PRINCIPAL RESULTS

The sedimentary succession (cored from 328.0 to 452.7 meters below seafloor [mbsf] and from 752.0 mbsf to basement at 1130.4 mbsf; Table 1) is dominated by nannofossil foraminifer chalk and limestone. We divided the sequence into three lithologic units and seven subunits ( Fig. 8). Unit I consists of ooze and chalk, with the transition between the two (at 337.6 mbsf) defining the boundary between Subunits IA and IB. However, because we did not core the upper part of the succession, only a single core of ooze (Subunit IA) was recovered. P-wave velocities in the ooze and chalk sections of Unit I increase downward from a mean of 1700 m/s to a mean of 2240 m/s. Oligocene chalk cored from 752.0 to 838.6 mbsf contains abundant volcanic ash layers and is designated Subunit IC.

Unit II is Paleocene to middle Eocene limestone. Conspicuous bands of chert between 838.6 and 958.3 mbsf characterize Subunit IIA, and the presence of zeolite-rich bands (probably altered ash layers) and less pronounced chert bands below 958.3 mbsf define Subunit IIB. The base of Unit II is placed at the lowest significant zeolite-rich horizon (986.6 mbsf), which coincides approximately with the base of the Cenozoic. Bulk density increases sharply in the lowest part of Subunit IIB, and both the water content and porosity decrease. A marked P-wave velocity increase occurs at the boundary between the chert-rich limestone of Subunit IIA (generally <2500 m/s) and the zeolite-rich limestone of Subunit IIB (generally >3000 m/s); the highest velocities of 3500-4400 m/s are at the bottom of the unit. Cretaceous limestone Unit III is subdivided on the basis of color, with the change from a white Subunit IIIA to a mottled gray and pinkish white Subunit IIIB at 1088.8 mbsf, corresponding approximately to a hiatus and condensation of the Cenomanian through upper Coniacian portion of the sequence. Subunit IIIB, directly above basement, contains microfossils (Eprolithis floralis and Leupoldina cabri) restricted to a short interval straddling the boundary between the early and middle Aptian.

The lowest 2 m of Subunit IIIB contain two intervals of vitric tuff, the lowermost of which is separated from the underlying basalts by a 25-cm-thick limestone bed. The main component of the tuff intervals is basaltic ash consisting of partly glassy to tachylitic fragments with abundant plagioclase microlites. Texturally, many of the fragments are similar to aphanitic pillow margins in the underlying basalts. Most fragments are nonvesicular, but some have vesicles or scalloped margins. Altered brown glass shards are also present; most are blocky and nonvesicular, but some are moderately vesicular. The tuff is composed of at least eight normally graded beds, several of which have scoured bases. The uppermost layer grades up through parallel-laminated to cross laminated beds, indicating deposition by turbidity currents or reworking by currents. The minor moderately vesicular basaltic glass shards in these tuff beds indicate formation by relatively shallow submarine eruptions, whereas the partly glassy basaltic ash that constitutes the dominant component could have been derived from shallow-water to subaerial hydroclastic eruptions, or by erosion of a volcanic source somewhere in the summit region of the main Ontong Java Plateau. These beds, and possibly a vitric tuff of similar age just above basement basalt at DSDP Site 289 (Andrews, Packham, et al., 1975) are the only evidence that a portion of the main plateau was relatively shallow and possibly subaerial.

We cored 80.7 m of basaltic basement, from 1130.4 to 1211.1 mbsf, recovering a total of 44.8 m (Table 1) at a low average penetration rate of 1.2 m/hr. We divided the basalt into eight units (Fig. 9), ranging in thickness from 0.36 to 25.70 m, on the basis of the presence of thin interbeds of recrystallized limestone and/or hyaloclastite. Microfossils in the interbeds indicate an age no greater than early Aptian. All eight units contain pillow basalt, defined by quenched glassy rims, grain-size variations from aphanitic near pillow margins to fine grained in the interiors, and vesicle patterns. Some of the glassy rims appear to be unaltered even though they are laced with calcite veins. The glass is preserved best in Units 4-7. Except near pillow margins, where small (1-2 mm), elongate vesicles are present, the basalt is essentially nonvesicular. Most of it is sparsely olivine ± plagioclase phyric, with a quenched to subophitic groundmass consisting of plagioclase, clinopyroxene, titanomagnetite, glassy mesostasis, and a trace of sulfide. The abundance of olivine phenocrysts increases slightly with depth in the succession, reaching a maximum of ~4%.

Shipboard major and trace element analyses show that the basalt flows at Site 1183 are tholeiitic and very similar in composition to those forming the >2.7-km-thick Kwaimbaita Formation on Malaita, nearly 1000 km to the south (Fig. 10, Fig. 11, Fig. 12, Fig. 13). Kwaimbaita-type basalt also has been sampled 533 km to the north of Site 1183 at Site 807 (Units C-G), in the single flow penetrated at Site 289, 183 km to the northeast, and at Sites 1185 and 1186 (see below). The upper group of basalt flows in Malaita, the ~750-m-thick Singgalo Formation, is compositionally distinct from the Kwaimbaita Formation (Fig. 4, Fig. 10, Fig. 11, Fig. 12, Fig. 13); Singgalo-type basalt also is found in Santa Isabel and forms the upper 46 m of flows (Unit A) at Site 807. Its complete absence from Sites 289 and 1183 suggests that basalt of this composition may not be present on the broad crest of the plateau.

Cumulate gabbroic xenoliths and plagioclase megacrysts are present in Units 2-7 at Site 1183 (Fig. 14). They are round to subround and ¾3 cm in diameter. Clinopyroxene in the xenoliths is partially to totally resorbed, whereas the plagioclase shows only minimal signs of reaction with its basaltic host. Interestingly, similar xenoliths and megacrysts have been found in lava flows of the Kwaimbaita Formation on Malaita (Tejada et al., 2000) and in the Units C-G flows at Site 807, as well as at Sites 1185 and 1186 (see below).

Pervasive low-temperature interaction of the basaltic basement with seawater-derived fluids under anoxic to suboxic conditions has resulted in alteration ranging from &It;5% to 20% of the rock. Olivine phenocrysts are completely replaced by smectite (probably saponite with subordinate nontronite and celadonite), Fe oxyhydroxide and, more rarely, calcite. Groundmass glass has been partly to completely replaced by the same secondary minerals, with minor amounts of pyrite. Black halos, ranging from 2 to 50 mm in thickness, are seen in hand specimen along surfaces previously exposed to seawater and, less commonly, along the margins of veins. Such halos are characteristic of an alteration process initiated during cooling of the lava and completed within 1-2 m.y. (e.g., Honnorez, 1981). Pyrite is associated with the black halos and scattered in the groundmass as far as several centimeters beyond the black alteration front. A third stage of alteration, olive halos containing Fe oxyhydroxide and brown smectite, is common in the upper part of the hole and decreases downhole. This stage of alteration corresponds to halmyrolysis or submarine weathering, which takes place at bottom seawater temperature (i.e., around 2°C) in highly oxidizing conditions and with large water:rock ratio.

Veins are relatively abundant (~20 veins/m) in the basaltic basement. Most result from symmetrical infilling of open cracks with minor or no replacement of the wall rock, and the vast majority contain the following succession of secondary minerals from vein wall to center: smectite and/or celadonite, Fe oxyhydroxide or pyrite, and calcite. Rare, small grains of native copper are present in veins in the upper part of the basement. Veins in the lower part of the basement sequence contain chalcedony and quartz as the final phases precipitated.

The natural remanent magnetization (NRM) of the Miocene ooze and chalk is weak, only slightly above the noise level of the pass-through magnetometer. The Oligocene to Aptian chalk and limestone, with ash layers rich in magnetic minerals, are more strongly magnetized. The NRM of the basalt is strong, but much of the material is broken into small pieces, and reliable magnetic directions are difficult to obtain. However, from detailed sampling of the larger intact pieces, we were able to characterize the intensity and direction of stable remanence. For each core from the hole, we defined magnetic polarity intervals from consistent values of magnetization and calculated a mean paleoinclination. The combination of polarity intervals and biostratigraphic data yields a magnetic stratigraphy for much of the cored interval below 770 mbsf, including the Cretaceous/Tertiary boundary. All basalt samples measured have normal polarity, consistent with formation of basement during the Aptian. Conversion of paleoinclinations to paleolatitudes, combined with age information, allowed us to construct a drift path for Site 1183 for times between ~120 Ma and the present (Fig. 15). The oldest sedimentary rocks indicate a paleolatitude of 25°-30°S. These values are slightly higher than those determined for basement lava flows at ODP Site 807 (Mayer and Tarduno, 1993) but slightly lower than obtained by Hammond et al. (1975) for basal sedimentary rocks and basement at DSDP Site 289. Results for Sites 1185, 1186, and 1187 are within error of those for Site 1183 (see below). Values for all of these sites are significantly less than both the predicted ~40°S early Aptian paleolatitude of the central plateau in the plate reconstruction of Neal et al. (1997) and the ~50°S latitude of the Louisville hot spot today.

The major results of drilling at Site 1183 are summarized as follows:

1. Depositional setting on the seafloor was primarily deep, oxygenated (pervasively bioturbated and no organic-carbon preservation), and quiet (no significant currents or redeposition events after the Aptian).

2. Calcareous microfossil assemblages in ~800 m of uppermost lower Aptian through upper middle Miocene sediments and sedimentary rocks at Site 1183 are for the most part poorly preserved.

3. Planktonic foraminifers and calcareous nannofossils can be used to detect 12 unconformities, of which the most significant are in the inter-Albian (spanning ~10 m.y.) and terminal Albian to upper Coniacian (~13 m.y.).

4. The preserved sediment was deposited above the CCD, and generally above the foraminifer lysocline. However, the regional pattern of the terminal Albian to upper Coniacian (99-86 Ma) condensation or hiatus is consistent with a progressive (relative) rise of the CCD through the Aptian and Albian to a level above the summit of the plateau, followed by a rapid descent of the CCD during the Campanian and early Maastrichtian. These trends may represent the posteruption subsidence history of the plateau coupled with oscillations in the Pacific CCD. Input of volcanic ash into the sedimentary succession is concentrated in two main periods: Paleocene to early Eocene (65-50 Ma) and late Eocene to Oligocene (40-20 Ma). The latter episode is probably related to the Melanesian arc.

The middle Eocene (50-40 Ma) chalk is chert rich and corresponds to a lull in the input of volcanic ash.

Emplacement of basaltic lava flows was entirely submarine at this site and ceased no later than the middle Aptian. The nonvesicular nature of the flows and the microfossil evidence suggest minimum paleodepths of several hundred meters. Microfossils at Site 807 indicate that basalt emplacement there ended about the same time or slightly earlier. Sedimentary interbeds in the upper levels of basement at both sites yielded microfossils no older than early Aptian. Two thin intervals of vitric tuff in the Aptian limestone at Site 1183 and a vitric tuff just above basement at DSDP Site 289 provide the only evidence that at least a portion of the high plateau was shallow.

The Site 1183 basalt flows are petrographically and chemically similar to those of the Kwaimbaita Formation, the lower of the two basalt formations defined on Malaita, and to the lower basalt units (Units C-G) at Site 807, despite the considerable distances separating Site 1183 from Malaita (~1000 km) and Site 807 (533 km). Very similar basalt is present at Site 1186 and in the lower basement units at Site 1185 (see below).