Niobium Mineralogy of Pliocene A1-Type Granite of the Carpathian Back-Arc Basin, Central Europe
Round 1
Reviewer 1 Report
Report on the manuscript “Niobium mineralogy of Pliocene A1-type granites of Central Europe” by Huraiova M et al.
1. General comments
The manuscript addresses the theoretical and practical implications of the presence in the Pannonian Basin mafic lavas of exceptional granitic xenoliths. The topic is of great interest and the manuscript presents an important contribution, which would be worth a publication in Minerals.
Unfortunately, beside a series of minor points (see Detailed comments below), the manuscript suffers from major defects which should be mended before any definitive acceptance. First, a solid Geological Setting section is lacking, and the summary of previous works on the same object is too short. Second, there are problems in relation to the partially molten nature of the xenoliths that are not properly addressed. Third, the discussion does not go enough in depth when petrogenesis of the granitic xenoliths is considered. I have made suggestions in the Detailed comments.
I thus conclude to acceptance with major revision.
2. Detailed comments
Introduction
L43: Partial or complete melting of an alkali-metasomatized (fenitized) crust has been proposed by Martin (2006). The inverse process of fractionation, i.e., very low degree of partial melting of a mafic rock (like an underplated basalt) has also been advocated (e.g., Kovalenko et al., 2009). You may also refer to the synthesis by Bonin (2007) or to Frost and Frost (2010) for a discussion of A-type classification.
Bonin B (2007) A-type granites and related rocks: Evolution of a concept, problems and prospects. Lithos 97:1-29.
Frost CD, Frost RB (2010) On ferroan (A-type) granitoids: their compositional variability and modes of origin. J Petrol 52:39-53.
Kovalenko VI, Naumov VB, Girnisa AV, Dorofeev VA, Yarmolyuk VV (2009) Peralkaline Silicic Melts of Island Arcs, Active Continental Margins, and Intraplate Continental Settings: Evidence from the Investigation of Melt Inclusions in Minerals and Quenched Glasses of Rocks. Petrology 17:410-428.
Martin RF (2006) A-type granites of crustal origin ultimately result from open-system fenitization-type reactions in an extensional environment. Lithos 91:125-136.
Materials and methods
It is rather unusual to present in detail the petrography and other characteristics of the studied samples in this kind of section: a specific section must be devoted to this, for instance a "Previous work" section, since what is presented here seems to have been already published. The better would be a short “Geological setting” section. You cannot rely only on past publications. In addition, you should give more information on the granite chemistry, in order to compare with the melt inclusions you find into fergusonite, on one hand, and to introduce the fractionation processes you are referring to in your Discussion section.
L71: From a terminological point of view, the presence of "clusters of interstitial glass" (whatever it may mean, please be more explicit) may rend the use of the name "granite" a little bit problematic: unless these are enclaves (but why would have they remained liquid?); it does suggest that the initial melt was not fully crystallized (or did he remelted, with magnetite restites? This point should be clarified) and the rock should consequently be named a porphyritic rhyolite ...
L73: Measurement of the full HFSE spectrum is indeed difficult, but is nevertheless currently effected in all good quality laboratories - no need to pretend here to have done a break. Yet, the detail of analytical procedure, as presented here, is indeed useful (but could be transferred to the Supplementary Material).
Results
Discriminating Nb-bearing phases: possibly better“Characterization of Nb-bearing phases”
L 90: Is there any indication of the relative chronology between these phases and/or with the major rock-forming minerals? From figure 2, it may be deduced that at least some of the accessory phases directly grew in the melt at an early stage of the liquid line of descent, and this should be emphasized. However, your statement in L154-sq suggests that things are more complicated, and a clarification would be appreciated. This is important if a comparison with other rare metal granites must be done, as, likely, many interested readers would intend to do.
Figure 2: Please hold to the IMA abbreviations for minerals. "Fer" is for feruvite and cannot be used for fergusonite. The correct abbreviation for magnetite is "Mgt", and for zircon, “Zrn”. Qtz, Kfs and Ilm are OK. When no abbreviation has been coined by IMA, you may feel free to coin yours, and thus "Col" and “Pyr” are acceptable. Note, however, that IMA favours the use of consonantic abbreviations: I personally use "Clb" for columbite, and I would suggest "Frg" for fergusonite.
L100: In comment b. “fergusonite is adhered” is uncorrect; you may write “fergusonite sticks to”. It is a pity not to have imaged in detail the zircon with uranothorite inclusions.
L 102: In commentc., you note that zoning is caused by U, Ti and Nb, whereas in the text the latter is not mentioned?
L 104: You speak of “myrmekitic rutile”, but in the figure, you write “glass + rutile”: myrmekite is therefore not the right term, and in any case the figured assemblage (although too much reduced) is more akin a symplectic association. In the text, however, it is no longer question of “glass + rutile”, and there is consequently a problem of internal consistency! (see comments to L 182).
L 109: What are the elements involved in the zoning here?
L112: Here you need to make clear your procedure, i.e., you must explain that you make use of a discriminant diagram, and give a clue to read Figure 3 (what are the coordinates, what is the reference- not sufficient to give it in Fig. 3 caption, etc.).
L 129: More than two decimal digits in apfu is unwarranted, given the analytical uncertainties.
L 158: You should give the full analytical data (obtained by EPMA? What conditions? How did you manage the loss in alkalis under the beam?) and compare them with the bulk analysis of the whole granite.
L 162: Strictly speaking, Y is not a REE!
L163-sq.: Please give more details about columbite! You should give the variation interval of the Nb to Nb+Ta (Nb#) and Mn to Fe+Mn (Mn#) ratios (saving the reader to go to supplementary data), and the classical Nb# vs. Mn# diagram should be provided, allowing an interpretation of the oscillatory zoning characteristics.
L 178: You refer here to an “intimate association” of columbite with Ca-niobate: yet, in Fig. 2 you only present a pyrochlore/Ca-niobate assemblage (not cited in the text). An image of this “intimate association” is needed, allowing the reader to evaluate the significance of the assemblage.
L 182-sq.: Despite the “low abundance” (but how much “low”?) of rutile, this phase would deserve a more expanded treatment. Indeed, with such high Nb contents, we may be close to wodginite composition: did you check this possibility? More important, you should give more details on the rutile-ilmenite relationships. The symplectic association between rutile and ilmenite? (text) or glass? (Fig. 2) is worth a more detailed presentation, and would demand a figure at higher magnification than in Fig. 3, allowing a discussion of the significance of the texture. You assure that the symplectite replaces ilmenite: but, what is the evidence? In the same way, “rutile exsolution” is an interpretation: what is the evidence?
Substitution of Nb and REE: better“Crystal chemistry of Nb-bearing phases”!
Figure 4 alludes to “reaction rims” between ilmenite and columbite; yet, they have not been described in the preceding section! You also mention “compact rutile replacing early ilmenite”, which is not described altogether. You must evidently fill these gaps!
Discussion
L 256: What are the age and geodynamic context of these “other A-type Carpathian granites”?
L 229: It would be better to keep in line with the introduction, where you classify the granite as “A1 type”.
L 263: As mentioned above, you should have emphasized this primary magmatic characteristic in the Results section.
L 270: The presence of fractionation process in the xenolith suite should have been presented in a preceding section (see my comments to the Materials and Methods section).
L 294: Until this point, your demonstration seems to hold. But here, you avoid a series of serious difficulties. First, as seen in Fig. 1, the granite enclaves are true fragmented xenoliths, not the co-magmatic enclaves implied by your simple model directly linking the basalts to the granites. Second, as far as may be understood from your text, you are dealing with a differentiated suite of granite xenoliths, suggesting that the basalts dismembered an evolved granite body. Third, the xenoliths are in reality frozen granitic mushes, which is apparently contradictory with the second point. This third difficulty may be easily resolved if you can demonstrate that the granite enclaves were partially remelted in the basaltic melt – a process which would have been greatly favoured if the dismembered granite body was still at high subsolidus temperature (in line with the Zrn age you provide: but what if these Zrn crystallized from a newly formed melt resulting from remelting?). Evidently, your discussion must be refined here.
L 295-sq.: There is evidently another delicate point here. In fact, your granite xenoliths display contradictory features, mixing characteristics of crustal-derived melts and of peralkaline, mafic-derived, melts. An easy explanation would be to envisage a mixing process between an evolved A-type melt and a peralkaline melt, as found for instance in the Nb-rich Huangshan granite in SE China (Zhu et al., 2018). The mantle-derived characteristics of the Zrn are not contradictory with this model, if the Zrn is considered to be issued from the peralkaline component (a likely possibility, given the uranothorite inclusions you mention in the text). In fact, you could try to inventory the accessory minerals relationships in order to find evidence of a dual origin. From this point of view, the rutile relationships could appear decisive.
Conclusions
L 313: Here again, you could refer to the Huangshan granite (Zhu et al. 2018).
Zhu Z.Y, Wang R.C, Marignac C, Cuney M, Mercadier J, Che X.D, Lespinasse MY (2018) A new style of rare metal granite with Nb-rich mica: The Early Cretaceous Huangshan rare-metal granite suite, northeast Jiangxi Province, southeast China. Amer Mineral 103:1530-1544
Author Response
General commentsThe manuscript addresses the theoretical and practical implications of the presence in the Pannonian Basin mafic lavas of exceptional granitic xenoliths. The topic is of great interest and the manuscript presents an important contribution, which would be worth a publication in Minerals.
Unfortunately, beside a series of minor points (see Detailed comments below), the manuscript suffers from major defects which should be mended before any definitive acceptance. First, a solid Geological Setting section is lacking, and the summary of previous works on the same object is too short.
Answer: Both sections have been extended.
Second, there are problems in relation to the partially molten nature of the xenoliths that are not properly addressed.
Answer: We address the problem of partial melting in the additional table containing compositions of interstitial glass, melt inclusions trapped in fergusonite and bulk composition of the xenolith. We link the increased alkalinity of the interstitial melt with the partial melting of feldspars during uplift.
Third, the discussion does not go enough in depth when petrogenesis of the granitic xenoliths is considered.
Answer: discussion was extended
Detailed comments
Introduction
L43: Partial or complete melting of an alkali-metasomatized (fenitized) crust has been proposed by Martin (2006). The inverse process of fractionation, i.e., very low degree of partial melting of a mafic rock (like an underplated basalt) has also been advocated (e.g., Kovalenko et al., 2009). You may also refer to the synthesis by Bonin (2007) or to Frost and Frost (2010) for a discussion of A-type classification.
Bonin B (2007) A-type granites and related rocks: Evolution of a concept, problems and prospects. Lithos 97:1-29. discussed
Frost CD, Frost RB (2010) On ferroan (A-type) granitoids: their compositional variability and modes of origin. J Petrol 52:39-53. discussed, check year
Kovalenko VI, Naumov VB, Girnisa AV, Dorofeev VA, Yarmolyuk VV (2009) Peralkaline Silicic Melts of Island Arcs, Active Continental Margins, and Intraplate Continental Settings: Evidence from the Investigation of Melt Inclusions in Minerals and Quenched Glasses of Rocks. Petrology 17:410-428.
Martin RF (2006) A-type granites of crustal origin ultimately result from open-system fenitization-type reactions in an extensional environment. Lithos 91:125-136.
Answer: the works of Bonin and Frost have already been included in the original version, the paper by Martin supplemented by paper from Jiang was added in the revised one. I do not have access to the paper by Kovalenko et al. The discussion about A-type granites in the Introduction was extended and additional references have been added, including the reference of Frost et al. 2001 pertaining to geochemical classification of granites.
Materials and methods
It is rather unusual to present in detail the petrography and other characteristics of the studied samples in this kind of section: a specific section must be devoted to this, for instance a "Previous work" section, since what is presented here seems to have been already published. The better would be a short “Geological setting” section. You cannot rely only on past publications. In addition, you should give more information on the granite chemistry, in order to compare with the melt inclusions you find into fergusonite, on one hand, and to introduce the fractionation processes you are referring to in your Discussion section.
Answer: The paper was reorganized by the addition of Geological Setting section. Geological map has also been added. Granite chemistry has been compared with the fergusonite-hosted silicate melt inclusions in a new table. The concise background geological information in the previous version was governed by the need of maintaining the number of references below 60, as required in the Instruction for authors.
L71: From a terminological point of view, the presence of "clusters of interstitial glass" (whatever it may mean, please be more explicit) may rend the use of the name "granite" a little bit problematic: unless these are enclaves (but why would have they remained liquid?); it does suggest that the initial melt was not fully crystallized (or did he remelted, with magnetite restites? This point should be clarified) and the rock should consequently be named a porphyritic rhyolite ...
Answer: The application of nomenclature of effusive rocks would bring the equal controversy, as the xenoliths are deep-seated igneous rocks.
L73: Measurement of the full HFSE spectrum is indeed difficult, but is nevertheless currently effected in all good quality laboratories - no need to pretend here to have done a break. Yet, the detail of analytical procedure, as presented here, is indeed useful (but could be transferred to the Supplementary Material).
Answer: The table was transferred into the electronic supplementary file.
Results
Discriminating Nb-bearing phases: possibly better“ Characterization of Nb-bearing phases”
Answer: accepted
L 90: Is there any indication of the relative chronology between these phases and/or with the major rock-forming minerals? From figure 2, it may be deduced that at least some of the accessory phases directly grew in the melt at an early stage of the liquid line of descent, and this should be emphasized. However, your statement in L154-sq suggests that things are more complicated, and a clarification would be appreciated. This is important if a comparison with other rare metal granites must be done, as, likely, many interested readers would intend to do.
Answer: The chronology is unclear in most cases, as the minerals form isolated clusters. Whenever possible, the chronology was emphasized, e.g. in Figure 5 and elsewhere. In addition, we disscuss the temporal relationships in one more paragraph and add succession scheme (Figure 4) for better explanation.
Figure 2: Please hold to the IMA abbreviations for minerals. "Fer" is for feruvite and cannot be used for fergusonite. The correct abbreviation for magnetite is "Mgt", and for zircon, “Zrn”. Qtz, Kfs and Ilm are OK. When no abbreviation has been coined by IMA, you may feel free to coin yours, and thus "Col" and “Pyr” are acceptable. Note, however, that IMA favours the use of consonantic abbreviations: I personally use "Clb" for columbite, and I would suggest "Frg" for fergusonite.
Answer: Suggested abbreviations accepted. We could not find a citable paper summarizing official IMA abbreviations for other than rock-forming or more abundant minerals.
L100: In comment b. “fergusonite is adhered” is uncorrect; you may write “fergusonite sticks to”. It is a pity not to have imaged in detail the zircon with uranothorite inclusions.
Answer: the zircon image is provided, suggestion above accepted.
L 102: In commentc., you note that zoning is caused by U, Ti and Nb, whereas in the text the latter is not mentioned?
Answer: clarified in the text
L 104: You speak of “myrmekitic rutile”, but in the figure, you write “glass + rutile”: myrmekite is therefore not the right term, and in any case the figured assemblage (although too much reduced) is more akin a symplectic association. In the text, however, it is no longer question of “glass + rutile”, and there is consequently a problem of internal consistency! (see comments to L 182).
Answer: The term myrmekitic was meant as morphologic rather than genetic term. We accept using the term symplectite, although it refers to integrown minerals. In our case, glass was also present.
L 109: What are the elements involved in the zoning here?
L112: Here you need to make clear your procedure, i.e., you must explain that you make use of a discriminant diagram, and give a clue to read Figure 3 (what are the coordinates, what is the reference- not sufficient to give it in Fig. 3 caption, etc.).
L 129: More than two decimal digits in apfu is unwarranted, given the analytical uncertainties.
Answer: not exactly right. This may hold true for elements with low-to-medium atomic mass.
L 158: You should give the full analytical data (obtained by EPMA? What conditions? How did you manage the loss in alkalis under the beam?) and compare them with the bulk analysis of the whole granite.
Answer: new table provided containing the required comparison
L 162: Strictly speaking, Y is not a REE!
Answer: no reason to treat Y separately from HREE
L163-sq.: Please give more details about columbite! You should give the variation interval of the Nb to Nb+Ta (Nb#) and Mn to Fe+Mn (Mn#) ratios (saving the reader to go to supplementary data), and the classical Nb# vs. Mn# diagram should be provided, allowing an interpretation of the oscillatory zoning characteristics.
Answer: data added, but the diagram not as it did not show any fractionation trend
L 178: You refer here to an “intimate association” of columbite with Ca-niobate: yet, in Fig. 2 you only present a pyrochlore/Ca-niobate assemblage (not cited in the text). An image of this “intimate association” is needed, allowing the reader to evaluate the significance of the assemblage.
Answer: the pyrochlore-Ca-niobate image was provided in the original version (now Figure 3f). The columbite-Ca-niobate image is of lower quality – not publication-worthy.
L 182-sq.: Despite the “low abundance” (but how much “low”?) of rutile, this phase would deserve a more expanded treatment. Indeed, with such high Nb contents, we may be close to wodginite composition: did you check this possibility?
Answer: wodginite does not contain Ti and is rich in Sn. Hence, this mineral has not been considered.
More important, you should give more details on the rutile-ilmenite relationships. The symplectic association between rutile and ilmenite? (text) or glass? (Fig. 2) is worth a more detailed presentation, and would demand a figure at higher magnification than in Fig. 3, allowing a discussion of the significance of the texture. You assure that the symplectite replaces ilmenite: but, what is the evidence? In the same way, “rutile exsolution” is an interpretation: what is the evidence?
Answer: the figure with higher magnification was included in the triangular plot together with other BSE images showing the ilmenite-rutile relationships. It is impossible to provide contrasting BSE image with rutile and ilmenite, showing also the zoning in columbite. For this reason, the color-coding was selected to make the rutile inclusions at least discernible. The evidence for replacing ilmenite by rutile is also provided. The interpretation of Rt1 as ilmenite-hosted exsolution is supported by Nb-rich rims and equal orientation within the ilmenite host. Furthermore, more resistant remnants of Rt1 protrude into the symplectite, thus providing the unequivocal evidence for the temporal relationships of Rt1-3.
Substitution of Nb and REE: better “Crystal chemistry of Nb-bearing phases”!
Answer: accepted
Figure 4 alludes to “reaction rims” between ilmenite and columbite; yet, they have not been described in the preceding section! You also mention “compact rutile replacing early ilmenite”, which is not described altogether. You must evidently fill these gaps!
Answer: relevant rutile-ilmenite types are now documented directly in the corresponding Figure.
Discussion
L 256: What are the age and geodynamic context of these “other A-type Carpathian granites”?
Answer: clarified, age added
L 229: It would be better to keep in line with the introduction, where you classify the granite as “A1 type”.
Answer: corrected
L 263: As mentioned above, you should have emphasized this primary magmatic characteristic in the Results section.
Answer: text improved
L 270: The presence of fractionation process in the xenolith suite should have been presented in a preceding section (see my comments to the Materials and Methods section).
L 294: Until this point, your demonstration seems to hold. But here, you avoid a series of serious difficulties. First, as seen in Fig. 1, the granite enclaves are true fragmented xenoliths, not the co-magmatic enclaves implied by your simple model directly linking the basalts to the granites. Second, as far as may be understood from your text, you are dealing with a differentiated suite of granite xenoliths, suggesting that the basalts dismembered an evolved granite body. Third, the xenoliths are in reality frozen granitic mushes, which is apparently contradictory with the second point. This third difficulty may be easily resolved if you can demonstrate that the granite enclaves were partially remelted in the basaltic melt – a process which would have been greatly favoured if the dismembered granite body was still at high subsolidus temperature in line with the Zrn age you provide: but what if these Zrn crystallized from a newly formed melt resulting from remelting ? Evidently, your discussion must be refined here.
Answer: This is not exactly the direction we wanted to follow, because it invokes speculations and requires replication of already published data. Our interpretation suggests the formation of granites by fractionation in crustal magmatic reservoirs, and scavenging the fully or partially solidified crystal-liquid mush within later portions of alkali basalt. This model does not contradict the fragmental shape of xenoliths, if crystal dominated and the residual granitic melt is volatile-free. Partial melting within the basalt is also possible and documented by the increased alkalinity of the interstitial melt. However, it can be ruled out that the zircon with typical mantle signature represents the newly-formed phase concidental with the introduction of a hypothethical peralkalic melt component: 1. zircon is wrapped within quartz and K-spars, 2. there is only one zircon generation and U-Pb zircon ages cluster along single isochron without any difference between Th-rich core and inclusion-free rim despite entirely different U, Pb, and Th concentrations, 3. undoubtely primary magmatic fergusonite and pyrochlore are present as inclusions within the clear zircon rims, or fill cracks intersecting the whole zircon, 4. The magmatic growth zoning and euhedral, bipolar zircon shapes cannot originate during the rapid ascent and decompression en route to surface limited to a maximum of several hours, which triggers the skeletal growth. The skeletal growth is indeed characteristic for interstitial ilmenite and orthopyroxene, 5. introduction of the hypothetical peralkaline, HFSE-bearing component during uplift is ruled out by the essentially Nb-free interstitial rutile. The origin of skeletally grown interstitial opx and ilmenite in our xenoliths can be safely linked with the free water release from the interstitial melt during decompression. This process has been addressed and modelled in one of previously published papers (Huraiova et al. 2017). In principle, we do not exclude the partial melting during scavenging within the host basalt, but this process is restricted to feldspars, and apart of ilmenite, orthopyroxene, and possibly also sanidine, no other new minerals have formed. It cannot be ruled out that symplectites as well as ilmenite resorption is linked with the decompression.
L 295-sq.: There is evidently another delicate point here. In fact, your granite xenoliths display contradictory features, mixing characteristics of crustal-derived melts and of peralkaline, mafic-derived, melts. An easy explanation would be to envisage a mixing process between an evolved A-type melt and a peralkaline melt, as found for instance in the Nb-rich Huangshan granite in SE China (Zhu et al., 2018). The mantle-derived characteristics of the Zrn are not contradictory with this model, if the Zrn is considered to be issued from the peralkaline component (a likely possibility, given the uranothorite inclusions you mention in the text). In fact, you could try to inventory the accessory minerals relationships in order to find evidence of a dual origin. From this point of view, the rutile relationships could appear decisive.
Answer: The peralkaline component was added as possible contaminant of the source. It remains, however, rather fictitious as it has been detected neither in silicate melt inclusions nor indirectly by a special assemblage of silicate minerals with high Zr, Nb contents typical of peralkaline granites. The alkaline character of interstitial melt within calcic and calc-alkalic granites xenoliths reselt merely from the partial melting of feldspars. The Huangshan granite is not the right example, as it does not contain Nb-oxides, is alkalic-to-alkali-calcic, meta-to-peraluminous.
Conclusions
L 313: Here again, you could refer to the Huangshan granite (Zhu et al. 2018).
Answer: accepted
Reviewer 2 Report
Dear Authors!
Manuscript ID: minerals-562040
I have read the manuscript with interest; however, there are some issues that need to be clarified and fixed before the manuscript can be considered for publication. My comments are attached to the manuscript and in general comments file.
Have a good luck
Comments for author File: 
Comments.pdf
Author Response
Reviewer 2.
First of all, I think the title should be slightly modified as follows: Niobium Mineralogy of
Pliocene–Pleistocene A1‑Type Granite of Carpathian Back-Arc Basin, Central Europe.
Answer: accepted
The article missing a brief review of Geological background and geological map of the
studied area, would also need to show the selected samples locations.
Answer: Geological background, geological map and sample location added
The authors used the term Pannonian back-arc basin (in conclusion) as the studied area,
while they used Carpathian back-arc basin in the text (as in abstract). The Authors should
differentiated between Pannonian back-arc basin and Carpathian back-arc basin.
Answer: terminology unified
Geological Setting of the studied area
Table 1 needs rearrangement and more explanations as follows: (1) Elements and lines
could be expressed as AlLα , BaLα, CaLα , CeLα ………etc. (2) Symbols P/BCT, LOD and
SD should be explained at the bottom of the Table 1
Answer: Table provided as electronic supplemental data. Symbols have been explained in the original and the revised version
Conclusions: Make sure any suggestions in the conclusion are clearly linked to earlier
material in the manuscript main text: it’s better to see more expansions and rewording the
conclusion of the manuscript.
The manuscript needs to be edited for its English. There are numerous instances such as this
throughout the manuscript.
Which ones ?
Round 2
Reviewer 1 Report
I have read the revised manuscript. A great deal of work has been done to greatly improve it and I now agree with its publication in Minerals.
Reviewer 2 Report
Dear Editor and authors!
Revised Version Manuscript ID: minerals-562040
I went through the revised paper and referees report along with authors reply very carefully.
The authors have shown a lot of efforts to improve the manuscript and this should be well appreciated.
I found the authors have addressed all my comments carefully and in detail by adding more materials in the text and more figures; such as Geological setting and Carpathian back-arc basin geological map. As a result, I now recommend the current form can be accepted for publication without further modification.
Best wishes
Reviewer 2
