Ascent of Ultrahigh-Pressure Rocks in Southeastern Papua New Guinea, as Revealed by Ti-in-Quartz Thermometry and Rb-Sr Dating

Abstract

Debate concerns the timing of ultra-high pressure (UHP) metamorphism and the kinematics of exhumation of the world’s youngest known eclogite-facies rocks (U-Pb ages of 4 – 8 Ma) in the Woodlark Rift of southeastern Papua New Guinea. End-member kinematic models that have been proposed for the crustal exhumation of metamorphic gneiss domes that host these young eclogites include detachment-related (asymmetric) gneiss doming (metamorphic core complexes), and pure shear-dominated (symmetric or diapiric) gneiss doming. The former is predicted to exhume the deepest structural levels of the domes adjacent to a major normal fault. The latter is predicted to yield a concentric pattern of exhumation levels with the deepest rocks located near the center of the domes. As far as can be determined, there are no mappable field gradients with respect to either the high-pressure metamorphism or the later pervasive amphibolite-facies overprint in the lower crust. This apparent uniformity contributes to uncertainty regarding the distribution of vertical strain in the body, and thus to the mode of dome emplacement. To delineate spatial differences in exhumation and cooling rates, I measured Ti content in quartz for 90 samples of quartzofeldspathic gneiss and eclogite distributed across four migmatitic gneiss domes using laser ablation inductively coupled plasma mass spectrometry. I calculated paleo-temperatures from these data using the Ti-in-quartz geothermometer (Thomas et al., 2010; Wark and Watson, 2006). The Ti concentration in quartz in these samples ranges from 10 to 20 ppm, corresponding to an apparent temperature range of 500°C to 600°C (calculated using the Wark and Watson, 2006 calibration). As the apparent temperatures do not have a clear correlation with lithology and the quartz grains exhibit amoeboid grain boundaries, I infer that the Ti content in the quartz grains is capturing information related to quartz recrystallization and grain growth during the final stages of dynamic recrystallization by high-temperature grain-boundary migration (GBM). The presence of pervasive partial melt in the rocks and of chessboard microstructures in the quartz grains implies that the rocks in the D’Entrecasteaux Islands once attained temperatures of >630°C; however the 93% of the apparent temperatures calculated here are <630°C. Based on the pervasive GBM and the Ti-in-quartz apparent temperatures, I infer that the mean Ti content in these quartz grains was chiefly dependent on the relative rates of cooling and recrystallization as the body ascended through the crust. If the recrystallization rate was less spatially variable relative to the cooling rates across the gneiss domes, more quickly exhumed rocks would retain a larger relict fraction of unrecrystallized, hightemperature quartz grains. Ti content in quartz can therefore be used to map spatial changes in mean exhumation rate. As a rule, Ti concentrations in quartz in the D’Entrecasteaux Islands increase concentrically inward from 2.5 – 25 ppm at the dome margins to 20 – >100 ppm in the core of the domes. Based on this apparent increase in Ti content in quartz, I interpret the most rapidly cooled and exhumed rocks to occur today near the center of the gneiss domes. Thus, the Ti content in quartz data presented here indicate that the gneiss domes were emplaced in the crust by predominantly pure-shear symmetric doming. The timing of rock fabric development at the (U)HP conditions and during the main amphibolite-facies retrogression at lower crustal depths in the D’Entrecasteaux Islands has not been dated. To that end, I present nine Rb-Sr internal mineral isochrons for samples of eclogite, quartzofeldspathic gneiss, and granitic rock from four gneiss domes. Rb-Sr internal mineral isochrons for two samples of eclogites from the core zone of the Mailolo gneiss dome (Fergusson Island) date the timing of the (U)HP eclogite-facies metamorphism at mantle depths to 5.7 ± 2.0 Ma and 5.6 ± 1.6 Ma (2σ), respectively. One sample preserves coesite and the other contains radial fractures around quartz inclusions in garnet, implying that coesite was once been present in this rock. From this data, I infer that the eclogites in the Mailolo dome were metamorphosed at UHP depths of >90 km at 5.6 ± 1.2 Ma (approximate 95% confidence interval for the weighted mean of the two samples). The Rb-Sr isochron ages of amphibolite-facies rock fabric development in a granodioritic orthogneiss and quartzofeldspathic gneiss from the Mailolo and Goodenough dome are 2.4 ± 1.4 Ma (2σ) and 2.38 ± 0.3 Ma (2σ) respectively. This similarity in these Rb-Sr ages implies that deformation at amphibolite-facies conditions took place nearly simultaneously in these gneiss domes. A granodiorite >40 km to the south of the Mailolo granodioritic orthogneiss yielded a Rb-Sr age of 3.90 ± 0.44 Ma. This latter age is 1.5 ± 1.1 Ma (approximate 95% confidence interval) older then the Mailolo and Goodenough samples, indicating an apparent westward younging in the age of the amphibolite-facies metamorphism in the northwestern D’Entrecasteaux Islands. Five samples of quartzofeldspathic gneiss and quartzose rock were analyzed for pressure and temperature estimates using the garnet-plagioclase-muscovite-quartz barometer, the garnet- Al2SiO5-plagioclase-quartz barometer, and the garnet-phengite exchange thermometer. These data indicate a temperature range of 640 – 720 °C and a corresponding pressure range of 10 – 17 kbar for the amphibolite-facies metamorphic overprint in the D’Entrecasteaux Islands. This pressure estimate is higher then previous estimates (7 – 11 kbar) by ~3 – 6 kbar, which implies the HP body was thicker then previously inferred when it ponded at the lower crust. Combining these new Rb-Sr ages with U-Pb zircon ages by Gordon et al. (in review), I calculate a time lag of 2.2 ± 1.2 Ma (approximate 95% confidence interval) between eclogite-facies deformation in the mantle and amphibolite-facies foliation development in the lower crust for rocks in the Mailolo dome. This time lag, when combined with the amphibolite-facies pressure estimates presented here and pressure estimates consistent with the preservation of coesite in one sample, implies a minimum unroofing rate of 19 ± 11 mm yr-1 (approximate 95% confidence interval) for the (U)HP body from the mantle to the lower crust. This minimum unroofing rate strongly supports previous inferences that the exhumation from the mantle to the surface of the gneiss domes in the D’Entrecasteaux Islands took place at plate tectonic rates.