Petrogenesis of the early Cretaceous Valle Chico igneous complex (SE Uruguay): Relationships with Paraná–Etendeka magmatism

Abstract

The early Cretaceous (∼130 Ma) igneous complex of Valle Chico (SE Uruguay) is made up of felsic plutonic and subordinate volcanic rocks and dykes cropping out over an area of about 250 km 2. This complex is strictly linked with the formation of the Paraná–Etendeka Igneous Province and the first stages of the South Atlantic Ocean rifting. The plutonic rocks range from quartz-monzonite to syenite, quartz-syenite and granite. The volcanic rocks and the dykes range from quartz-latite to trachyte and rhyolite; no substantial differences in term of chemical composition have been found between plutonic and volcanic rocks. Only a sample of basaltic composition (with tholeiitic affinity) has been sampled associated with the felsic rocks. The Agpaitic Index of the Valle Chico felsic rocks range from 0.72 to 1.34, with the peralkaline terms confined in the most evolved samples (SiO 2>65 wt.%). Initial 87Sr/ 86Sr (130) of the felsic rocks range from 0.7046 to 0.7201, but the range of 87Sr/ 86Sr of low-Rb/Sr samples cluster at 0.7083; 143Nd/ 144Nd (130) ratios range from 0.5121 (syenite) to 0.5117 (granite). The tholeiitic basalt show more depleted isotopic compositions ( 87Sr/ 86Sr (130)=0.7061; 143Nd/ 144Nd (130)=0.5122), and plots in the field of other early Cretaceous low-Ti basaltic rocks of SE Uruguay. The radiogenic Sr and unradiogenic Nd of the Valle Chico felsic rocks require involvement of lower crustal material in their genesis either as melt contaminant or as protolith (crustal anatexis). In particular, most of the Valle Chico (VC) felsic rocks define a near-vertical array in Sr–Nd isotopic spaces, pointing toward classical EMI-type composition; this feature is considered to reflect a lower crust involvement as observed for other mafic and felsic rocks of the Paraná–Etendeka Igneous Province. Decompression melting of the lower crust related to Gondwana continental rifting before the opening of the South Atlantic Ocean or the presence of thermal anomalies related to the Tristan plume may have induced the lower crust to partially melt. Alternative hypothesis considers contamination of upper mantle by a mafic/ultramafic keel composed of lower crust and uppermost mantle after delamination and detachment processes. This interaction may have occurred after the continent–continent collision during the last stages of the Panafrican Orogeny. This “lower crust” model does not exclude active involvement of upper crust as contaminant, necessary to explain the strongly radiogenic 87Sr/ 86Sr (130) isotopic composition of some VC SiO 2-rich rocks. Mineralogical (sporadic presence of pigeonite, Ca–Na and Na clinopyroxene, calcic- and calco-sodic amphibole) and geochemical evidences (major and trace element as well as Sr–Nd isotopic similarities with the felsic early Cretaceous volcanic rocks of the Arequita Formation in SE Uruguay) allow us to propose for the VC rocks a transitional rock series (the most abundant rock types are of syenitic/trachytic composition) preferentially evolving towards SiO 2-oversaturated compositions (granite/rhyolite) also with a strong upper crustal contribution as melt contaminant. This conclusion is in contrast with previous studies according which the VC complex had clear alkaline affinity. Many similarities between VC and the coeval Paresis granitoids (Etendeka, Namibia) are evidenced in this paper. The genetic similarities between VC and the rhyolites (s.l.) of SE Uruguay may find counterparts with the genetic link existing between the early Cretaceous tholeiitic-alkaline Messum complex and the quartz latites (s.l.) of the Awahab Formation (Etendeka region, Namibia).