Dear Jovid,

       Interesting question. Can you be more specific on the type of granitic rock? Or, are you inferring mainly to the igneous granites in your question.

                       Regards,

                        M.Jayabalan.

The standard answer is that peraluminous granites are formed by the melting of metasediments (or have suffered crustal contamination); that metaluminous granitic melts are formed by the melting of meta-igneous rocks (or are the fractionation products of mantle-derived magmas); and that peralkaline granitoids are somehow related to rifts or mantle plumes. There is, however, new experimental evidence, from Othmar Müntener and collaborators, that peraluminous felsic magmas can also be formed by extended fractionation of mantle-derived melts.

I would not confine the concepts of metaluminosity or peralkalinity to granitic rocks only. Also "basaltic" rocks (i.e., ultrabasic-basic-intermediate volcanic rocks with variable MgO content) can be classified as one of these categories. For example lamproitic rocks are Al-poor alkali-rich volcanic rocks and, consequentially, plot in the peralkaline field.

Among the others I would avoid to assign a geodynamic significance (i.e., peralkaline rocks = mantle plumes or rift) because there is no strict relation among these terms. For example, in SW Sardinia island (Western Mediterranean) there are peralkaline rocks (the type locality of peralkaline rhyolites = comendites) genetically associated to a subduction setting, strictly connected with calcalkaline andesites and dacites. It is also true that the other type locality of peralkaline rhyolites (pantellerites) is Pantelleria island, in the Sicily Channel, W Mediterranean, in this case clearly in a rift setting on the foreland of the Apennine subduction system.

There are also peralkaline evolved rocks in Ischia Island (Tyrrhenian Sea) characterized by peralkaline character, but genetically classified among the subduction-related composition (of the old Potassic to Ultrapotassic Roman Magmatic Province).

In altered rocks the peraluminosity can artificially increase as consequence of the easy lisciviation of alkalies, with Al remaining fixed in the altered minerals.

 I agree with Lustrino. 

The association between some geochemical features and a particular geodynamic environment or genetic model, is an idea introduced decades ago and should be revisited. Certainly it is true that into specific geodynamic environments there are peculiar associations of igneous rocks but, as pointed out by Prof. Lustrino often they are not unique. The simple determination of a petrogenetic index cannot provide any kind of information if it is not considered together other characteristics, such as texture, mineralogy, mineral chemistry, trace and major elements variation, isotopic composition.

Regards

 Dear all, thank you very much for your answers. I will take into account the suggestions of Prof. Lustrino. In my case I am dealing with granitic rocks  sensu lato, from  quartz diorites to true granites and only for petrogenetic purposes. To reveal the geodynamic settings a comprehensive analysis of  the data of isotope geochemistry, geochemistry and mineral chemistry will be used. 

Plese Jodiv, pay a lot of attention on the numbers coming out from a thermal ionization mass spectrometer or a multiple collector ICPMS. Isotopes are isotopes, not rocks or books or certificates.

Isotopic ratios are geochemical concepts, not necessarily and authomatically associated to geological processes (e.g., continental rifting, subduction setting, mantle plumes and so on). In every case you have to study the local geological evolution.

Remember that the mantle has a very strong memory. Partial melts can reflect ancient modification processes of their sources (i.e., melt extraction or source enrichment), not necessarily associated to the geodynamic forces acting in the precise moment of igneous activity.

Dear Prof. Lustrino, do you mean that special attention should be paid  to the numbers of the isotope ratios such as 143Nd/144Nd or  87Sr/86Sr instead of plotting these two ratios together? 

I mean that isotopes are geochemical concepts that cannot be authomatically transported into geological features. Isotopic ratios used in igneous petrology (above all Sr-Nd-Pb) are very useful toos, but they must be ancillary and cannot be the base over which a tectonic or geodynamic model is built.

Unfortunately many geochemists or igneous petrologists believe much more to isotopic ratios than to the geological evidences.

Well said Prof. Michele Lustrino. It is chemical laws that determine the isotopic and geochemical characteristics of rocks and minerals but not the tectonic or geodynamic environments.

It must be considered with other parameters. For example, it combination with differentiation index could give an idea of the fractional crystallization process during magma evolution.