Sr Isotope Zoning in Plagioclase from Parinacota Volcano (Northern Chile): Quantifying Magma Mixing and Crustal Contamination

JOURNAL OF PETROLOGY VOLUME 55 NUMBER 6 PAGES 1203^1238 2014 doi:10.1093/petrology/egu023 Sr Isotope Zoning in Plagioclase from Parinacota Volcano (Northern Chile): Quantifying Magma Mixing and Crustal Contamination C. GINIBRE1,2* AND J. P DAVIDSON1 1 DEPARTMENT OF EARTH SCIENCES, DURHAM UNIVERSITY, SOUTH ROAD, DURHAM DH1 3LE, UK 2 DEPARTEMENT DE MINERALOGIE, UNIVERSITE DE GENEVE, 13 RUE DES MARAICHERS, 1205 GENEVE, SWITZERLAND RECEIVED JUNE 18, 2008; ACCEPTED APRIL 7, 2014 We present analyses of Sr isotope zoning by microdrilling and thermal ionization mass spectrometry in plagioclase crystals from Parinacota volcano (Central Volcanic Zone, northern Chile), which were analysed for major and minor element zoning in a previous study. Although the isotopic range of the bulk-rock samples is small at this volcano (0·7067^0·7070, except for one flow of mafic andesite at 0·7061), significant variations are seen (0·70649^0·70700) within and between plagioclase crystals. A general negative correlation is observed between Sr isotope composition and Sr concentration in the liquid in equilibrium with each plagioclase zone, as calculated from chemical zoning data and partition coefficients. Additional scatter is superimposed on this general trend, indicating a decoupling between isotopic and chemical variations for Sr. In one dacite sample a detailed isotopic profile shows increasing contamination during crystal growth, except for an abrupt decrease correlated with a dissolution surface and interpreted as a recharge event. We apply energy-constrained recharge, assimilation and fractional crystallization modelling to the melt evolution recorded in the chemical and isotopic zoning in this crystal. Results suggest 20% assimilation of the local wall-rock gneiss, at high initial temperatures. The isotopic data confirm the involvement of two contrasting mafic magmas, which are sampled at flank cinder cone vents. One (Lower Ajata) has a low Sr content with high 87Sr/86Sr, the other (Upper Ajata) has a high Sr content with lower 87Sr/86Sr. In some samples from Parinacota, the isotopic composition of plagioclase crystal rims or groundmass crystals is significantly higher than that of the high 87 Sr/86Sr mafic magma. In others, where chemical zoning profiles suggest that recharge was from the low 87Sr/86Sr magma, the 87 Sr/86Sr of the groundmass and crystal rims is higher than expected. This indicates either additional parent magmas to the two previously identified, or further crustal assimilation, either at lower crustal *Corresponding author. E-mail: depths, before crystallization of plagioclase, or just after the last recharge. Our results illustrate the complexity of magma^crust interaction beneath Parinacota, which is likely to be representative of many other Central Andean volcanoes formed on thick crust. Such complex interactions can be revealed by combined study of chemical and isotopic zoning in plagioclase (in a textural petrographic context), despite a small whole-rock isotopic range.The distinct contamination patterns of various samples suggest an important role for the geometry, location and evolution of the magma plumbing system and, in general, variations of the thermal and compositional structure of the crust underneath the volcano. KEY WORDS: crustal assimilation; magma mixing; isotopic microsampling; plagioclase; zoning patterns; Parinacota I N T RO D U C T I O N The influence of the various components (mantle wedge, lithospheric mantle, slab melt or fluids and continental crust) in the sources of arc magmatism is a subject of considerable debate and clearly varies between arcs. The role of lower continental crust in modifying magma compositions has been shown to be important all along the Andean volcanic arc (e.g. Hildreth & Moorbath, 1988; Wo«rner et al., 1992; Davidson & de Silva, 1995; Garrison et al., 2006). The Andean Central Volcanic Zone (CVZ) represents a segment of the continental arc where the crust is thickest; therefore, the chemical and isotopic characteristics of the magmas are expected to be strongly influenced by the nature of the crust beneath the volcanic ß The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@ oup.com JOURNAL OF PETROLOGY VOLUME 55 front. Several studies document the effect of this thick crust (Harmon et al., 1984; Davidson et al., 1991; Wo«rner et al., 1994) and identify a difference in crustal composition in different domains in the lower crust [e.g. north and south of the Pica Gap, 19^218S (Wo«rner et al., 1994); the Arequipa Domain, to the north, is isotopically distinct from the domains to the south (Mamani et al., 2008)]. McMillan et al. (1993) have also identified a relationship between the modification of crustal signature and the timing of crustal thickening. A lower crust signature is particularly clear in the Nevados de Payachata volcanic region, located in northern Chile (188S), next to the Bolivian border, including the Pleistocene to Holocene twin composite volcanoes Pomerape and Parinacota (Wo«rner et al., 1988; Davidson et al., 1990; McMillan et al., 1993). The eruption products of the Parinacota stratovolcano are mainly andesitic and range from basaltic andesite to rhyodacite; most Parinacota samples have only moderately elevated Sr concentrations (900^1200 ppm) similar to other volcanic centres of the northern CVZ. However, two basaltic andesite units from recent flank eruptions (Ajata Flows) have contrasting concentrations of Sr (up to 1800 ppm vs 800 ppm) and other incompatible trace elements. These probably reflect different amounts of crustal assimilation at different depths, the high incompatible element contents being typical of a lower crustal signature. The study of the relationships between such mafic magmas and the products of the main edifice is therefore a way in which to investigate magma^crust interactions beneath this volcano. The influence of mixing with two mafic end-member magmas during the whole history of Parinacota has been implied by Bourdon et al. (2000) based on U^Th data. Ginibre & Wo«rner (2007) showed, based on trace element zoning in plagioclase, that such mixing occurred with alternating magma recharge events of increasing frequency through time. However, these studies do not allow detailed identification of crustal assimilation. For that purpose, we have examined Sr isotopic zoning in plagioclase at high spatial resolution. The way in which two mafic magmas interact with the crust, and are modified by processes such as fractional crystallization and assimilation, will depend on the details of the plumbing system (magma chamber size, location and interconnectivity) as well as the nature of the crust. This study thus allows us to make further inferences about the magma plumbing system of Parinacota Volcano. Sr isotope microsampling has been used previously to elucidate disequilibrium and mixing (...truncated)