Micro-Analytical Studies of the Petrogenesis of Silicic Arc Magmas in the Taupo Volcanic Zone and Southern Kermadec Arc, New Zealand

Abstract

The petrogenesis of silicic arc magmas is controversial with end-member models of fractional crystallisation and crustal anatexis having been invoked. A prime example of this is the archetypical continental Taupo Volcanic Zone and the adjacent oceanic Kermadec Arc. Insights into the genesis and timescales of magmatic processes of four continental rhyolitic magmas (Whakamaru, Oruanui, Taupo and Rotorua eruptives) and an oceanic (Healy seamount) rhyodacitic magma are documented through micro-analytical chemical studies of melt inclusions and crystal zonation of plagioclase and quartz. Electron probe microanalysis, laser ablation inductively coupled plasma mass spectrometry and Fourier transform infrared spectroscopy have been used to measure major, trace and volatile element concentrations, respectively, of melt inclusions and crystals. Melt inclusions are high silica (e.g. 74 - 79 wt%) irrespective of arc setting and display a wide range of trace element compositions (e.g. Sr = 17 - 180 ppm). Taupo Volcanic Zone melt inclusions exhibit higher K2O and Ce/Yb relative to Healy melt inclusions reflecting the assimilation of continental lithosphere. Quantitative trace element modelling of melt inclusion compositions: (a) demonstrates that magma genesis occurred through 62 - 76% fractional crystallisation at Healy whereas assimilation of continental lithosphere (greywacke) in addition to 60 - 80% fractional crystallisation is required for the Taupo Volcanic Zone magmas; and (b) suggests the presence of crystal mush bodies beneath silicic magma chambers in both continental and oceanic arc environments. Water concentrations of melt inclusions ranged between 1.4 - 5.1 wt% for the Whakamaru, Taupo and Healy samples. However, the inconsistency in the measured molecular water to hydroxyl concentrations of melt inclusions relative to those determined experimentally for groundmass rhyolitic glasses provide evidence for the degassing of inclusions prior to quenching, by diffusion of hydroxyl groups through the crystal host. Thus, partial pressures of water estimated from the inclusions and inferred depths of the crystallising magma bodies are underestimated. Chemical profiles of mineral zonation, however, indicate a more complex origin of silicic melts than simple fractionation and assimilation. For example, trace element modelling of Whakamaru plagioclase suggests that the three distinct textural plagioclase populations present in Whakamaru samples crystallised from four physiochemically discrete silicic melts. This modelling indicates a strong petrogenetic link between andesitic and silicic magmas from the chemical variation of selected Whakamaru plagioclase crystals possessing high anorthite (45-60 mol %) cores and low anorthite (~ 30 mol %) rim compositions and the interaction of greywacke partial melts. Furthermore, Sr diffusion modelling of core-rim interfaces of the same plagioclase crystals indicate the amalgamation of the magma chamber occurred continuously over the 15,000 years preceding the climactic eruption. Conversely, the major element zonation of Taupo plagioclases implies magma genesis occurred solely through assimilation and fractional crystallisation without the incorporation of evolved crystal mush magmas, indicating a spectrum of magmatic processes are occurring beneath the Taupo Volcanic Zone with each eruption providing only a snapshot of the petrogenesis of the Taupo Volcanic Zone.