Yes, in subduction setting it is like this if NMORB normalisation

This is a typical arc signature. This is also a MORB characteristic features. Although, one should look for more correlative features to state your evidence.

Yes, it is a typical characteristic of subduction-related magmtism.

Yes, it is typical of oceanic subduction, normalized to NMORB.  A continental setting could be more variable, and sometimes enriched in Zr and Hf.  As a supporting chemical argument, you might also consider the water-soluable LILEs which typically show enrichment from a metasomatized mantle overlying the subducting plate.

Yes and I agree with Victor. Subduction-related magmatic rocks normalized to NMORB and Primitive mantle display negative Nb anomaly.

I believe that this trace-element pattern could also be due to crustal contamination of an asthenospheric melt in a continental within-plate setting? The actual mapped geology of the dykes  and associated rock types could be more diagnostic of the tectonic setting. Are they intrusive into a folded volcanosedimentary package of similar age to them, or are they clearly intruded into cratonic orthogneisses and granitoids?

I believe the combined Nb-Ta and Ti anomaly is a likely chemical signature of subduction-related magmas. However when studying the northern segment of a continental magmatic arc in Brazil, Araçuaí orogen, we found the following: "Although marked by negative Nb–Ta, Sr and Ti anomalies, typically associated with subduction-related magmas, the combined Sr, Nd and Hf isotopic data characterize a crustal signature related to anatexis of metamorphosed igneous and sedimentary rocks, rather than fractional crystallization of mantle-derived magmas." You can get more details at: http://www.sciencedirect.com/science/article/pii/S1342937X15001951. I hope it helps somehow. Good luck ;)

Yes, it is one of the chemical distinguishing features of such type subduction-related magmatic rocks in their MORB-normalized patterns 

I'm interested in what Mr Evans said. Nb-Ta negative anormaly is a  typical signature of continental materials. Almost all of the granitoids in continents have significant Nb-Ta negative anormaly, which may be attributed to a widely accepted fact that the granitoids are formed by remelting of continental materials, mostly metamorphic sedimentary rocks.

For basalts, Nb-Ta negative anormaly has been the most important diagnostic of subduction-related magmtism. However, whether the Nb-Ta negative anormaly can be due to crustal contamination of an asthenosphere- or mantle-derived melt, as Mr Evans said?

Nice comment, Lu Lu.

The anomaly, as you say, is typical of crustal material. Crustal material can be recycled into mantle for long distances, as pointed out by the great Australian petrologist Ringwood long ago.

Cheers

So, a tough question is raised now that how to distinguish the basaltic magmatism in the subduction-related setting from those in the within-plate setting? I agree with what Mr Evans said. Associated rock types maybe more diagnostic of the tectonic setting, given that different tectonic setting would have different igneous assemblage.

What Leonid Shumlyanskyy said is nice and helpful.

Thank you all for your comments and discussions. As I believe, the Nb, Ta, Zr, Hf, all being the HFSE group of elements should behave in a similar manner. Moreover, the negative Nb-Ta anomaly in a rock is generally due to the presence of Ti bearing phases during partial melting as these phases can accommodate Nb, Ta and other HFSEs. So negative Nb-Ta anomaly must be accompanied with negative Zr-Hf signature for a subduction related rock.In our case no Zr-Hf anomaly is observed(???).

Another issue whether subduction or crustal contamination as mentioned by Mr. Lu Lu is very crucial. As mentioned by Leonid Shumlyanskyy, two possibility arises either crustal contamination or presence of the subducted crustal material in the plume source.Other geochemical and published literature supports for the second possibility indicating Archean subduction signature as the studied dykes are of Paleoproterozoic age.

Reply to D.M.Evans: These dykes are intruded into  cratonic orthogneisses and granitoids.

yes, the subduction-immobile character of Nb-Ta makes them depleted in the samples originated in arc-regions.

Dear all,

I am still trying to understand if negative Ta-Nb and Ti anomalies for granitoids just happens in subduction-related environments or could also reflect another tectonic setting. Does anyone have a paper showing a different interpretation rather than a magmatic arc building? Thanks a lot. All the best. ;)

The answer to your question is: not always. Post-collisional or intraplate magmatism sourced from a hybrid mantle (crust+DM) presents geochemical (incompatible elements) and isotopic characteristics of the crustal component. Couzinié et al. (2016) (Earth and Planet. Sci. Lett.; Vol. 546) have demonstrated that a small (~10-25%) crustal component, transported to great depths in subduction zones as sediment or mechanically eroded sialic crust, is able to metasomatize/hybridize depleted mantle so that a rock sourced (e.g. high-K basalts) from this hybrid mantellic domain could provide incompatible element signatures (including Nb-Ta troughs), as well as radiogenic isotopic values (i.e. Hf, Nd), that dominantly reflect the crustal component. Therefore, let's picture a hypothetical scenario. Subduction during the Paleoproterozoic would generate arc-related magmatism and metasomatize/hybridize large swaths of the suprasubduction mantle. Right after continental collision, orogenic extensional collapse would partially melt this hybrid mantle and generate basalts (and fractionates) with a crustal-like geochemical signature, and therefore Nb-Ta anomaly. Extensional processes could (and probably would) continue during much longer periods due to periodic extensional/rifting. Therefore, basalts with Nb-Ta anomalies and other geochemical fingerprints of subduction zones could be generate in post-collisional or intracontinental settings. Nevertheless, they need to be sourced from mantle that was previously enriched and metasomatized by a (minor) crustal component delivered by an active subduction zone. Hope this helps. Cheers.

Dear Rodrigo S. Marimon

Thanks a lot for your answer and for the suggested reference.

I am really interested in such geochemical parameter. I am going to read the paper in details and see if it applies to the biotite hornblente tonalite and granodiorite I have been worked with. Best regards, Léo.

Dear Leonardo Gonçalves

No problem. I forgot to mention, but crustal melts could also generate a Nb-Ta anomaly inherited from the parental rock. This could be the case specially for more evolved rocks, such as tonalites and granodiorites. You might wanna take a look at the papers of Mora et al. (2014) and Rocha et al. (2018), that focus on the Socorro-Guaxupé domain of the Southern Brasília Orogen (SBO). They studied diatexites and provided chemical analyses for the considered rocks. It is interesting to note that leucosomes show Nb-Ta anomalies, even though they're clearly from crustal origin. Some anatectic batholiths in the SBO are thought to be derived from high-grade melting of the middle-lower crust and also show conspicuous Nb-Ta anomalies (Janasi, 1997, 2002; Vinagre et al., 2014; Toledo et al., 2018). Contrary to the rocks you mentioned, these anatetic granitoids are K rich. However, I suspect that this has something to do with the (mostly) mafic parental rocks that are also high in K (see paleosome of Rocha et al., 2018). At the end of the day, Nb-Ta anomalies alone are not a reliable indication of subduction-related magmatism.

Cheers,

Rodrigo

Dear Rodrigo S. Marimon ,

Thanks again for the papers you suggested.

I am going to read them and see if negative Ti anomaly is also a chemical parameter of the studied rocks. The biotite hornblende tonalite and granodiorite I mentioned are metaluminous to slightly peraluminous, so I think they are quite different from those migmatites studied in the SBO. Anyway thanks a lot for the discussion and provided references. Cheers, Léo.

Maybe the samples underwent crustal contamination