DISCUSSION

Davies' (1980b) model for the tectonic assembly of Papua New Guinea calls for pre-Eocene northeastward subduction of the Australian plate beneath the Pacific plate, creating an island arc. This collision resulted in underthrusting and metamorphism of Australian plate continental rocks along with southward obduction of oceanic crust and mantle from the Pacific plate (Davies, 1980a). Subduction polarity reversed to southwestward subduction following Eocene collision (Cooper and Taylor, 1987), creating the Trobriand Trough and a volcanic arc above an ophiolitic basement, which overlies metamorphosed Paleocene arc and subducted Australian crustal rocks. In eastern Papua New Guinea, Paleogene arc convergence, continent/arc collision, and ophiolite emplacement preceded continental rifting and the transition to oceanic crustal development and growth (Davies and Smith, 1971; Davies 1973, 1980b; Weissel et al., 1982; Baldwin et al., 1993; Benes et al., 1994).

The presence of ubiquitous basalt, diabase, and gabbro with grain size increasing with depth suggests that the Moresby Seamount may be comprised, at least in part, of oceanic crust (Taylor, Huchon, Klaus, et al., 1999). Sedimentary evidence for subaerial deposition of diabase conglomerate places these rocks above sea level prior to regional rift-related subsidence. Thermochronologic results from this study indicate that diabase and gabbro recovered during ODP Leg 180 did not form as a result of Pliocene-Holocene seafloor spreading in the western Woodlark Basin but instead likely represents latest Cretaceous-early Paleocene oceanic crust. Furthermore, the diabase recovered during Leg 180 has not been thermally affected by Miocene-Pliocene rift-related events. We infer that these rocks have remained at shallow and cool levels in the crust (i.e., upper plate) since they were partially reset as a result of middle Oligocene hydrothermal alteration.

It is tempting to correlate the basalt, diabase, and gabbro recovered during Leg 180 with rocks of the Papuan Ultramafic Belt (PUB) of southeast Papua New Guinea. The PUB is believed to represent oceanic crust and upper mantle obducted over the Australian continental lithosphere during arc-continent collision (Davies, 1980b). K-Ar ages on gabbroic and basaltic portions of the PUB are 147-150 and 116 Ma, respectively (Davies and Smith, 1971; Davies 1980b). Although interpretation of this data is problematic, these K-Ar ages are considerably older than the zircon ²³⁸U/²⁰⁶Pb age (66.4 ± 1.5 Ma) interpreted to date crystallization of the diabase recovered during Leg 180. In the Musa-Kumusi divide of Papua New Guinea, 66- to 56-Ma K-Ar and ⁴⁰Ar/³⁹Ar hornblende ages from granulites from the sole of the PUB have been interpreted to date the emplacement of the PUB (Lus et al., 1998). These data suggest the PUB represents Late Jurassic to Early Cretaceous(?) oceanic crust that was emplaced in the Paleocene.

ODP Leg 180 basalt, gabbro, and diabase may be related to tholeiitic lavas and boninite lavas from the Dabi Volcanics on the Papuan Peninsula, which have ⁴⁰Ar/³⁹Ar total fusion and plateau ages of 58.9 ± 1.1 and 58.8 ± 0.8 Ma, respectively (Walker and McDougall, 1982). These data are concordant with the ⁴⁰Ar/³⁹Ar isochron age on plagioclase from the least-altered diabase (from Site 1109). Furthermore, Leg 180 basalt, gabbro, and diabase may be petrogenetically related to undated gabbro and ultramafic rocks exposed in the D'Entrecasteaux Islands (Davies, 1973), where they occur in the upper plate of active metamorphic core complexes (Hill et al., 1992).

Diabase and gabbro recovered during Leg 180 are petrologically and chemically similar between sites with respect to mineralogy, texture, and nonmobile major and trace elements but are heterogeneous with respect to degree of alteration and incorporation of "excess" or inherited ⁴⁰Ar. Hydrothermal alteration has had a drastic effect on the resetting of argon systematics within these rocks. A correlation clearly exists involving increasing hydrothermal alteration within plagioclase grains and decreasing apparent ⁴⁰Ar/³⁹Ar ages (Fig. F11). ⁴⁰Ar/³⁹Ar isochron ages constrain the age for plagioclase cooling at 58.9 ± 5.8 Ma for unaltered Core 180-1109D-45R and provide a maximum age of K alteration of 31.0 ± 0.9 Ma for pervasively altered Core 180-1118A-70R.

Plagioclase step heat experiments for all analyzed samples revealed nonatmospheric (⁴⁰Ar/³⁶Ar)_i ratios indicative of incorporation of excess argon. Although the mechanism for incorporation of excess argon is not completely understood, it is commonly observed in plagioclase, as has been demonstrated in this and many other studies (cf. Harrison and McDougall, 1981; Pringle, 1993; Claesson and Roddick, 1983). Empirical evidence suggests that argon retentivity in plagioclase is somewhat lower than for biotite (McDougall and Harrison, 1999). Thus, the plagioclase isochron age for the least altered sample (Core 180-1109D-45R) provides evidence that the diabase cooled to temperatures below ~200°-250°C by ~59 Ma.

Rhyolite and microgranite clasts analyzed from Sites 1108, 1110, and 1111 were most likely derived from the vicinity of the Moresby Seamount. These clasts record two tectonic events related to prerift volcanism and synrift magmatism. Zircons from rhyolites with a ²³⁸U/²⁰⁶Pb age of 15.7 ± 0.3 Ma provide evidence for mid-Miocene volcanic activity in the region. One zircon had an inherited 96 ± 6 Ma (2 ) core, indicating the presence of Cretaceous crust in the area during the formation of the rhyolites. The rhyolites were subsequently hydrothermally altered at 10-11 Ma, perhaps as the result of subsequent magmatic activity. Ash fall tuff layers from Site 1115 have similar ⁴⁰Ar/³⁹Ar apparent ages of 13.8 and 14.0 Ma (Lackshewitz et al., 2001). The tuffs are present in the Miocene section below an unconformity, which has been interpreted to represent the transition from subduction and forearc development to rift-related regional subsidence. Thus, protoliths for volcanic clasts from Holes 1110B and 1111A most likely formed during a period of Miocene arc magmatism. Additional evidence for magmatism at this time is found in the Milne Basic Complex on the Papuan Peninsula where 12- to 16-Ma felsic rocks intrude the undated Milne Basic Complex (Smith, 1982).

Although no in situ continental basement was recovered during Leg 180, the ~3-cm-diameter microgranite clast from Site 1108 (Core 180-1108B-6R) provides at present the only evidence for a granitic protolith in the vicinity of the Moresby Seamount that can be temporally related to rift-related magmatism in the western Woodlark Basin. Its small grain size and late Pliocene ²³⁸U/²⁰⁶Pb zircon, ⁴⁰Ar/³⁹Ar biotite and feldspar, and apatite fission track ages suggest that the microgranite was emplaced at shallow (i.e., cool) levels of the Earth's crust. The protolith for this clast crystallized at 3.0 Ma, cooled rapidly by ~2.2 Ma, and was subsequently eroded, transported, and deposited as a clast meters above the northern bounding normal fault of the Moresby Seamount. Similar granitic protoliths exist in the D'Entrecasteaux Islands (Baldwin et al., 1993), suggesting that the clast may have been transported longitudinally from the west. Alternatively, it may have been derived from a more proximal, but presently unknown, source in the vicinity of the Moresby Seamount.

Although the Moresby Seamount is a crustal block bounded on its northern flank by a seismically active low-angle normal fault (Abers, 1991) in an area of continental extension just west of the active seafloor spreading tip, it does not possess many of the features characteristic of metamorphic core complexes (cf. Crittenden et al., 1980; Coney and Harms, 1984). Leg 180 did not recover in situ crustal rocks exhumed from great depths as is the case for lower plate rocks of the D'Entrecasteaux Islands to the west (Baldwin et al., 1993). However, the diabase and gabbro recovered during Leg 180 may possibly have correlations with undated ultramafic and mafic rocks found in the upper plate above detachment faults on the D'Entrecasteaux Islands. Active crustal extension in the vicinity of the Moresby Seamount is being accommodated by normal faulting within latest Cretaceous to early Paleocene oceanic crust.