Linking crust and mantle:  the origin of Cenozoic alkali magmatism in the 
southwest Pacific
Finn, C.A., Müller, R.D., Panter, K. and Moresi, L.

Understanding the origin of < 50 Ma alkaline igneous rocks in West 
Antarctica, eastern Australia, parts of New Zealand and the Pacific plate 
extending east of the Australia-Antarctic discordance and west of the 
Antarctic Peninsula requires linking crustal tectonic and magmatic studies 
with mantle dynamic models through a variety of data sets.  The magmatism has 
been conjecturally linked to rifting, mantle plumes, or hundreds of hot 
spots, but all of these associations have flaws.  For example, plate 
reconstructions demonstrate that the last episode of major extension in west 
Antarctica, eastern Australia and New Zealand occurred during the Mesozoic 
break-up of Gondwana, and potential-field and radar data do not image large 
Cenozoic basins suggesting that the widespread magmatism cannot be explained 
by rifting alone.  Estimates of volumes of magmas erupted in west Antarctica 
and Australia, are low compared to areas associated with plumes.  Geochemical 
models and low velocity zones imaged in seismic tomography data from the 
region indicate that the melts originate from a restricted zone between 
depths of 100-200 km, unlike areas underlain by deep-seated plumes.  Uplift 
and doming typically associated with mantle plumes are also largely absent; 
the age of a proposed Cenozoic dome in Marie Byrd Land, Antarctica is poorly 
constrained.  Also, to explain the areal distribution of the volcanism, an 
unusually large plume would have to underlie the entire southwest Pacific, or 
there would have to be hundreds of hot spots.  Therefore, new models that 
incorporate models of crustral structure, mantle temperature and geochemical 
variations, and mantle dynamics are required to find the cause of magmatism.  
We propose an alternative model that links the sudden detachment and sinking 
of several dense, subducted slabs into the lower mantle in the early Eocene, 
with a slight rise of upper mantle material away from subduction zones.  This 
ascent of warm mantle may induce melting of metasomatized mantle.  Consequent 
upon minor crustal extension, small batches of magma leak through faults and 
fractures. Larger volumes of magmas that form large volcanoes erupt in areas 
with moderate extension such as documented near Antarctica (the Adare Trough, 
Terror Rift), and the Balleny Islands regions.  Superimposed on this 
regionally extensive alkaline magmatic province are volcanic chains (in 
eastern Australia and the Tasman Sea and Louisville Ridge) related to well-
documented hot spots, such as the Bass and Louisville hot spots, 
respectively.  Geodynamic models on mantle flow in response to slab 
detachment, improved age control on volcanoes in the region and detailed 
seismic tomographic studies in Antarctica could help resolve appropriate 
models for magmatism.