Since the advent of plate tectonic modelling, the Ramu-Markham
valley has been thought the site of a classical island-arc/conti- nent collision zone. In this interpretation the units south of the
Markham Valley are held to be island are / micrcontinental mate- rial accreted to Australia and those north of the valley are thought
to represent a separate Finisterre island-arc terrane. Recent
remapping by the Geological Survey of Papua New Guinea shows:
(1) there is no terrane boundary along the Ramu-Markham valley
(2) the Finisterre Volcanics are probably part of either a fore-arc sequence or oceanic plateau obducted onto a microcontinent/
volcanic-arc composite terrane, accreted to the continental
Australian plate. A new tectonic model is proposed for northern PNG and involves:
(I) Eocene to Late Oligocene subduction below a microcontinent
remote from Australia to produce the Sepik are and related
intrusives, the Finisterre Volcanics (fore-arc or possible oce- anic plateau), together with the Alife blueschist. (2) obduction of the Sepik arc in Late Oligocene-Early Miocene
times over the advancing Australian plate. This resulted in
crustal melting, some of the products of which are the Mid- dle to Late Miocene Akuna and Bismarck intrusive complexes
of the Maramuni arc. (3) obduction of the Finisterre Volcanics began in Early Pliocene
times and continues today. (4) post-Middle Miocene mineralisation is the result of melting
after obduction of the Sepik microcontinent.