the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Oct 2023
Melt-enhanced strain localization and phase mixing in a large-scale mantle shear zone (Ronda peridotite, Spain)
Jolien Linckens
Gernold Zulauf
Download
- Final revised paper (published on 26 Oct 2023)
- Supplement to the final revised paper
- Preprint (discussion started on 02 Jan 2023)
- Supplement to the preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1348', Jacques Précigout, 17 Feb 2023
Dear Editors,
This paper deals with the petrological and micro-structural features of the large-scale shear zone that affected the western Ronda peridotite massif (Southern Spain). Based on a significant amount (35) of samples across the shear zone, including most of them in the mylonitic domain (30) and 5 samples in the tectonite domain nearby the mylonite, the authors provide detailed documentations of the chemical and micro-structural features of the peridotites, focusing on secondary phases (mostly pyroxenes). The paper is well written, correctly organized and up to date. I also think that the data deserve to be published, but I cannot recommend this paper to published in its present form, mostly because the data are over-interpreted.
Indeed, the authors give evidence of melt infiltration and melt-rock interactions through the whole shear zone, including the mylonitic domain, and they attribute strain localization to this presence of melt during deformation. I do not question that melt infiltration may have occurred through the peridotites that compose the Ronda shear zone, but the authors do not provide any feature or data that suggest syn-kinematic infiltrations. The process of refertilization described here could thus be pre-kinematic only. Furthermore, if melt was involved to localize strain in Ronda, we should see a significant difference in modal composition of the peridotites (i.e., the amount of pyroxenes) between the mylonite and tectonite. But the authors - and previous papers, including mine - did not document any difference. Finally, the gradient of Mg# documented here across the shear zone is very interesting, but first, this feature is only based on 3 samples (out of 35), and in case of strain localization induced by melt infiltration, we would expect increasing melt-rock interactions (i.e., a decreasing Mg#) with increasing strain. The authors show the opposite.
What I could suggest to improve the paper is: 1) better synthesizing the micro-structural features, documentations of which are a bit unbalanced with respect to chemical features, 2) strengthening the point of Mg# gradient by performing chemical analyses on more samples (this would help to better document the melt-rock interactions front), and 3) reconsidering the main axis/interpretation of the paper by focusing on the evidence of melt-rock interactions front across the shear zone, and not speculating - although it could be discussed - on the role of melt in triggering strain localization in Ronda (based on the data presented here, it is not plausible). I therefore encourage the authors to reconsider their paper in this perspective, so that they give more chance to their study to be published in the near future.
With my Best Regards,
Jacques Précigout
Minor comments to the authors
Geological setting: You could be interested in having a look at the two papers of Bessière et al. (2021) that expand on the geodynamic of the Ronda peridotite.
Line 142: underlying not underlaying
Line 159: Citing a paper rather than the phd thesis of Dirk Van der Wal would be more appropriate here. And I think some other hypothesis (and references) need to be mentioned, including the one described in our paper (Précigout et al., 2013).
Line 167: were, not was.
Line 202: 100 grains is not enough to calculate a J or Mindex (Skemer et al., 2005). You could also see our recent paper dealing with this feature (Précigout et al., 2022, sci. rep.)
Line 204: We commonly use 10° as halfwidth angle. Otherwise, it smoothes a lot the data.
Line 223: Plane, not plain
Line 242: This feature has been already described in Précigout et al. (2013), so it should be cited here.
Figure 4a: Where are the olivine dots?
Figure 4: the olivine-rich matrix represents the major part of the peridotite, so you cannot exclude it from the grain size dataset, whatever the reason.
Line 334: B-type fabric has been documented in Précigout et al. (2014), so it has to be mentioned here.
Figure 6: The number of grains (or datapoint) has to be shown by each pole figure.
Line 387: what do you mean by « tend to be higher »? The AR is higher or not.
Figure 8A: Avoid writing labels up side down.
Line 527: Discussing about processes of mantle refertilization, the paper of Le Roux et al. (2007) should be discussed, at least cited, somewhere.
Line 539: what do you base on to say that this CPO is atypical ? Is there any reference that mention that.
Line 581: when you discuss a feature that is not described in your paper, you have to cite the publication that describe it. For instance, boudinaging of pyroxenite layers has been described in Précigout et al., 2013.
Line 652: Saying constant grain size is not correct here, but constant dynamically recrystallized grain size may be correct.
Figure 12 and 13: to be frank, your model is very difficult to understand based on the figures. Could you please make them more clear?
Line 701: what trend?
Citation: https://doi.org/10.5194/egusphere-2022-1348-RC1 - AC1: 'Reply on RC1', Sören Tholen, 01 May 2023
-
AC2: 'Reply on RC1', Sören Tholen, 01 May 2023
Dear Editors, dear Andréa Tommasi, dear Jacques Précigout,
First of all, the authors want to thank both reviewers, Andréa Tommasi and Jacques Précigout, for their detailed and constructive comments, which help to improve the manuscript. The authors decided to reply in one statement because the major remarks of both referees largely overlap or complement each other. In general, all proposed changes and comments of both reviewers were considered in the revised version of the manuscript. The changes tracked by line number are in the second part of this reply. However, at first we will briefly comment on the main points of criticism.
According to both reviewers, the three main points of criticism are:
- The nature of metasomatism: Refertilizing or late stage, fluid-rich melt?
- The timing of the metasomatism and its effect on deformation.
- The insufficient quantification of the olivine grain size.
The nature of metasomatism:
We agree with both reviewers, that the irregular grain/phase boundaries and grain shapes as well as extensive phase mixing are robust microstructural evidence for metasomatism in the entire NW Ronda shear zone. A metasomatism of parts of the investigated mylonite unit by refertilizing melts was postulated by Soustelle et al. (2009). This process was adapted in the original manuscript for our transect. However, both reviewers reject the interpretation of a refertilizing melt because of low syn-kinematic temperature estimates (800–900 °C, 1.95–2.00 GPa (Garrido et al. (2011)) and the missing modal change to increased fertile components. The authors agree with both objections and discuss the nature of the metasomatism in the reviewed version by taking into account the suggestions made by the reviewers. Following the annotations made by Andréa Tommasi, the microstructural similarity as well as the matching PT-estimations for the grt/spl-mylonites from Rondas counterpart from the Morrocan limb of the Gibraltar arc, the Beni Bousera peridotite massif, point to a consistent genesis. Frets et al. (2012, 2014) suggested a metasomatism of small fractions of fluids or highly evolved melts, which did not reset the quilibrium temperatures in Beni Bousera. Matching all observations made in our samples, the authors agree, that the metasomatism is most likely attributable to highly evolved melt. A fluid-driven metasmomatism as proposed for the plagioclase-tectonite unit in Ronda by Hidas et al. (2016) is in the authors opinion less likely because of the low abundance of amphibole in the dominant mixed matrix and the absence of ultramylonites, which were obsereved to form by fluid-rock reactions. Based on these observations, the rewritten section 5.1 “Microstructural implications – Formation” now includes a discussion and evaluation of the different potential metasomatic agents. In this regard, the authors agree with Jacques Précigouts remark of the small geochemical data base for a geochemically based model of the shear zone’s evolution. To resolve the geochemical trend in detail, an additional study would be needed with the focus on the transitional area between the mylonite and tectonite unit. According to Jacques Précigouts suggestions, the reviewed discussion was focused on the microstructural evidence.
The timing of the metasomatism and its effect on deformation:
The rewritten section 5.2 “Microstructural implications – Deformation” now discusses the timing of the metasomatic event and its effect on the deformation. Microstructural similarities especially of the film/wedge-shaped orthopyroxenes in the mylonitic part of the shear zone to mylonites and ultramylonites investigated by Dijkstra et al. (2002) and Hidas et al. (2016) indicate a syn-kinematic metasomatism with dissolution-precipitation reactions being active. This assumption is supported by Frets et al. (2014), who argued for the corresponding grt/spl-mylonites of Beni Bousera for syn- to late kinematic metasomatism. As both reviewers criticize an overinterpretation of the data in terms of the importance of the metasomatic event for the genesis of the NW Ronda shear zone, the discussion was fundamentally shortened in this regard. Therefore, the main focus of section 5.2 lies now on the active deformation mechanisms (dislocation creep, dissolution-precipitation creep), the dominant deformation mechanism (dislocation creep) and the potential impact of phase mixing and melt presence on the deformation. The authors agree that an irrevocable argument for the trigger of the shear zone by metasomatic processes cannot be given. However, the comparison with other upper mantle shear zones (section 5.4) indicates a general strong relation between reactions and localized deformation in the upper mantle. With the data presented, the NW Ronda shear zone lines up or at least does not contradict this picture.
The insufficient quantification of the olivine grain size:
The complete data was reprocessed to quantify the original olivine grain size using the method suggested by Andréa Tommasi. The new data were added to the microstructural data of figure 3 and of supplementary data S2. However, even with a larger spread and coarser grain sizes, olivine follows the general trend of constant grain sizes in the entire mylonite unit formerly reported by Johanesen & Platt (2015). Moreover, Frets et al. (2014) report for the Grt/Spl-mylonite unit of Beni Bousera a similar range of mean olivine grain size (90-160 µm). The statistics of 7375 olivine grains analyzed in the mixed matrix and the consistency with the published data indicate a robust data set of constant olivine grain size over the entire mylonite unit with local variations but no obvious trend.
--------------------------------------------------------------------------------------------------------------------------
Detailed list of corrections for remarks of Jacques Précigout sorted by line numbers. In the authors answers the first line number refers to reviewed manuscript without changes marked, the second line number to the version with changes marked.
Jacques Précigout:
Additional comments:
1. Better synthesizing the micro-structural features, documentations of which are a bit unbalanced with respect to chemical features.
The complete discussion was rewritten with the focus on the microstructural implications. The authors agree, that for solid geochemical interpretation additional measurements are necessary on samples that are not present at the moment. The focus of the geochemical investigation should be on the transition between mylonites and tectonites.
2. Strengthening the point of Mg# gradient by performing chemical analyses on more samples (this would help to better document the melt-rock interactions front).
The authors decided to focus with this manuscript on the microstructural analysis of the mylonite unit. However, agreeing with the remark a future study in the geochemistry is planned.
3. Reconsidering the main axis/interpretation of the paper by focusing on the evidence of melt-rock interactions front across the shear zone, and not speculating - although it could be discussed - on the role of melt in triggering strain localization in Ronda (based on the data presented here, it is not plausible)
In accordance with the answer on the main points of criticism and the comments above the discussion was rewritten with the focus on the microstructural implications for the formation of and deformation in the mylonitic unit.
Minor comments to the authors
Geological setting: You could be interested in having a look at the two papers of Bessière et al. (2021) that expand on the geodynamic of the Ronda peridotite.
Thanks for the interesting suggestion.
Line 142: underlying not underlaying.
Changed (l. 143/ 160)
Line 159: Citing a paper rather than the PhD thesis of Dirk Van der Wal would be more appropriate here. And I think some other hypothesis (and references) need to be mentioned, including the one described in our paper (Précigout et al., 2013).
The authors decided to delete the emplacement hypothesis from the introduction as the focus lays on the microstructures and not on the overall tectonics (l. 160/ 183). As the reviewer annotates correctly otherwise more hypotheses should be discussed.
The reference was changed to the most relevant paper.
Line 167: were, not was.
Changed (l. 168/ 192)
Line 202: 100 grains is not enough to calculate a J or Mindex (Skemer et al., 2005). You could also see our recent paper dealing with this feature (Précigout et al., 2022, sci. rep.)
A minimum of 150 grains was set for the M- and J-Index (l. 202/ 226).
Line 204: We commonly use 10° as halfwidth angle. Otherwise, it smoothes a lot the data.
Thanks for the remark. Using 15° as halfwidth was adapted from previous studies (Tholen et al 2022, Linckens et al 2021). For the next dataset we will compare 10° and 15°.
For this dataset, the CPOs are mostly strong if present and therefore the changes in the dominant olivine CPO for example are easily recognizable.
Line 223: Plane, not plain.
Changed, thanks (l. 229/ 257)
Line 242: This feature has been already described in Précigout et al. (2013), so it should be cited here.
Citation was included (l. 244/ 276).
Figure 4a: Where are the olivine dots? The olivine-rich matrix represents the major part of the peridotite, so you cannot exclude it from the grain size dataset, whatever the reason.
Now figure 3A. As both reviewers requested to include the olivine grain size, we recalculated all EBSD data to reconstruct the original grain size. The reconstruction was performed using the proposed method in MTEX (Matlab) by Andréa Tommasi.
Line 334: B-type fabric has been documented in Précigout et al. (2014), so it has to be mentioned here.
Citation was added (l. 354/ 403).
Figure 6: The number of grains (or datapoint) has to be shown by each pole figure.
Numbers of grains and color bars for ODFs were added to all orientation figures.
Line 387: what do you mean by « tend to be higher »? The AR is higher or not.
Changed (l. 411/ 460).
Figure 8A: Avoid writing labels up side down.
Now figure 9. Label was flipped.
Line 527: Discussing about processes of mantle refertilization, the paper of Le Roux et al. (2007) should be discussed, at least cited, somewhere.
Refertilization was discussed in section 5.1.1. The citation was added.
Line 539: what do you base on to say that this CPO is atypical ? Is there any reference that mention that.
Reference was added (l. 690/ 747). The dominant CPOs for opx in deformed mantle rocks commonly have [001] in the lineation.
Line 581: when you discuss a feature that is not described in your paper, you have to cite the publication that describe it. For instance, boudinaging of pyroxenite layers has been described in Précigout et al., 2013.
Very correct. Thanks for the annotation, citation was added (l. 628/ 684).
Line 652: Saying constant grain size is not correct here, but constant dynamically recrystallized grain size may be correct.
Section was rewritten (now 5.2)
Figure 12 and 13: to be frank, your model is very difficult to understand based on the figures. Could you please make them more clear?
Complete discussion was rewritten with the focus on the microstructural implications. The model was dismissed.
Line 701: what trend?
See comment above.
-
RC2: 'Comment on egusphere-2022-1348', Andrea Tommasi, 06 Mar 2023
Review of ms. Melt-enhanced strain localization and phase mixing in a large-scale mantle shear zone (Ronda peridotite, Spain) by Tholen et al.
This ms. presents microstructural (EBSD) and mineral composition (EPMA) data for peridotite mylonites along a transect across the (structurally) upper section of the Ronda peridotite massif in southern Spain. The deformation that formed these mylonites resulted in extensive development of a fine-grained well-mixed polyphase matrix composed of olivine-orthopyroxene-clinopyroxene±spinel±amphibole. The present study focusses on this matrix, aiming to unravel the processes involved in its formation and their impact on strain localization.
Based on morphological characteristics of the microstructure of the polyphase matrix, on the decrease in Mg# of all phases and of Ti contents of pyroxenes across the transect and decrease in Cr content between orthopyroxene porphyroclasts and matrix at the local scale, the authors propose that:
(1) The development of this matrix resulted from pre- to syn-deformational melt infiltration from the structurally deeper melting (or recrystallization) front, which separates the tectonites from the coarse-granular peridotites in the Ronda massif.
(2) Strain localization was triggered by melt-enhanced deformation and grain size reduction by crystallization of interstitial pyroxenes, leading to subsequent activation of a grain size-sensitive deformation mechanism.
The microstructures, in particular the highly lobate phase boundaries and the wedge or film-like shapes along the foliation plane of pyroxenes, undoubtedly point to a major role of dissolution-precipitation processes on the formation of the polyphase fine-grained matrix in these peridotites. However, the interpretation that the fluid allowing for these reactions is a melt producing refertilization comes against the equilibration conditions recorded for these rocks. Previous studies on the same mylonites, in particular Garrido et al. (2011), show solid evidence for equilibrium conditions that evolved from prekinematic minimum conditions of 2.4–2.7 GPa and 1020–1100 °C to early syn-kinematic conditions of 800–900 °C and 1.95–2.00 GPa, with no record of a heating event neither predating nor synchronous to the deformation. Under such conditions, which are definitely sub-solidus, a "normal" mantle-derived melt cannot move by porous flow over large distances (here hundreds of meters) without crystallizing.
The geochemical arguments for refertilization of the studied mylonites are also not robust. There is no evidence for net increase in basaltic components in whole-rock or modal compositions. The observed compositional changes, in particular the decrease in Ti and Fe contents and increase in Mg# towards the uppermost structural levels of the massif, are inconsistent with reactional percolation of a basaltic melt leading to refertilization (cf. Leroux et al. 2007, Bodinier and Godard 2014
By consequence, the fluid phase responsible for the dissolution-precipitation reactions for which there is abundant evidence in these mylonites cannot be a basaltic melt, similar to that proposed to explain the refertilization just atop of the recrystallization front by Soustelle et al. (2009). It has to be either an aqueous fluid or a highly evolved volatile-enriched small melt fraction derived from the partial crystallization of basaltic melts in the structurally lower parts of the shear zone. Indeed, the observed microstructures are very similar to those described in shear zone that crosscuts the plagioclase-tectonite domain of the Ronda peridotite, in which orthopyroxene geothermometry indicated deformation under similar temperatures, which were interpreted by Hidas et al. (2013), based on microstructural and geochemical evidence, as produced by dissolution-precipitation mediated by aqueous fluids.
It is noteworthy that similar microstructures were already described and interpreted by Frets et al. (2013) as resulting from dissolution-precipitation processes in response to syn- to late kinematic reactive percolation of small fraction of fluids or highly evolved melts in the garnet and spinel mylonites of the upper section of the Beni Bousera peridotite massif. In this massif, which is the counterpart of the Ronda peridotite in the Rif belt (the continuation of the Betics on the other side of the Alboran arc), the exposure conditions allowed to better document the continuity of the structures and variation of synkinematic pressure and temperature conditions and hence to unravel that the entire peridotite massif records the functioning of a transtensional shear zone, a few kilometers wide, which accommodated exhumation of the base of the lithosphere from 90 to 60 km depth (Frets et al. 2013). Fast decompression of the footwall resulted in partial melting in the lower structural levels (the coarse-grained domain in both peridotite massifs), without the need for exotic heat sources. The strong thermal gradient across the shear zone was preserved because deformation (advection) is much faster than heat diffusion. Reactive percolation with partial crystallization of the decompression melts into the upper levels of the shear zone accounts for the observed refertilization features in the central section of the shear zone, the tectonite domain (e.g., Soustelle et al. 2009 for Ronda) and reactive percolation of small fractions of highly evolved volatile-rich melts or fluids derived from these reactions may explain development of extensive dissolution-precipitation without any evidence for heating in the uppermost levels of the shear zone, that is, the microstructures observed in the spinel and garnet mylonites of both Beni Bousera and Ronda massifs.
My second concern relates to the data. The authors put a lot of emphasis on the grain size determinations of the neoblasts by Johanesen and Platt (2015). However, these data are based on optical redrawing of a limited number of thin sections of partially serpentinized peridotites. EBSD data, such as that used by Precigout et al. (2007) for Ronda and Frets et al. (2013) for the equivalent units in Beni Bousera is more reliable, since it allows to use crystal orientations to identify the grains, despite fracturing and serpentinization, and larger areas are analyzed. Both previous EBSD-based studies deduced a variation in neoblasts grain size. And, here comes to a point that I really do not understand: why do the authors do not use the advanced grain reconstruction tools developed in the MTEX toolbox (see https://mtex-toolbox.github.io/EBSDFilling.html) to determine the olivine neoblasts grain sizes in their EBSD maps? Even if there will be uncertainty in such values for highly serpentinized peridotite samples, this approach would provide first-order estimates of the neoblast sizes in all analyzed maps, which will certainly be more precise than those derived from the analysis of optical micrographs. A last important remark concerning the EBSD data in the present study: in most, if not all, maps shown in the article, the areas analyzed are too small to represent representative elementary volumes. In the same line, in contrast to what is stated in line 251 of the ms., twenty grains are not enough for a statistical analysis of grain sizes, shapes, or orientations.
The last major comment concerns the interpretations. The presence of fluids certainly facilitated the deformation as it allowed for activation for additional deformation processes: dissolution-precipitation creep and probably locally lubricated grain boundaries, easing grain boundary sliding. However, there is no clear evidence justifying that the fluids triggered (1) strain localization and (2) transition from dislocation creep to a deformation dominated by grain-size sensitive processes. There is no clear evidence for fluids percolation predating the formation of the mylonites and as the authors clearly document, even in the fine-grained polyphase matrix, the clear crystallographic preferred orientations point to significant contribution of dislocation creep to the deformation.
Additional comments:
- The introduction and discussion sections have repetitions and may be significantly shortened, so that there will be more space to present the data, which in the present form of the ms. is largely presented as Supplementary material. The section of the amphibole-bearing veins is also not essential to the article.
- Please add a map of the Sierra Bermeja massif with foliations and lineations. Even if you focus on the microstructures, the structural context is important. In general, the description of the structural data is too vague. The orientation of the mylonitic foliations is much more varied then stated in l. 115 - their trend follows on average that of the limits between the tectonometamorphic domains, see maps from Darot (1973) and data reported in later studies (Obata, Van der Wal et al 1993, 1996, Soustelle et al. 2009…). Same for the lineations.
- The description of the sampling referring to the shear zone boundary is not always clear. Better state that the samples were collected at increasing distance from the northeastern limit of the massif. This limit is not necessarily the limit of the shear zone as the contact between the peridotites and the Jubrique unit may have been reworked. Similarly, in line 371, the use of distal may lead to confusion.
- How do the area fractions of porphyroclasts and matrix vary as a function of distance along the transect? This information might allow to better evaluate the continuity (or not) of the evolution of the deformation conditions along the transect.
- How are the limits between porphyroclasts tails and the matrix defined? Is it really important to discriminate between these two microstructural domains?
- To discuss the variations in olivine CPO patterns along the transect as it is done lines 331-335 and 550-555, the full dataset needs to be presented in the article. It is stated in the text that A-type patterns dominate. Yet most figures presented in the main text show AG-type patterns.
- Deformation mechanisms in the matrix: please complete this point by showing and discussing the internal deformation of the neoblasts... If they deformed by dislocation creep, as stated in the ms., they should display, to some extent, a substructure (GNDs) consistent with this deformation.
- CPOs in the recrystallization tails: do the observations hint for inheritance of orientations from the porphyroclasts?
- Line 558: What are the observations that indicate that deformation was enhanced by the presence of melts in the early stages of shearing? Why early stages?
- Line 559: Piezometric data cannot document the activation of a grain size sensitive mechanism. At best, given all the uncertainty and hypotheses inherent to this method, it allows an estimate of the active stresses. And this estimate is only valid if grain size reduction is controlled by dislocation creep, since these are the conditions prevailing in the experiments used for the calibration.
- Line 560: Where does the evidence for GBS is shown?
- Not all pyroxenites in Ronda were interpreted as resulting from replacement of previous gt-pyroxenites by melt-rick reaction. An important volume of pyroxenites was interpreted as formed by partial (reactive) crystallization of percolating melts. Moreover, in the mylonites, gt-pyroxenites predominate (Garrido and Bodinier 1999).
- Referencing is imprecise in some places. For instance, Passchier and Trow (1996) is not the best citation for viscoplastic anisotropy due to crystal orientation.
- Lack of cross-cutting relations between gt-mylonites and sp-tectonites was also reported by Soustelle et al (2009).
Minor comments / questions:
- In all figures presenting CPO the color bars indicating the intensities of the contours are missing.
- In fig. 2 it is impossible to see the elongated opx porphyroclasts. Yet they are clearly visible in the field. Moreover, due to serpentinization and fine-grained nature of these peridotites, to define in the field variations in mineralogical composition is very difficult. Is this figure really useful?
- In figure 3, the microstructure of the tectonite is not visible.
- Line 150: the peridotite solidus is not a temperature, it depends on temperature and pressure, and composition, volatiles…
- What are the arguments (=observations) used to define an intergrowth between olivine and pyroxenes (line 313)?
- What do the arrows in Fig. 12 mean? In Soustelle et al. (2009), only in two samples the analyzed pyroxenes were clearly identified as secondary, that is, resulting from partial crystallization from melts. In addition, the area concerned by this study is not indicated in Fig.1 as stated in the figure caption.
Citation: https://doi.org/10.5194/egusphere-2022-1348-RC2 -
AC3: 'Reply on RC2', Sören Tholen, 01 May 2023
Dear Editors, dear Andréa Tommasi, dear Jacques Précigout,
First of all, the authors want to thank both reviewers, Andréa Tommasi and Jacques Précigout, for their detailed and constructive comments, which help to improve the manuscript. The authors decided to reply in one statement because the major remarks of both referees largely overlap or complement each other. In general, all proposed changes and comments of both reviewers were considered in the revised version of the manuscript. The changes tracked by line number are in the second part of this reply. However, at first we will briefly comment on the main points of criticism.
According to both reviewers, the three main points of criticism are:
- The nature of metasomatism: Refertilizing or late stage, fluid-rich melt?
- The timing of the metasomatism and its effect on deformation.
- The insufficient quantification of the olivine grain size.
The nature of metasomatism:
We agree with both reviewers, that the irregular grain/phase boundaries and grain shapes as well as extensive phase mixing are robust microstructural evidence for metasomatism in the entire NW Ronda shear zone. A metasomatism of parts of the investigated mylonite unit by refertilizing melts was postulated by Soustelle et al. (2009). This process was adapted in the original manuscript for our transect. However, both reviewers reject the interpretation of a refertilizing melt because of low syn-kinematic temperature estimates (800–900 °C, 1.95–2.00 GPa (Garrido et al. (2011)) and the missing modal change to increased fertile components. The authors agree with both objections and discuss the nature of the metasomatism in the reviewed version by taking into account the suggestions made by the reviewers. Following the annotations made by Andréa Tommasi, the microstructural similarity as well as the matching PT-estimations for the grt/spl-mylonites from Rondas counterpart from the Morrocan limb of the Gibraltar arc, the Beni Bousera peridotite massif, point to a consistent genesis. Frets et al. (2012, 2014) suggested a metasomatism of small fractions of fluids or highly evolved melts, which did not reset the quilibrium temperatures in Beni Bousera. Matching all observations made in our samples, the authors agree, that the metasomatism is most likely attributable to highly evolved melt. A fluid-driven metasmomatism as proposed for the plagioclase-tectonite unit in Ronda by Hidas et al. (2016) is in the authors opinion less likely because of the low abundance of amphibole in the dominant mixed matrix and the absence of ultramylonites, which were obsereved to form by fluid-rock reactions. Based on these observations, the rewritten section 5.1 “Microstructural implications – Formation” now includes a discussion and evaluation of the different potential metasomatic agents. In this regard, the authors agree with Jacques Précigouts remark of the small geochemical data base for a geochemically based model of the shear zone’s evolution. To resolve the geochemical trend in detail, an additional study would be needed with the focus on the transitional area between the mylonite and tectonite unit. According to Jacques Précigouts suggestions, the reviewed discussion was focused on the microstructural evidence.
The timing of the metasomatism and its effect on deformation:
The rewritten section 5.2 “Microstructural implications – Deformation” now discusses the timing of the metasomatic event and its effect on the deformation. Microstructural similarities especially of the film/wedge-shaped orthopyroxenes in the mylonitic part of the shear zone to mylonites and ultramylonites investigated by Dijkstra et al. (2002) and Hidas et al. (2016) indicate a syn-kinematic metasomatism with dissolution-precipitation reactions being active. This assumption is supported by Frets et al. (2014), who argued for the corresponding grt/spl-mylonites of Beni Bousera for syn- to late kinematic metasomatism. As both reviewers criticize an overinterpretation of the data in terms of the importance of the metasomatic event for the genesis of the NW Ronda shear zone, the discussion was fundamentally shortened in this regard. Therefore, the main focus of section 5.2 lies now on the active deformation mechanisms (dislocation creep, dissolution-precipitation creep), the dominant deformation mechanism (dislocation creep) and the potential impact of phase mixing and melt presence on the deformation. The authors agree that an irrevocable argument for the trigger of the shear zone by metasomatic processes cannot be given. However, the comparison with other upper mantle shear zones (section 5.4) indicates a general strong relation between reactions and localized deformation in the upper mantle. With the data presented, the NW Ronda shear zone lines up or at least does not contradict this picture.
The insufficient quantification of the olivine grain size:
The complete data was reprocessed to quantify the original olivine grain size using the method suggested by Andréa Tommasi. The new data were added to the microstructural data of figure 3 and of supplementary data S2. However, even with a larger spread and coarser grain sizes, olivine follows the general trend of constant grain sizes in the entire mylonite unit formerly reported by Johanesen & Platt (2015). Moreover, Frets et al. (2014) report for the Grt/Spl-mylonite unit of Beni Bousera a similar range of mean olivine grain size (90-160 µm). The statistics of 7375 olivine grains analyzed in the mixed matrix and the consistency with the published data indicate a robust data set of constant olivine grain size over the entire mylonite unit with local variations but no obvious trend.
--------------------------------------------------------------------------------------------------------------------------
Detailed list of corrections for comments by Andréa Tommasi, sorted by line numbers. In the authors answers the first line number refers to reviewed manuscript without changes marked, the second line number to the version with changes marked.
Additional comments:
1. The introduction and discussion sections have repetitions and may be significantly shortened, so that there will be more space to present the data, which in the present form of the ms. is largely presented as Supplementary material. The section of the amphibole-bearing veins is also not essential to the article.
Both sections were revised and reworked. Parts were shortened and the complete CPO data was included in the results section. The authors decided to leave the section on amphibole-bearing veins (4.1.3) in the manuscript to make readers aware of this feature which is potentially interesting to investigate late-stage fluid-peridotite interaction and was not described so far.
2. Please add a map of the Sierra Bermeja massif with foliations and lineations. Even if you focus on the microstructures, the structural context is important. In general, the description of the structural data is too vague. The orientation of the mylonitic foliations is much more varied then stated in l. 115 - their trend follows on average that of the limits between the tectonometamorphic domains, see maps from Darot (1973) and data reported in later studies (Obata, Van der Wal et al 1993, 1996, Soustelle et al. 2009…). Same for the lineations.
A structural map of the Sierra Bermeja massif including foliation, lineations and major faults was added to figure 1. The studied area of Soustelle et al. 2009 was indicated in this map. The section on the structural data was rewritten to clarify the variations in the foliation/lineation and the dominant orientation of both in the area of investigation (ll. 139/ 156, ll. 225/ 251).
3. The description of the sampling referring to the shear zone boundary is not always clear. Better state that the samples were collected at increasing distance from the northeastern limit of the massif. This limit is not necessarily the limit of the shear zone as the contact between the peridotites and the Jubrique unit may have been reworked. Similarly, in line 371, the use of distal may lead to confusion.
Shear zone boundary (SFZ) was changed to NW boundary of the Ronda peridotite massif (NW-B). The abbreviation was necessary for graph axes titles etc.. The use of “proximal” and “distal” was avoided in the complete manuscript.
4. How do the area fractions of porphyroclasts and matrix vary as a function of distance along the transect? This information might allow to better evaluate the continuity (or not) of the evolution of the deformation conditions along the transect.
Descriptions and discussions of the variations for the abundances of porphyroclasts and the proportion of recrystallized matrix respectively are added in lines 392/ 440, 435/ 485 and 653/ 709.
5. How are the limits between porphyroclasts tails and the matrix defined? Is it really important to discriminate between these two microstructural domains?
Neoblast tails of pyroxene porphyroclasts are characterized by a phase (pyroxene dominated, amphibole-bearing), grain shape (equaxial), grain size (coarse) composition and CPOs (AG- and B- type). All these microstructural characteristics differ distinctly from the surrounding mylonitic matrix with ol-dominated, mostly amph-absent composition, elongated grain shape, smaller grain size and strong A-type CPOs. Therefore, their limits are defined by all these microstructural parameters which enable an easy distinguishment between matrix and tails.
6. To discuss the variations in olivine CPO patterns along the transect as it is done lines 331-335 and 550-555, the full dataset needs to be presented in the article. It is stated in the text that A-type patterns dominate. Yet most figures presented in the main text show AG-type patterns.
The complete CPO data was added as figure 4.
7. Deformation mechanisms in the matrix: please complete this point by showing and discussing the internal deformation of the neoblasts... If they deformed by dislocation creep, as stated in the ms., they should display, to some extent, a substructure (GNDs) consistent with this deformation.
A GND reconstruction map for the mylonitic mixed matrix was added in figure 7 corroborating dislocation creep as dominant deformation mechanism
8. CPOs in the recrystallization tails: do the observations hint for inheritance of orientations from the porphyroclasts?
Yes, there is a strong inheritance for opx neoblasts and a slightly weaker one for cpx. Description and discussion were added in lines 418/ 468, 464/ 515 and 677/ 734.
9. Line 558: What are the observations that indicate that deformation was enhanced by the presence of melts in the early stages of shearing? Why early stages?
For the discussion of the timing of the metasomatism please see above in the discussion of the main points of criticism. The potential effects of melt-presence in the studied rocks on the deformation and its microstructural implications are discussed in reworked section 5.2. The authors therein agree, that the initiation of strain localization cannot be conclusively attributed to the investigated metasomatic event. However, the comparison to other upper mantle shear zone shows a strong association of reactions, phase mixing and shear zones.
10. Line 559: Piezometric data cannot document the activation of a grain size sensitive mechanism. At best, given all the uncertainty and hypotheses inherent to this method, it allows an estimate of the active stresses. And this estimate is only valid if grain size reduction is controlled by dislocation creep, since these are the conditions prevailing in the experiments used for the calibration.
Correct, the paragraph was changed accordingly (ll. 722/ 779) .
11. Line 560: Where does the evidence for GBS is shown?
Evidence for GBS is tricky and in most cases no distinctive feature for GBS. Therefore, we refer to the research of Précigout et al. (2007) who argued for DisGBS as dominant deformation mechanism (ll. 724/ 782).
12. Not all pyroxenites in Ronda were interpreted as resulting from replacement of previous gt-pyroxenites by melt-rick reaction. An important volume of pyroxenites was interpreted as formed by partial (reactive) crystallization of percolating melts. Moreover, in the mylonites, gt-pyroxenites predominate (Garrido and Bodinier 1999).
Very correct. Thanks for the annotation, the text was changed accordingly (ll. 619/ 674).
13. Referencing is imprecise in some places. For instance, Passchier and Trow (1996) is not the best citation for viscoplastic anisotropy due to crystal orientation.
Citations have been checked and updated (e.g., l. 44/ 52). The authors would like to thank both reviewers for their paper suggestions.
14. Lack of cross-cutting relations between gt-mylonites and sp-tectonites was also reported by Soustelle et al (2009).
Thanks for the hint, the citation for Soustelle et al. (2009) was added (l. 128/ 145).
Minor comments / questions:
- In all figures presenting CPO the color bars indicating the intensities of the contours are missing.
Numbers of grains and color bars for ODFs were added to all orientation figures.
- In fig. 2 it is impossible to see the elongated opx porphyroclasts. Yet they are clearly visible in the field. Moreover, due to serpentinization and fine-grained nature of these peridotites, to define in the field variations in mineralogical composition is very difficult. Is this figure really useful?
Figure 2 was dismissed.
- In figure 3, the microstructure of the tectonite is not visible.
Contrast of tectonite overview has been increased in figure 2.
- Line 150: the peridotite solidus is not a temperature, it depends on temperature and pressure, and composition, volatiles…
Of course. In this regard it is just a citation of the estimated T conditions. P conditions were added (l 152/ 176).
- What are the arguments (=observations) used to define an intergrowth between olivine and pyroxenes (line 313)?
Highly lobate boundaries to bordering olivines and weird shaped protrusions (see figure 7; l 333/ 379).
- What do the arrows in Fig. 12 mean? In Soustelle et al. (2009), only in two samples the analyzed pyroxenes were clearly identified as secondary, that is, resulting from partial crystallization from melts. In addition, the area concerned by this study is not indicated in Fig.1 as stated in the figure caption.
Arrow description was added in figure 13. In figure 1 area of Soustelle et al. (2006) was added.
Peer review completion
