Neogene Evolution of the Pacific - Australia Plate Boundary Zone in NE Marlborough, South Island, New Zealand

Abstract

Six new palaeomagnetic localities in NE Marlborough, sampled from Late Cretaceous - Early Tertiary Amuri Formation and Middle Miocene Waima Formation, all yield clockwise declination anomalies of 100 - 150º. Similarity in the magnitude of all new declination anomalies and integration of these results with previous data implies that clockwise vertical-axis rotation of this magnitude affected the entire palaeomagnetically sampled part of NE Marlborough (an area of ~700sq. km) after ~18 Ma. Previous palaeomagnetic sampling constrains this rotation to have occurred before ~7 Ma. The regional nature of this rotation implies that crustal-scale vertical-axis rotations were a fundamental process in the Miocene evolution of the Pacific - Australia plate boundary in NE South Island. The Flags Creek Fault System (FCFS) is a fold-and-thrust belt that formed in marine conditions above a subduction complex that developed as the Pacific - Australia plate boundary propagated through Marlborough in the Early Miocene. Thin-skinned fault offset accommodated at least 20 km of horizontal shortening across a leading-edge imbricate fan. Mesoscopic structures in the deformed belt indicate thrust vergence to the southeast. The palaeomagnetically-determined regional clockwise vertical axis rotation of ~100º must be undone in order to evaluate this direction in the contemporary geographic framework of the thrust belt. Therefore the original transport direction of the thrust sheets in the FCFS was to the NE, in accordance with NE-SW plate motion vector between the Pacific and Australian plates during the Early Miocene. The two new palaeomagnetic localities that are within ~3 km of the active dextral strike-slip Kekerengu Fault have the highest clockwise declination anomalies (up to 150º). Detailed structural mapping suggests that the eastern ends of the FCFS are similarly clockwise-rotated, by an extra 45º relative to the regional average, to become south-vergent in proximity to the Kekerengu Fault. This structural evidence implies the presence of a zone of Plio-Pleistocene dextral shear and vertical-axis rotation within 2-3 km of the Kekerengu Fault. Local clockwise vertical-axis rotations of up to 50º are inferred to have accrued in this zone, and to have been superimposed on the older, regional. ~100º Miocene clockwise vertical-axis rotation. The Late Quaternary stratigraphy of fluvial terraces in NE Marlborough has been revised by the measurement of five new optically stimulated luminescence (OSL) dates on loess. This new stratigraphy suggests that the latest aggradation surface in the Awatere Valley (the Starborough-1 terrace) is, at least locally, ~9 ka old, several thousand years younger than the previous 16 ka thermoluminescence age for the same site. This new surface abandonment age implies that terrace-building events in NE Marlborough lasted well after the last glacial maximum (~17 ka). The timing of terrace aggradation in this peri-glacial region is compared with oxygen isotope data. Downstream transport of glacially derived sediment at the time of maximum deglaciation/warming is concluded to be the primary influence on the aggradation of major fill terraces in coastal NE Marlborough. This interpretation is generally applicable to peri-glacial central New Zealand. Patterns of contemporary uplift and directions of landscape tilting have been analysed by assessing the rates of stream incision and by the evolution of drainage networks over a wide tract of NE Marlborough that includes the termination of the dextral strike-slip Clarence Fault. Relative elevations of differentially aged terraces suggests an increase in rates of incision over the last ~10 ka. Uplift is highest in the area immediately surrounding the fault tip and is generally high where Torlesse basement rocks are exposed. Independently derived directions of Late Quaternary tilting of the landscape display a similar pattern of relative uplift in a broad dome to the north and west of the fault tip. This pattern of uplift suggests dissipation of strike-slip motion at the Clarence Fault tip into a dome-shaped fold accommodating: 1) crustal thickening (uplift) and 2) up to 44° of vertical-axis rotation of a ~40 km2 crustal block, relative to more inland domains, into which the fault terminates. The distribution of incision rates is compared with the pattern of crustal thickening predicted by elastic models of strike-slip fault tips. The observed pattern and spatial extent of uplift generally conforms with the distribution of thickening predicted by the models, although the rate of incision/uplift over the last ~120 ka has been variable. These differences may be due to variability in the strike-slip rate of the Clarence Fault, superimposition of the regional uplift rate or to interaction with nearby fault structures not accounted for in the models.