EVOLUTION OF HIGH-K ARC MAGMA, EGMONT VOLCANO, TARANAKI, NEW-ZEALAND - EVIDENCE FROM MINERAL CHEMISTRY

Citation
Rb. Stewart et al., EVOLUTION OF HIGH-K ARC MAGMA, EGMONT VOLCANO, TARANAKI, NEW-ZEALAND - EVIDENCE FROM MINERAL CHEMISTRY, Journal of volcanology and geothermal research, 74(3-4), 1996, pp. 275-295
Citations number
57
Categorie Soggetti
Geosciences, Interdisciplinary
ISSN journal
03770273
Volume
74
Issue
3-4
Year of publication
1996
Pages
275 - 295
Database
ISI
SICI code
0377-0273(1996)74:3-4<275:EOHAME>2.0.ZU;2-Q
Abstract
Magmas from Egmont volcano contain xenocrysts and glomerocrysts entrai ned from melt zones at or near the base of the crust. Lava whole- rock geochemistry therefore reflects mixing between these crystals and mel t and no longer represents melt compositions. As an alternative to usi ng whole-rock analyses, mineral chemistry is used to determine process es occurring during early stages of magma evolution. Primitive magmas at Egmont volcano were hydrous high-magnesian basalts with original f( O2) 2.5-3 log units above FMQ. By the time of eruption this had fallen to 0.8-1.0 log units above FMQ. Early fractionation of olivine (Fo(87 )) and chromite (Cr# 0.7-0.8 and Fe3+# 0.24-0.31), and later olivine clinopyroxene + titanomagnetite drove the evolution of the magma to a high-alumina basalt composition. At the base of the crust these evolv ed magmas entered the amphibole stability field and reaction between b oth anhydrous mafic cumulates and wall rocks crystallised amphibole, b uffering the melt composition to basaltic andesite. Tapping of these m elts to higher levels took them out of the amphibole stability field, resulting in decompressive melting of amphibole phenocrysts and incong ruent melting of amphibole in lower to mid-crustal wall rocks in conta ct with the melt. K2O-rich liquids from incongruent melting were a maj or source of potassium in the Egmont high-K andesites. Some plagioclas e fractionation occurred in higher-level magma chambers but melt segre gation was also an important process. H2O-saturated melts fractionated amphibole as the amphibole stability field was again intersected and these melts evolved along a calc-alkaline trend to dacite. In contrast drier melts did not fractionate amphibole and evolved only to andesit es. As most of the K2O present is in the groundmass of the lavas, that is, in the liquid phase, the melts formed bl-the highest degree of me lt extraction are the most potassic and these are the melts which tend to evolve to dacite.