Petrology and Geochemistry of Arc Tholeiites on the Palau-Kyushu Ridge, Site 448, Deep Sea Drilling Project Leg 59

Abstract

This chemical and petrologic study of rocks from Site 448 on the Palau-Kyushu Ridge is designed to answer some fundamental questions concerning the volcanic origin of remnant island arcs. According to the reconstruction of the Western Pacific prior to about 45 m.y. ago (Hilde et al., 1977), the site of the Palau-Kyushu Ridge was a major transform fault. From a synthesis of existing geological and geophysical data (R. Scott et al., this volume), it appears that the ridge originated by subduction of the Pacific plate under the West Philippine Basin. Thus the Palau-Kyushu Ridge should be a prime example of both initial volcanism of an incipient arc formed by interaction of oceanic lithospheric plates and remnant-arc volcanic evolution. The Palau-Kyushu Ridge was an active island arc from about 42 to 30 m.y. ago, after which initiation of back-arc spreading formed the Parece Vela Basin (R. Scott et al., this volume; Karig, 1975a). This spreading left the western portion of the ridge as a remnant arc that separates the West Philippine Basin from the Parece Vela Basin. In spite of numerous oceanographic expeditions to the Philippine Sea, including the two previous DSDP Legs 6 and 31 (Fischer, Heezen et al., 1971; Karig, Ingle et al., 1975), and even though the origins of inter-arc basins have been linked by various hypotheses to that of remnant island arcs (Karig, 1971, 1972, 1975a, and 1975b; Gill, 1976; Uyeda and Ben-Avraham, 1972; Hilde et al., 1977), very little hard data are available on inactive remnant arcs. Continental orogenic belts contain complexes such as those in Newfoundland and the Caucasus Mountains that have been interpreted as structural fragments of marginal basin-magmatic arc complexes. Thus recognition of processes involved in continental-crust generation requires a thorough understanding of the genesis of both remnant arcs and marginal basins. To this end several questions must be addressed: (1) Is the PalauKyushu Ridge actually a remnant arc? (2) If so, does this remnant arc have any petrologic features unique to the related marginal-basin formation? (3) Does the arc represent the initial phases of magmatic-arc evolution; that is, does volcanism on the Palau-Kyushu Ridge have an arc-tholeiitic character followed by progressively more calc-alkalic volcanism on the West Philippine and Mariana Ridges? Or is the initial arc-tholeiite-magmatic phase repeated each time a new arc is formed after sundering of the previous arc? (4) Is it possible to use the combination of petrologic and geomorphologic characteristics of the Palau-Kyushu Ridge to specify how and where the magmatic arc was sundered? (5) What is the temporal relationship between the cessation of arc volcanism and the sundering of the arc (with subsequent back-arc basin formation)? (6) Is there any indication that immature, largely submarine arcs have established widespread hydrothermal activity that may be the precursor of ore-deposition mechanisms characteristic of mature island arcs? The most extensive petrologic studies of PalauKyushu Ridge rocks prior to Leg 59 were reported by the International Working Group on the IGCP Project Ophiolites (1977), which used suites of samples collected during cruise 17 of the Dmitry Mendeleev between June and August of 1976. Rocks were collected from two dredge stations (1397 and 1396) located on the steep western face of the Palau-Kyushu Ridge (see Site 448 report for a location map, this volume.). The rocks collected included subalkalic basalts, basaltic andesites, andesites, two-pyroxene gabbros, and low-grade pumpellyite-prehnite or greenschist-facies hydrothermal metamorphic rocks. These volcanic rocks are reported to be highly vesicular, typical of shallow-marine or even subaerial volcanism. Two analyses of typical basaltic andesites are presented in Table 1 for comparison with our own findings. The International Working Group concluded that the highly vesicular nature and the petrologic trends of these rocks are typical of island-arc magmatic products. The high Fe/Mg ratios and low Ni, Cr, Sr, and Ba contents are suggestive of arc tholeiites. The high K is probably the result of sea-water weathering. The low Ba and Sr and high FeO/FeO + MgO ratio also rule out the possibility that these may be calcalkalic series differentiates even though the Al values are high. Results from these dredged samples also suggest that volcanic activity ceased by the end of the Oligocene.