not qualify Rochereil as an aggragation site. Our sentence ‘’Accumulation sites such as this must therefore be integrated into our understanding of the “structured poses” of the aggregation/dispersion cycle by which we often imagine Magdalenian annual mobility strategies (Conkey et al., 1980; Rivero, 2014).’’ suggests that Rochereil is part of the variability of sites used by mobile foragers for their various activities including specific location for symbolic productions.

Review C. Dupont
A main comment we have after reading the paper is: is the caching of raw material the only hypothesis you can propose? Caching has been removed from the text
- were the intact shells found in a certain form of accumulation? We explain below that little information on object provenience is available
- which argument are in favour of a caching? - cf. supra
Are there other raw materials evocate caching? cf. supra
A second comment is about the used modern collections. Some of them are represented by less than 60 individuals. It is difficult to use them as a reference for a whole population. At least one of the two reference collections for each shell species reach 60 speciments (or nearly). cf bellow
I suggest bibliographical references below. Although general references are given to describe the geographical origin of shells, it will be more relevant to give more precise references for each species and information (location, fossil or not, usual sizes…). We have added references

Abstract
The abstract is concise and presents the main findings of the study. The term “raw material” can induce that the shells are raw material to make an object. See if you keep only this hypothesis according to the comments. Abstract has been modified but we keep the hypothesis of the shells as raw material (cf discussion bellow)

Introduction
Was the Rochereil collection knew from Vanhaeren and Taborin when they wrote their syntheses: did these two researchers known the shells from Rochereil’s excavation, did they interpret them No. We have precise in the text. ‘These shells have not been previously studied, with just a list of material published without evaluation of function (Taborin 1992).’’
Line 44 erase A.L. done

Archaeological context and objectives
The different stages are describe but what about the sampling (with water, sieving, size of the mesh...). The authors write about “spatial records”, but nothing is said about the spatial organisation of these shells.
we have added : ‘Precise findspot information is unavailable, and records only attribute material to archaeological stratum (Jude 1960). While the sediment was not screened, the excavation seems to have been meticulous and comprehensive, as recent water-screening of the backdirt with a 4mm mesh has not resulted in the recovery of additional shell remains (P. Paillet, unpublished). ’’
Lines 79-88 are difficult to understand with the figure 1, because the number of the layer are not clearly mentioned on the legend of the stratigraphy of figure 1. fixed
Taxinomy
L111-112 The criteria used for bivalve determination are lacking. text modified accordingly
L119 A more recent check of Latin names can be interesting. Some of them can change regularly. We have checked recent nomenclature and changes the text accordingly
L119 "Biodiversity Heritage Library": reference? reference added
https://www.biodiversitylibrary.org/

Morphological and morphometric analyses
L135 “naked eye” English. Maybe “by sight”? naked eye is ok
L137-140 Explain a little more: why have you chose these reference collections? How have they been chosen and collected? We have added to the text : “Marine reference collections were made by 2 collectors, collected over 45 minutes on two beaches targeted for their relative proximity to Rochereil.”
L140 Ocenebra erinacea The Latin name is Ocenebra erinaceus modified accordingly
Table 1 With less than 60 individuals it is difficult to use a modern collection. This number is not efficient to describe a population. For each shell species at least one of the reference collections reaches 60 specimens (or nearly).

Microscopic analysis
taphonomic and anthropogenic modifications: can you explain how you did that more precisely? Have you done a systematic analysis in sharing the shell into different parts? We have recorded modifications on each anatomical part of the shells (apex, spire whorls, aperture, lip, umbo, ventral margin, dorsal and ventral sides). Considering the characteristic of the assemblage (mostly unmodified shells) and the limit in interpreting the results due to the context and excavation, record and sampling field methods, a more specific and time consuming methodology would not have been relevant.

Results
Shell identification
L177 377 shells: precise which ones there are: all shells intact and modified?
Yes, functional analysis comes in the following section
L177 “to Layer II” : what about shells from other layers? no other shells in other layers
L183 “shores in the Pleistocene.” Reference / You cite the references in the methodology but we need more precise references for each origin of raw material. References have been added
However, few data are available on natural Pleistocene shell deposits. New approaches may fill this blank in our knowledge on past shell species distribution, as the recent paper of Bosso et al (The rise and fall of an alien: why the successful colonizer Littorina saxatilis failed to invade the Mediterranean Sea) in which ECNM is used to map the past distribution of a shell species, but this is not the purpose of the present paper.
L184-187 Why don’t you test the shape of the Tritia reticulata? It can be a good mean to know if all of them are coming from the same coast. As this species is polymorph, it can be difficult to compare the archaeological collection to the modern collection
Even if the morphology of the shell brings additional information, the main and useful information for the present study is the allochthonous origin of the material demonstrated in the paper. Tritia nitida is very similar to T. reticulata and has sometimes been described under the name. T. reticulata in the literature. In the case of Rochereil the two species may be present in the collection but as they can share the same biotope, the main information of the paper (allochtonous origin of the material) remains unchanged.
L185 “gastropod shell is conical, featuring of axial ribs containing 7 to 9 whorls, and is also 186 present along Atlantic and the Mediterranean coasts.” references added
L195-197 references added
L198-202 references added
L209 “Among the four valves possibly belonging to the Cardiidae family” possibly has been removed
L225 references added
L230 “probably belonging to the genus Spondylus” explain which criteria you use to say that? probably removed, description of the shell improved in the text
L235 “comparison with previously published inventories show that 3 shell species are missing from the current collection” The reverse is also true, some of the shells in your inventory had not been identified. How do you explain this? We do not explain it (and cf. above), all the shells currently present in the material are shown in photos in our article. The studies of Jude and Taborin do not provide exhaustive iconography of the material, preventing any comparison. the text has been improved.
L254-255 “The fragments of the bivalves Mytilus sp., Pecten maximus, Spondylus sp. and the two fossil Cardiidae show no anthropogenic modification.” The edges are not used or eroded? No.
L279 Is the size of perforations not small for natice’s predation? No based on Rojas reference. Some worms can do double perforations.
Another reference maybe with complementary parameters: Cabral J.P., Monteiro-Rodrigues S. 2015 - Orifícios de predação em conchas de moluscos marinhos. Um modelo experimental para o estudo de perfurações em conchas arqueológicas La Investigación Arqueomalacológica en la Península Ibérica: Nuevas Aportaciones, in Gutiérrez-Zugasti1 I., Cuenca Solana D., Gonzalez Morales M.R. (dir.), Actas de la IV Reunion de Arqueomalacologia de la Peninsula Iberica, Santander, 22-24 mai 2014, Santander, Nadir Ediciones, p. 241-250
text modified and reference added
L309 (Figure 6d, e) / L268 space or not fixed
L317 size of the natural perforation fixed
L323-327 Are you sure that the shape is stable for the murex? Have you got sufficient shells to work on such comparison? https://www.researchgate.net/publication/237730687_Comparative_morphology_of_Pliocene_Quaternary_and_Recent_shells_of_Ocenebra_erinaceus_Linnaeus_1758_and_O_brevirobusta_Houart_2000_Mollusca_Muricidae_Ocenebrinae_reflections_on_the_intra-_and_interspe

The publication of V. Berrou includes shells coming from a large geographic extent from Normandy to Morocco and the Mediterranean shores. In the present study, we have focused on the shore located at a shorter distance from the site. Our aim is not to discuss Pleistocene natural shell populations variability but to infer anthropogenic selection of the shells according to their size. One of our reference samples reaches 60 individuals which is a good size sample. However, in any case, the low number of Murex within the archaeological collection is a limit for the statistical analysis. We have dealt with what we have, trying to keep a balanced methodology and research aims considering sample size and chose the most parsimonious hypotheses.
L364 Why tusk shells are animals and have an orientation: anterior and posterior. Why do you use proximal and distal? modified in the text
L408-409: reference for fossil specimens? reference removed.
L409 “By comparison, the length of the scaphopods from Rochereil is, on average, 5 times greater than their maximum diameter (Table 4).” Please precise your idea? What is the consequence of such difference? It means that the scaphopods from Rochereil are shorter than the shells from the reference collections, short sentence added to the text
L417 Moreover, the “rare?” larger and longer scaphopods from Rochereil fall outside the size range observed in the fossil reference collection. rare has been removed

Regional comparison
A your description of Rochereil is mainly based on sizes of shells: are there other archaeological collections with such data? no. very few authors deal with shell size. Most of the published studies focus on sourcing, functional analyses and taphonomy of the surfaces of the shells.

Tables and figures
Table 2 Problem with the reference of the figure for Glycymeris / line associated to the symbol ° for figure Tritia reticulata fixed
Figures: The orientations of shells should be the same. The columella of gastropod should be vertical when photographed one by one. See orientation for bivalves. fixed
Figure 3: italic for the last name fixed
Figure 4 italic “striations on the ventral margin of the shells” Only the margin? I see them on the extern part of the shell but not only on the margin.
yes, they are not of the same origin. those external striations are more large, less deep and may be post depositional. This shell will be published in a dedicated publication with the residue analysis.

Figure 8: orientation with vertical columella. fixed
Figure 9: The shape of this shell is dependant of its environment. Difficult to describe this figure with so few data. cf comment above about T. nitida
Figure 10: italic/ posterior and anterior would be better. fixed
Figure 11: italic fixed

Discussion
L513-515: “The absence of anthropic modification on these shells suggests they were naturally deposited in the cave sediment, and not intentionally introduced by Magdalenian occupants.” Why do you say that? Have you stratigraphic difference between these shells and the other ones? Have you other elements from the karstic limestone near these shells?
As shown in the site presentation, Rochereil has not been excavated according to modern excavation standards. Provenience data are missing, and no spatial recording of the material is available. We mention in the article the fossil origin, the patina, the absence of anthropic modification, and we consider the natural presence of Miocene fossil shells in the Miocene karsts of the region. The most parsimonious hypothesis is the natural origin of those specific shells. Archaeological literature is polluted with dubious objects interpreted as personal ornaments based on poor functional, technological, taphonomic, and use wear data (see Rigaud et al. 2009). Here we secure the anthropogenic assemblage by excluding any other artifacts for which no secure data support an anthropic accumulation.
L524 similar: on which criteria? A possible Atlantic origin has been added to the text
L535 generally or mainly? yes mainly, changed in the text
L538 The region is “Poitou-Charentes” and now Nouvelle Aquitaine. We are using geographic areas instead of administrative regions.
L556 “The location of the site at a considerable distance from either coast indicates that the mollusks were not collected live for consumption.” Are they known to be eaten after archaeological records? Even if they had been eaten at some sites, it does not weaken the argument provided here.
L575-576 Precise that this result is based on archaeological observations and that it is an hypothesis. The sentence has been modified accordingly
L578-580: reference ref added
L567-568 There are references on murex. Examples: Berrou V., Merle D., Dommergues J.-L., Crônier C. & Néraudeau D. 2004. — Comparativemorphology of Pliocene, Quaternary and Recent shells of Ocenebra erinaceus (Linnaeus,1758) and O. brevirobusta Houart, 2000 (Mollusca, Muricidae, Ocenebrinae): reflections onthe intra- and interspecific variations. Geodiversitas 26 (2) : 263-295. [CL1] / Fossil and recent Muricidae of the world , Merle, Didier Editeur : ConchBooks. Hackenheim, Germany,2011[CL2] . Can you verify if there are data on their size and shapes? We have checked. No data available
L593-594 “The fact that most of the shells recovered from Rochereil were unmodified suggests that the objects were collected with the intention of being transformed into ornaments,” I am not agree with this consequence. It is not because that these shells are unmodified that human have preview to do it.
The article provides several argument for ornamentation: selection of the larger scaphopods, the human made perforation on few T. reticulata, the evidence that shell’s introduction into the site was not for consumption (shore too far from the site), the common use of similar shell species as personal ornaments at other contemporaneous archaeological sites, no evidence of other function for those shell species in the archaeological record.
L599 3) are waiting to be exchange? 4) Personal collection? Yes, many uses and functions of perforated shells may be guessed. We prefer to keep a parsimonious hypothesis concerning their use as personal ornaments, without mentioning any possible individual or inter-individual functions that would be pure speculation.
L593-599 Have you got data that can may the difference between a storage or a personal collection? Are there concentrated shells in context? Were all species mixed at Rochereil when they were discovered? Maybe a reference that can interested your paper. It is about unmodified shells. There is a mention to the Taborin’s observation that is linked to your paper. Dupont C., 2019- Archaeological evidence for collecting empty shells along the French Atlantic coast: A major activity for coastal populations. Journal of Ethnobiology. 39(2), 223-239, https://doi.org/10.2993/0278-0771-39.2.223
thank you for the reference, it has been added to the ‘objectives’ section.
This is a very interesting paper that shows clear differences between the assemblages shown in the paper and our present study. For the wall decoration, no selection by the size is demonstrated and only perforations from a taphonomical origin are present in few specimens. moreover the shell species used for wall decoration are mostly not used for personal ornamentation according to the litterataure. At rochereil we have evidence of human made modification on the T. reticulata,, selection of the dentalium by their size, and the use of similar species at other contemporaneous sites for personal ornamentations.
For the child ‘Tresor’, the malacospectra is more diverse than at Rochereil, with few specimens per species, and it includes many shells not commonly used as personal ornaments in the archaeological record.
Moreover, the author is using the context of discovery, including precise spatial data for the Tresor and pieces of walls with shells still embedded in the mortar. We cannot expect such data for paleolithic sites and when considering mobile foragers. That is why here we provide more precise technological, morphometric data and regional comparisons to support our claim, data not provided in the Dupont 2019 paper.

Conclusions
L644 “storage” How do you define this term after your data? storage has been removed from the text
L646 “before modification”: necessary? we have keep it as we insist on the arguments provided all along the text to justify the shells were dedicated to be used as personal ornaments
L 657 ‘embedded caching’ How do you define this term after your data? removed from the text
L659 “a deviation in normal” excessive: what is normality? Can’t we consider populations as a diversity? Deviation is a negative word for me, a value judgment. change for diversion

References
L504 Taborin 1993: reference?,fixed
L632 Taborin 2007 reference? fixed
L652 White 1997: reference? fixed
L712/ L723/ L728-729/ L741 italic fixed
L826 & ?fixed
L848 Sandrine ? fixed
L936 abbreviation? fixed
References of a same author are not in a chronological order fixed

Reviewed by Lawrence Straus, 19 May 2022 06:50
My only suggestions are that the authors cite and use the seminal works of Polly Wiessner on the hxaro networks of southern Africa (shells traded over long distances from the coasts by San foraging peoples), References added
that they provide more information on why Rochereil might have been a central node (and important place) in such networks (because of geographical position, avenues of communication, physical characteristics of the cave vis a vis other Upper Magdalenian sites in this core area of UM settlement in SW France), this sentence has been removed
and that they explain why the taxonomic name Dentalium is used (I had thought it had been replaced by Antalis---a footnote is in order explaining this). text modified