Late Cenozoic Deformation Along the Pacific-Australian Plate Margin, Dannevirke Region, New Zealand

Abstract

The Dannevirke Basin records a history of late Cenozoic deformation in part of the New Zealand transpressive margin. The structure of the basin was investigated by means of field mapping and reinterpretation of a seismic reflection survey. Southeast verging thrust faults had developed by the early Waipipian (3.6Ma, the approximate age of the oldest basin sediments) and deformation was contemporaneous with sedimentation in a marine environment until about early Nukumaruan (approximately 2.3Ma). During this time, marine mudstones and limestones of the Te Aute Group (Mangatoro, Awapapa Limestone, Raukawa Mudstone and Te Onepu Limestone Formations) were deposited. The thrust faults presently have dips of between 30° and 65° and displacements of around 1000m and up to 3000m on basement. Analysis of the major element glass chemistry of rhyolitic tuffs within the terrestrial sediments of the Mangatarata Formation (Castlecliffian) allows the Dannevirke Basin tuffs to be correlated to radiometrically dated tuffs in the Wanganui Basin, leading to provisional identification of the Potaka Pumice (0.64Ma), Rewa Pumice (0.74Ma) and Pakihikura Pumice (1.06Ma) within the Dannevirke Basin. Age limits for the Mangatarata Formation are estimated at (1.1Ma for the base and 0.54Ma for top of the formation. Tuff marker horizons identified by glass chemistry have demonstrated a facies change away from the Ruahine Range in the upper Mangatarata Formation. The near-range, proximal facies consists of very poorly sorted sub-angular greywacke-clast conglomerates which are considered to be alluvial fan deposits, and these give way within about 3km to distal facies deposits consisting predominantly of siltstones which contain occasional thin beds of rounded conglomerate. On the basis of this evidence, older conglomerates in the middle of the basin are inferred to be the distal facies of proximal conglomerates. The proximal facies alluvial fan deposits are considered to indicate uplift of the Ruahine Range. The oldest conglomerates are around 1Ma, and the range is therefore inferred to have uplifted rapidly since then. Thrust faults within the Dannevirke Basin bound small crustal blocks of 2 to 5km width, and trend sub-parallel to the Wellington Fault, a major dextral strike-slip fault. The basin is the most westerly region of thrust faulting, as strike-slip and high-angle reverse and normal faults become the typical style of deformation on the west side of the Wellington Fault. The strain is considered to be partitioned in the Dannevirke region of the plate margin, with the margin parallel component being taken up on the strike-slip faults west of the basin, and the margin normal component being accommodated on the thrusts within and east of the basin. Four horizons were picked on seismic data. Their ages were estimated using biostratigraphy and assumed sedimentation rates and are: Horizon D - 2.3Ma; Horizon C - 2.8Ma; Horizon B - 3.3Ma; Horizon A - 3.6Ma. Amounts of deformation on each horizon along with estimated ages of the horizons have allowed rates of deformation to be estimated. The basin was shortening at an average rate of 13%/Ma over the period 3.6 to 2.3Ma. Fault displacement rates in the same period range from 0.2 to 2.3mm/a, with the majority being between 0.5 and 1mm/a. Rotation rates of the strata in fault block hanging walls range from 3°/Ma to 30°/Ma, with the majority being close to 7°/Ma. From around 2.3Ma to 0.5Ma a period of tectonic quiescence is indicated by zero or very low rates of shortening, displacement and rotation. A second shortening event is constrained in time by field observations which indicate that Castlecliffian strata were not folded until after their deposition. Shortening began at around 0.5Ma, with associated uplift preventing further sediment accumulation in the Dannevirke Basin. The event is inferred to have ended at some time before present because field observations and the low level of crustal seismicity in the area suggest that the faults are presently inactive. The average shortening rate for the period 0.5 to about 0.1Ma was 15%/Ma; fault displacement rates were around 1mm/a, with an inferred maximum of about 5mm/a; and rotation rates were about 25°/Ma. This shortening event reactivated some of the pre-existing basin faults, and caused steep asymmetrical folds to develop above fault tips. The folds trend sub-parallel to the faults, and commonly have sub-vertical eastern anticlinal limbs. Total shortening on basement averaged over the basin is 23%. A Pliocene erosion surface on the top of basement on the west side of the Ruahine Range is tilted at 8° to the west, and the same surface is inferred to underlie the Dannevirke Basin, where it is estimated to have a minimum westward tilt of 5°. Calculations of the expected deflection of the top surface of the Australian Plate assuming it has a free edge at the Wellington Fault and quite a low flexural rigidity (α = 80km) indicate that a dips 5° to 8° would not develop as a result of lithospheric flexure. A model is therefore proposed which involves westward backtilting of large crustal blocks of approximately 30km width which have bounding faults which penetrate the Australian Plate. The Dannevirke Basin is considered to have formed in the low-lying area at the west side of one of these blocks, and the adjacent Ruahine Range on the upthrown east side of another block.