Unraveling the petrogenesis of the Miocene La Peña alkaline intrusive complex, Mendoza, Argentina: Insights from the study of the disregarded late dykes

Abstract

The La Peña Complex (LPC) is a silica-undersaturated alkaline potassic intrusive system, with a subduction-related signature, linked to the early Miocene retroarc magmatism of the Southern Central Andes, in the flat slab segment. The LPC is composed of several intrusions, predominantly plutonic (clinopyroxenite, malignite and syenite), cross cut by a voluminous swarm of radial and annular dikes with mostly volcanic-subvolcanic textures and variable compositions (foid-bearing alkali feldspar trachyte, trachyte, benmoreite, ledmorite, syenite, tephrite, tephriphonolite and alkaline lamprophyre). In the TAS classification these rocks plot in the alkaline series covering a wide spectrum of compositions following two different trends: 1) alkaline (potassic) strongly silica-undersaturated series, from tephrite, phonotephrite to tephra-phonolite, and 2) mid-alkaline, less silica-undersaturated series, ranging from basaltic trachyandesite to trachyandesite (benmoreite), and trachyte. Dikes from the alkaline series show higher K2O/Na2O ratios and Sr, La, Ce, contents compared to those from the mid-alkaline series. Rocks of the alkaline series are richer in K-feldspar, sodalite, leucite (pseudoleucite), biotite, potassic-ferro-pargasite and garnet than the less silica-undersaturated (trachytic) rocks, reflecting a stronger alkaline potassic affinity. A review of geochemical, isotopic and mineralogical data, and a new geochemical modeling performed on the LPC dikes, suggests that both trends represent separated magmatic series that evolved from two different parental magmas lodged ∼30 km deep in the crust. Our results suggest that the compositional variations observed in LPC dikes, cannot be explained by a simple magmatic evolution via fractional crystallization from a unique parental magma, and that an assimilation and fractional crystallization (AFC) process is required to explain some compositional differences. Our results suggest an upper crustal contaminant (evolved rocks) with a Grenvillian isotope signature. On the other hand, analyses of feldspar crystals from the tephriphonolitic dikes indicate local mixing effects, between an evolved tephriphonolitic melt and a less evolved and hotter mafic magma. The origin of both parental magmas could be explained by different melting degrees of the same mantle source, a phlogopite-bearing spinel lherzolite metasomatized by subduction derived fluids. We consider as a possible explanation that alkaline and coeval calc-alkaline magmatism in this part of the Andes, is due to local heterogeneities in the mantle source, and different degrees of partial melting Similar isotopic compositions of the LPC dikes, with those from other Miocene magmatic occurrences with arc-signature and similar age (e.g., Paramillos de Uspallata, Las Máquinas basalt, Abanico Fm and Farellones Fm) suggest an analogous mantle source for these rocks, from arc and retroarc in the Pampean flat slab regions. However, our results suggest that the isotopic trend contamination of LCP is different from that of Paramillos de Uspallata and other arc rocks of the Southern Volcanic Zone. The crustal contaminant of LPC possibly has another composition that those of Precordillera and Principal Cordillera Miocene rocks. The age of LPC rocks (∼19 Ma) and their arc-related signature agree with the eastward broadening of the arc magmatism between 17 and 19 Ma in this part of the flat slab. According to our interpretations, the LPC is a singular occurrence of two alkaline magmatic series on destructive plate margins, associated with calc-alkaline magmatism, occurring closely in time and space.