Hi Pierre-Arthur,

It seems pertinent to use all layers of information, including field observations. For instance, mafic enclaves and uneven distribution of magmatic Fe-Mg silicates would imply some mixing processes.

Under a microscope, the reverse zoning of plagioclases, pyroxenes and other magmatic solid solutions, as well as any signs of chemical disequilibrium (e.g., reaction rims) also indicate the likely interaction between at least two magmas with different temperature and composition. However, multiple sources are not mandatory in this case, because chemically distinct magmas may represent the various stages of evolution of the same parental melt.

By common opinion, the best evidence of the contribution from chemically different sources is the isotopic heterogeneity of magmatic rocks. For altered rocks, the information from Nd isotopes is a bit more trustworthy than that from Sr and Pb because REE's host phases are relatively resistant to low temperature processes.

The trace element contents (especially the ratios of highly incompatible elements) may also be considered as an argument if the points of your samples definitely follow a mixing line on diagrams (again, it would be better to use HFSE and/or REE ratios instead of LILE, if your rocks are altered). But the information from rock chemistry is more ambiguous than that from isotopes because magma composition may be influenced by crystal fractionation.

Regards,

Petr

Hi Petr,

thank you very much for you precise and detailed answer, it does bring me some ways that I could follow in order to solve this issue. I realized reading your answer that I totally forgot to mention that the sequence is polydeformed and metamorphosed to the amphibolite-facies meaning that, unfortunately, primary mineral textures are lost. however, it is still possible to recognize primary rock textures like enclaves, pillows, breccia, hyaloclastites, just to mention a few. Unfortunately outcrops are generally limited and do not show extensive relationships between the difference lithologies (its one of the main problem when you work in an area covered by vegetation and you don't have access to drillcores). However, I do have a couple of Nd isotopes values and am waiting for more. You are right, isotopes might bring me the definitive answer to this question. I tried to work with HFSEs and REEs but also elements such as Cr, V, Ni and so far the icelandites were falling in between the Fe-Ti basalts and the andesite-dacite but without any clear trend that could be interpreted as a mixing line. I will try to find some papers related to this topic and see if I can find some immobile elements ratios that could do the job.

I will keep you in touch if I find a way that works,

Cheers,

Pierre-Arthur

Dear Dr. Groulier.

I read with interest your question and then your additional post about the history of your magmatic rocks. And then, I am honest to say this, I shook my head, as I read your comment:” I totally forgot to mention that the sequence is polydeformed and metamorphosed to the amphibolite-facies meaning that, unfortunately, primary mineral textures are lost…. Even if I am not familiar with your regional geology, this means finding the "fragments of a broken needle in a haystack". I am sure there will be as set of nice data after all the isotope analyses have been finished to accomplish a nice paper in one of the petrological journals but to you believe that this has anything to do with reality and how will you prove this model which we can expect. After more than 40 years working on all six continents from the Precambrian to the Quaternary I am always surprised what a strange pathway geoscience sometimes goes.

Did you meticulously separate the deformational and compositional alteration processes affecting your primary /parent material (I guess it took place during the Precambrian ?)?

Which minerals come closest to what you may call the original material ?

Do what extent is the material dislocated by the deformation into allochthonous nappe stacks etc. which have nothing to do with chemical mixing ?

Sorry for the “vox diaboli” of an old field geologist who wants to see geologists deal with the real rock.

With kind regards H.G.Dill

Dear Dr. Gill,

thank you very much for your interest and your answer. No do not worry it's always better when you have someone that does the "vox diaboli".

I might have been too negative in my previous answer. We have access to very high quality maps with great details and we do have a precise idea of the stratigraphy. I also had the chance to do 3 summers of field mapping on the area. And thanks to the combination of field observations, petrography, lithogeochemistry, U-Pb geochronology and Sm-Nd isotopes we do know now that we are dealing with a ca. 1.5 Ga peri-Laurentia island-arc, rifted arc and back-arc sequence built on older crustal material. This peri-Laurentian arc was then accreted to Laurentia during the Great Proterozoic Accretionary orogen, around 1.4 Ga and was then strongly impacted by the Grenvillian orogeny at about 1 Ga.

1) "Did you meticulously separate the deformational and compositional alteration processes affecting your primary /parent material (I guess it took place during the Precambrian ?)?"

I do agree with you, it is always a challenge to work on hydrothermally altered, deformed and metamorphosed Precambrian sequences, but we can still manage to overpass these challenges. I have done an enormous work using 1) field observations, 2) structural history, 3) petrography, and 4) lithogeochemistry to discriminate between primary igneous processes and hydrothermal alterations. Also, it is in my opinion easier to separate hydrothermally altered rocks from unaltered rocks when they have been metamorphosed to high-grade metamorphism (upper greenschist to amphibolite facies) because they will develop unusual metamorphic mineral assemblages (to name just a few ortho-amphibole, cordierite, garnet, alumino-silicate, etc...), they are therefore easily mappable. Following the previously enumerated steps I did separate least altered from altered samples. I separated the chemical elements based on their mobility, finding the best elements to use in oder to understand the primary igneous signature of the rocks.

2) "Which minerals come closest to what you may call the original material ?"

I would say zircon, and possibly plagioclase in some porphyritic basaltic rocks, the others minerals are clearly metamorphic (for example amphibole-plagioclase with more or less quartz-epidote in basalt; quartz-feldspar-biotite and/or amphibole in andesitic to rhyolitic rocks).

3) "Do what extent is the material dislocated by the deformation into allochthonous nappe stacks etc. which have nothing to do with chemical mixing ?"

We have a clear dome and basin map pattern which indicates that the sequence was at least impacted by two major deformation events. Within the syncline structures, the supracrustal rocks do exhibit foliation(s) and foldings (and they are sometimes sheared) but S0 is usually well-preserved. Like I said, you can still easily recognized primary volcanic and sedimentary structures (pillows, breccia, hyaloclastite, cross-bedded stratifications, lapillis, etc...). I am confident in my sampling, I avoided areas of possible tectonic mixing and heterogeneous samples (for example rocks with enclaves).

I hope I was clear enough in my answers, it is a pleasure to discuss about this topic with you!

Best regards,

Pierre-Arthur

Dear Dr. Groulier,

I am glad to see that the "Geoscientific chain of Command" is respected and followed up: Geology-Mineralogy-Geochemistry-Advanced level techniques (e.g. age dating, isotopes). Too often this trend is reversed among younger colleagues forced by the doctrine "Publish or Parish". Your summary you gave sounds good and I learnt that there is a lot of "classical way of thinking".

Therefore, I can only wish you much success to your project.

Mit herzlichem Glückauf (traditional German miners´ salute)

Harald G. Dill

Hi Pierre-Arthur,

As usual, there is nothing to add after Prof. Dill's comment. Also, I see you work is well organized, and all major sources of information are involved. I would just note that if you may perform in situ isotopic analysis of zircon, you can measure the variations of eHf values and their spatial distribution within a crystal. The mixing of chemically contrasting magmas results in significant eHf zonation, and zircon is conventionally the best shelter for residual magmatic elemental and isotopic ratios under amphibolite-grade conditions. Although, Lu/Hf ratio may be affected by crystal fractionation, especially in silicic magmas.

Regards,

Petr

Thank you Dr. Gill and Petr for your feedbacks. It definitively gave me some options to look at.

Best regards,

Pierre-Arthur

Hi Pierre,

I have found this site very useful for looking at similar problems in West Australian Archean and Proterozoic magmatic terranes:

https://www.scotthalley.com.au/free-papers-and-workshops

The author uses a simple, systematic approach to classifying geochemical populations according to composition and alteration. Specifically, I would refer you to the "Raglan" lecture video and accompanying powerpoint. Here the author discusses how to identify fractionation trends in an Archean magma series. The author has also published this process in a step-by-step guide which you can find here;

https://pubs.geoscienceworld.org/segweb/economicgeology/article-abstract/115/3/489/583381/Mapping-Magmatic-and-Hydrothermal-Processes-from?redirectedFrom=fulltext

At the very least, this process will help eliminate features that aren't due to mixing (i.e. fractionation).

I hope this helps,

Zeb

Hi Zeb,

whoa thank you very much! it is really informative I will have a close look at it!

Thanks a lot!

Cheers,

Pierre-Arthur