Abstract:
Andesitic magmas are the product of a complex interplay of processes including fractional
crystallisation, crystal accumulation, magma mixing and crustal assimilation. Recent studies have
suggested that andesitic rocks are in many cases a complex mixture of a crystal cargo and melts
with more silicic compositions than andesite. In situ glass- and mineral-specific geochemical
techniques are therefore key to unravelling the processes and timescales over which andesitic
magmas are produced, assembled and transported to the surface. To this end, this thesis presents a
detailed in situ glass- and mineral-specific study of six Holocene eruptions (Kaupokonui,
Maketawa, Inglewood a and b, and Korito) at Mt Taranaki to investigate the petrogenetic processes
responsible for producing these sub-plinian eruptions at this long-lived (130 000 yr) andesitic
volcano. Mt Taranaki is an andesitic stratovolcano located on the west coast of New Zealand’s
North Island and as such it is distinct from the main subduction related volcanism. Crystal-specific
major and trace element data were combined with textural analysis and quantitative modelling of
intensive magmatic parameters and crystal residence times to identify distinct mineral populations
and constrain the magmatic histories of the crystal populations.
Least-squares mixing modelling of glass and phenocryst compositions demonstrates that the
andesitic compositions of bulk rock Mt Taranaki eruptives results from mixing of a daciticrhyolitic
melt and a complex crystal cargo (plagioclase, pyroxene, amphibole) that crystallised
from multiple melts under a wide range of crustal conditions. Magma mixing of compositionally
similar end members that mix efficiently also occurred beneath Mt Taranaki, and as such only
produced prominent disequilibrium textures in a small proportion of the minerals in the crystal
cargo. The chemistry of the earliest crystallising amphibole indicates crystallisation from an
andesitic-dacitic melt at depths of ca. 20-25 km, within the lower crust. Magmas then ascended
through the crust relatively slowly via a complex magmatic plumbing system. However, most of
the crystal cargo formed by decompression-driven crystallisation at depth so 6-10 km, as is
indicated by the dominance of oscillatory zoning and the equilibrium obtained between mineral
rims and the host glasses.
Taranaki magmas recharge on timescales of 1000-2000 yrs. The eruptions investigated here
provide a snapshot of the end of one cycle and the beginning of another. The younger Kaupokonui
and Maketawa eruptions (ca. 2890 - <1950 yr BP) are the least evolved magmas, record a stronger
mixing signal in the crystal cargo, and are volumetrically smaller than the earlier Inglewood a and
b and Korito eruptions (ca. 4150-3580 yr BP). The Kaupokonui and Maketawa eruptions may
reflect arrival of a new pulse of magma from the lower crust, or that these are early eruptions
within a recharge sequence, which have not had as much time to further differentiate and evolve as
the earlier Inglewood a and b and Korito eruptions that represent the end of a magma recharge
cycle. One of the six investigated eruptions was identified to come from Fantham’s Peak on the
basis of its distinctive glass and mineral chemistry and petrology. Glass trace element data indicate
that this eurption’s magmatic system was distinct from that of the other main vent Holocene
eruptions investigated in this study.
Crystal residence times were investigated using Fe-Mg interdiffusion in clinopyroxene and indicate
that magma bodies stall in upper crustal storage chambers for timescales of a few months to years.
The younger eruptions of the least evolved magmas with the strongest mixing signal return the
shortest residence times, which may indicate that magma mixing events occurring a few months
before eruption may have been the trigger for these eruptions at Mt Taranaki. Amphibole
geospeedometry for these eruptives reveal rapid magma transport from depths of 6-10 km to the
surface on timescales of < 1 week.