Evaluating the relative roles of crustal growth versus reworking through continental arc magmatism: A case study from the Ross orogen, Antarctica

Abstract

Arc magmatism in subduction zone settings reflects variable contributions of melts sourced from the mantle wedge, downgoing oceanic slab and sediments, and from partial re-melting of the upper plate. Magmas sourced from the mantle wedge that ascend into and differentiate in the upper plate—evolving to more felsic and buoyant compositions—represent new additions of continental crust, while re-melting of the upper plate may be an important process in the “refinement” of pre-existing continental crust. It is clear that island arc magmatism generates new continental crust, but determining the predominance of either crustal growth or reworking in continental arc settings is more challenging. Granites (sensu lato) are a conspicuous and ubiquitous feature in eroded continental arcs. However, there is no consensus on whether they represent predominantly reworked pre-existing crust or new additions of continental crust through the differentiation of mantle-derived (juvenile) magmas. We assess the relative magnitudes of crustal growth and reworking in a 500-km-long segment of the Ross orogen of Antarctica—an archetypal example of a long-lived “Cordilleran-style” continental arc—utilizing an extensive set of zircon Hf (~70 samples) and whole rock Nd and Sr (15 samples) isotopic data for igneous rocks with compositions ranging from gabbro to granite. Initial ɛNd and 87Sr/86Sr values range from +0.1 to −10.3 and ~0.7044 to 0.7137, respectively. Initial ɛHf values (weighted means of individual analyses from each sample) are predominantly negative—ranging from +4.3 to −12.3. The moderately enriched isotope compositions may be qualitatively interpreted as a record of a dominant crustal component in the source of the granitoids. However, inherited zircon domains provide evidence for significantly less-radiogenic ancient crust in the unexposed basement of the orogen. Additionally, the least evolved samples (SiO2<52wt%; Mg#>69) range in initial ɛHf from approximately +0.8 to −4.0, representing juvenile magmatism with enriched isotopic compositions. While a broader range and lower ɛHf values (+4.3 to −12.3) in more evolved samples from large granitic plutons likely reflect variable degrees of crustal assimilation during differentiation, overlap with the isotopic compositions of primitive samples permits differentiation with relatively minor degrees of crustal assimilation. This qualitative interpretation is generally corroborated by binary mixing and assimilation-fractional crystallization (AFC) models, suggesting that the large granitic batholiths of the Ross orogen reflect significant additions of new, evolved continental crust with subordinate crustal reworking. Some ambiguity in the juvenile/reworked ratio in the most evolved samples is introduced when MELTS is used to track the bulk compositional effects of fractionation-assimilation processes. However, a predominant juvenile contribution to evolved arc magmas is supported by thermal and mass balance considerations from other studies. Numerous assumptions intrinsic to fractionation-assimilation models limit precise quantitative estimates of juvenile versus reworked components in arc magmas, although constraints can be made through careful evaluation of radiogenic isotope records. The large isotope dataset also expands our knowledge of the Ross orogen in the greater context of tectono-magmatic processes along the margin of East Gondwana in the late Neoproterozoic through the early Paleozoic.