Discussion of depositional history of the sites drilled during Leg 180 can be conveniently divided into, first, those on the hanging wall and northern margin of the active rift basin (Sites 1108, 1109, 1110-1113, 1115, and 1118) and, second, those on the footwall (Sites 1114 and 1116). The northern margin sites are relatively little deformed, well dated, and can be correlated accurately using seismic reflection data. By contrast, the footwall sites are more highly deformed, less well dated, and difficult to correlate using seismic stratigraphy. In addition, owing to faulting, differential sedimentation, and erosion, the footwall and hanging wall sites cannot be correlated with confidence across the rift basin.

The oldest well-dated sedimentary rocks recovered during Leg 180 were from Site 1115, the most northerly location of the south-north transect (Fig. F9), and are of middle Miocene to late Miocene age. They are interpreted as part of a forearc basin located north of the calc-alkaline Trobriand Arc and south of the Trobriand Trough subduction zone. These sediments are mainly turbidites deposited in upper bathyal water depths (150-500 m) and derived mainly from basic extrusive rocks. Upward in the middle Miocene succession there is an incoming of redeposited shallow-water carbonate and minor pebbly sandstone. Above this, the upper Miocene interval records shallowing up to a major unconformity.

The uppermost Miocene succession above the unconformity at Sites 1115 and 1109 is emergent to lagoonal. At Site 1115, coring and FMS logging data show the presence of ~3 m of conglomerate composed of well-rounded clasts of basalt together with fine-grained sediments, including possible root traces. At Site 1109, drilling terminated in massive dolerite. This could alternatively (1) form part of a regional Paleogene arc/forearc (part of the Papuan ophiolite belt), (2) record part of the Neogene Trobriand forearc, or (3) relate to rifting of the Woodlark Basin. The dolerite is overlain by conglomerate, as confirmed by interpretation of FMS data. The conglomerates at both sites were apparently deposited in a fluvial to swampy setting. At Site 1109 the nonmarine sediments are thicker and include goethite concretions and altered basaltic material. The overlying lagoonal facies is brackish (alternating fresh and hypersaline) at Site 1109 to initially relatively enclosed and then more open marine at Site 1115. A similar facies (Ruaba unit) is developed at the Nubiam-1 well ~100 km to the northwest where the underlying unconformity is dated as younger than 9.63 Ma (Francis et al., 1987). At that time, Site 1118 was subaerial. An early Pliocene or older conglomerate, at least 50 m thick, was deposited at Site 1118 (Fig. F16), with clasts up to 0.5 m thick, again composed of basalt and dolerite. Lateritic paleosol and breccia are present there in the interstices between clasts.

Each of the three Sites 1115, 1109, and 1118 shows evidence of subsidence from paralic to shallow-marine to deeper (bathyal) water conditions. This transition was diachronous, and began in latest Miocene to early Pliocene time (Fig. F10). The transition from inner to outer neritic and then to upper bathyal water depths occurred first at Site 1109 where shallow-marine carbonates of early Pliocene age accumulated on an open-shelf setting, influenced by traction currents. Similar transitions occurred shortly thereafter at the northern site, Site 1115. At Site 1118, situated on a paleotopographic high, a slightly younger transgression was marked by deposition of early Pliocene or older shallow-marine conglomerates and then lagoonal carbonate with abundant calcareous algae. Then the sediments pass disconformably into upper bathyal calcareous sands in the latest early Pliocene.

Site 1115 sediments exhibit deepening and rapid deposition (344 m/m.y.) (Fig. F10) mainly by turbidity currents during early Pliocene time (3.9-3.6 Ma). Calcium carbonate increases generally upward in the Pliocene-Pleistocene interval, which correlates with a general increase in the pelagic carbonate component through time (Fig. F23). At Site 1109, ~250 m of hemipelagic mud and fine-grained turbidites accumulated during middle to late Pliocene time. At Site 1118 sediments above the shallow-water succession (sandstone, siltstone, and volcaniclastic sandstone) accumulated in deeper water (150-500 m) at very high sedimentation rates (435-485 m/m.y.; Fig. F10) during latest early to late Pliocene time.

At each of Sites 1118, 1109, and 1115 the Pliocene-Pleistocene successions are dominated by deep-marine turbiditic and pelagic lithologies, although with subtle differences in the facies and age at each location. The Pliocene-Pleistocene successions are thickest and show the highest sedimentation rates in the south (Site 1118), decreasing northward through Sites 1109 and 1115 (Fig. F10).

At Site 1118 ~500 m of turbidites and hemipelagic sediments accumulated from upper bathyal depths to middle bathyal depths during middle Pliocene to late Pliocene-Pleistocene? time. Ash-rich interbeds are common (Fig. F24). Reddish, finely laminated, less bioturbated sediments are interspersed with darker colored mainly turbiditic sediments. The interval of reddish sediments can be traced northward on a seismic profile until it disappears before reaching Site 1109 (Fig. F25). Geophysical logs suggest that the noncored interval at Site 1118 (above 205 mbsf) records turbiditic and hemipelagic deposition of mainly Pleistocene age.

At Site 1109 very rapid, thick (~250 m) deposition of hemipelagic sediments and turbiditic mud, silt, and minor sand occurred during middle to late Pliocene time at initially upper bathyal then middle to lower bathyal depths. Volcaniclastic sand and silt and mud of turbidite origin became more abundant at middle bathyal depths. Finally, during the Pleistocene, calcareous nannofossil-rich pelagic ooze and calcareous clay were interstratified with volcaniclastic silt, sand turbidites, volcanic ash, and rare calciturbidites.

During early Pliocene time (3.9-3.6 Ma) at Site 1115 there was a transition to a deeper water succession (150-500 m) with rapid deposition (284 m/m.y.) of muds and volcaniclastic sands, mainly by turbidity currents. Deposition continued in an upper bathyal setting at a very rapid depositional rate during middle Pliocene time (3.6-2.5 Ma). Volcanic glass-rich beds and local concentrations of pumice fragments (Fig. F26) are interpreted as ash fallout. Further deepening took place during late Pliocene-Pleistocene time to upper middle bathyal depths, and the overall sedimentation rate decreased to 63 m/m.y. Pleistocene deposition was dominated by nannofossil ooze with volcanic ash (see below), and sedimentation slowed further (34 m/m.y.) after 0.5 Ma.

A counterpart of the Pliocene-Pleistocene successions at Sites 1109, 1115, and 1118 is found at Site 1108, which is very close to the depocenter of the rift basin. The deeper part of the recovered succession is dominantly turbiditic, whereas the upper part is dominated by talus derived from the Moresby Seamount. The recorded sedimentation began in the middle to late Pliocene with rapid deposition (400 m/m.y.) of fine-, medium-, and coarse-grained sandstones, and minor conglomerate, interpreted mainly as deposits from turbidity currents and debris flows. Clasts and mineral grains were derived from basic and acidic volcanic rocks, and to a minor extent from plutonic rocks (including ultrabasic rocks), metamorphic rocks, and neritic carbonates (Figs. F27, F28). During the late Pliocene (1.95-2.58 Ma) there was an interval of fine-grained turbidity current deposition, affected by faulting. The turbiditic sediments were followed by emplacement of talus (as much as ~50 m thick) comprised of angular clasts of mainly metadolerite of possible ophiolitic provenance, for which the obvious source is the Moresby detachment fault system. Finally, during the late Pleistocene deposition was relatively slow (15 m/m.y.) with accumulation of calcareous, nannofossil-rich clay with minor silt and sand, including volcanic ash.

A small amount of Pleistocene sediments was also recovered at Sites 1111, 1112, and 1113. Data on turbidites and other inferred mass-flow deposits from Site 1110 record deep-water Quaternary sedimentation in a setting of extreme sediment instability. Metamorphic clasts are most likely to be talus deposits derived from the nearby Moresby Seamount. At Site 1111 individual clasts of igneous, metamorphic, and rare sedimentary rocks were recovered. At Site 1112 there was minor recovery of talus of mainly metamorphic rocks intercalated with clay and silty clay, silt and sand, and several volcanic ash beds.

Sedimentary evidence from the footwall sites (Sites 1114 and 1116) on the Moresby Seamount was limited by poor recovery, incomplete logging data, and imprecise dating (Fig. F21).

At Site 1114 the recovered section comprises classical turbidites, mud turbidites, and paraconglomerates possibly deposited by high-density turbidity currents, or debris-flow deposits, that accumulated at middle bathyal depths (500-2000 m) during middle to late Pliocene time at an average sedimentation rate of >176 m/m.y. for the entire succession (Fig. F10). The sediment was apparently derived from calc-alkaline, metamorphic, and possibly ophiolitic sources.

At Site 1116 Pleistocene sediments are absent, and the deep-water succession is dominated by coarse-grained turbidites, debris flows, and fine-grained deposits with a minor input of shallow-water carbonate. The whole of the recovered succession accumulated during middle late Pliocene time at middle bathyal depths (500-2000 m) at an average sedimentation rate of >70 m/m.y. (Fig. F10). The provenance of sediments at Site 1116 was again mainly from a calc-alkaline arc terrane, with a minor metamorphic (Fig. F29) and possibly an ophiolitic contribution.

In summary, the overall sedimentary regime at Sites 1114 and 1116 is consistent with a relatively proximal rift setting. By contrast, sediments in the hanging wall sites were more distally derived, with a variable volcanogenic input, as summarized below.

The presence of volcanogenic ash layers in the sediments at Sites 1118, 1109, and 1115 is displayed in terms of thickness and number of volcanogenic ash layers per million years for the Pliocene-Pleistocene time (Fig. F17). Only one volcaniclastic ash layer was recovered in the middle Miocene section and none from the latest middle Miocene to earliest Pliocene time interval. The probable source of the volcanic ash was the Trobriand Arc, including former volcanoes in the Amphlett Islands and Equm Atoll.

The volume of ash fallout layers is volumetrically minor compared to the supply of volcaniclastic turbidites. However, the presence of numerous ash fallout layers indicates that explosive volcanism, possibly related to continental arc rifting, did take place during Pliocene-Pleistocene time. These volcanic ash interbeds are composed of rhyodacitic to rhyolitic bubble-wall shards and highly vesicular and pipe-vesicular shards with minor phenocrysts. By contrast, the volcaniclastic component consists of brown and colorless glass shards, crystals of plagioclase, quartz, biotite, amphibole, pyroxene, and opaque minerals. Phenocrysts may reach up to 50% by volume of the sediment.

The temporal and spatial distribution of the recovered ash is quite variable (Fig. F17). In the sediments from the early to middle Pliocene transition (3.5-3.6 Ma), volcaniclastic ash layers are particularly abundant at the southernmost site, Site 1118, with 100 layers and a thickness of 8.7 m per 0.1 m.y. By comparison, the frequency at Site 1115 peaks at 10 layers per 0.1 m.y. and at Site 1109 at 15 layers per 0.1 m.y. Several layers of middle to late Pliocene age with abundant volcanogenic material (mainly volcaniclastic) were recovered at Site 1118, in contrast to Site 1109 and Site 1115, where only minor input was dominated by air-fall silicic ash. A prominent episode of mainly volcaniclastic sedimentation took place at Site 1109 in early Pleistocene time, but this was not recorded in the sediments at Site 1115 further north. Finally, the Pleistocene section at both Sites 1109 and 1115 was marked by abundant, dominantly silicic, air-fall ash of platy and bubble-wall types indicative of explosive volcanism.

Porosity profiles reflect a combination of loading history, lithologic effects (e.g., differences in grain size, grain shape, mineralogy, strength, etc.), and chemical effects (e.g., degree of pore filling by cementation and permeability differences that affect dewatering rates). Typically, seafloor porosities of marine oozes are high (70%-80%). For homogeneous sediments that are not overpressured, porosity loss with depth follows an exponential law (e.g., Terzaghi, 1925; Athy, 1930).

As shown in Figure F11A, porosity data from Sites 1109, 1111, 1115, and 1118 show a widely scattered, but consistent downhole trend, which starts at ~70%-80% seafloor porosity and then decreases exponentially to values between ~30% and ~55% at ~700 mbsf. Other than ~120 k.y. of missing sediment at the top of Site 1115, no obvious hiatuses have been noted above the regional unconformity for the sedimentary record recovered from these sites. In contrast, porosity data from Sites 1108, 1114, and 1116 differ remarkably from the above trend, ranging from ~15% to ~45%, although the boreholes were rather shallow (~150 to ~485 mbsf). These anomalously low porosities may reflect erosion of overlying material, although other factors such as cementation may also contribute.

The thickness of eroded material was estimated at these sites by using a regression least-squares exponential fit to the data from each site (e.g., Athy, 1930). The estimates regarding the thickness of removed deposits are ~400 m at Site 1108, ~220 to >400 mbsf at Site 1114, and ~960 m at Site 1116 (see Table T2). However, the limited amount of data and poor curve fits reduce the reliability of these estimates.

In a second approach presented here, porosity profiles from Sites 1108, 1114, and 1116 are compared to the range shown by data from the non- (or minor-) erosion sites (Fig. F11B). This approach assumes that the lithologies and compaction histories of the sites allow them to be used as references for Site 1108 in the graben and for Sites 1114 and 1116, now uplifted on the Moresby Seamount. This comparison indicates that porosities at Sites 1108, 1114, and 1116 correlate to much deeper sections of the low-erosion sites (Fig. F11B) and the data were shifted in depth to obtain a reasonable visual fit. Site 1108 porosity data were shifted 500 m above 165 mbsf, and 700 m below 165 mbsf. The additional 200 m of removal is because of faulting. Similarly, 750 m of erosion was applied to the upper part of the sedimentary succession at Site 1114, resulting in a broad agreement of the data with that from the "reference" sites. Finally, porosities from Site 1116 were shifted down 1000 m (Table T2).

In summary, consistent results were obtained for Sites 1108 and 1116 using the two methods, but at Site 1114, the application of an exponential relationship yielded a significantly lower amount of sediment removed relative to that estimated by the comparison of these sites to the "reference" sites. One possible explanation is that the second approach overestimates erosion where fluid-enhanced cementation, as suggested from filled veins at Site 1108 or 1114, contributes to the low porosities observed. Proximity to active fluid flow along fault zones may cause these sites to be preferentially cemented.

Magnetic susceptibility reflects changes in magnetic mineralogy (e.g., lithologic variations) and was obtained routinely as part of the multisensor track (MST) measurement of sediment cores from Sites 1108, 1109, 1114, 1115, 1116, and 1118. The quality of magnetic susceptibility data commonly degrades from APC- to XCB- to RCB-cored sections because of a combination of reduced core diameter and core fracturing. For these reasons Leg 180 magnetic susceptibility data show considerable scatter.

Intersite correlations are based on recognizing characteristic features in the magnetic susceptibility data sets. For Sites 1118, 1109, and 1115, a first-order difference exists between a consistently high-amplitude susceptibility zone within the upper part of the stratigraphic section and a lower zone identified by a dramatic decrease in susceptibility amplitude (Fig. F12). This high-/low-amplitude combination is found within the turbidite units that characterize much of the basin infill. The magnetic susceptibility profiles shown in Figure F12 have been aligned relative to the top of the Mammoth polarity event, a datum recognized at both Sites 1115 and 1118. In general, the high-amplitude susceptibility zone correlates with the presence of high-frequency, clayey silt/silty clay turbidites. In contrast, the low-amplitude magnetic susceptibility zone relates to high-frequency, silty, and, on occasion, sandy turbidites. Lithologically, these differing clay- and silt-dominated turbidite units have been termed distal and proximal, respectively. By using the paleontologic- and paleomagnetic-determined sedimentation rates for each site, it is possible to display the magnetic susceptibility as a function of time (Fig. F13). This format removes the effect of varying sedimentation rate between sites. Within the accuracy of the age determinations (100-200 ka), it would seem that the transition between the first-order low- and high-amplitude magnetic susceptibility variation is coeval between the sites. Second-order trends are represented by the relatively high amplitude magnetic susceptibility variations that occur at the base of each stratigraphic succession.

Explaining the origin of the magnetic susceptibility variations within and between the Leg 180 sites remains enigmatic. Although the magnetic susceptibility generally correlates with the remanent magnetization intensity, there is no simple relationship with either grain size, main lithologic boundaries, and/or gamma-ray count. A general correspondence does exist between magnetic susceptibility and remanent magnetic intensity, indicating a common magnetic mineralogy. However, X-ray diffraction failed to identify this mineralogy although the high remanent intensities suggest that magnetite is the dominant mineral with respect to the remanence. Ferromagnesium-rich clays such as smectite and, less frequently, chlorite, are likely important contributors to the magnetic susceptibility.

The mixing between magnetic and nonmagnetic clays and sands, in addition to the approximately synchronous change in magnetic susceptibility character across the sites, implies sediment input and mixing from multiple provenances.

Apart from clasts and crystals in volcanogenic sandstones and a few clasts of hornblende basalt at Site 1116, dolerites, metadolerites, and quartz gabbros were recovered at Sites 1108, 1109, 1110-1114, 1117, and 1118. Several conclusions have been drawn from preliminary analyses of these rocks:

1. Ubiquitous dolerite and some gabbro (Fig. F18) occur on both the footwall and the hanging wall of the detachment. Dolerite is the major rock type in the upper levels of the basement of Moresby Seamount.
2. Preliminary chemical comparison shows similar major and trace element compositions in the dolerite and gabbro from the seamount sites (Sites 1114 and 1117) and the dolerite from Site 1118. All of these differ from the Site 1109 dolerite, which is generally less altered. The relationship between these units is not yet clear, but further chemical evidence may provide clarification to provenance, as may dating the time(s) of their formation.
3. Close to the major faults at Sites 1114 and 1117 on the footwall of the Moresby detachment, the rocks underwent a similar history of deformation, metamorphism, and alteration. Metamorphism of these rocks has progressed from the formation of foliation and veining. Subsequently, these structures were refolded and subjected to greenschist facies conditions and hydrothermal alteration. A final phase of brittle deformation is shown by fractures filled with quartz and calcite.

Figure F30 shows average values of major and trace elements from Sites 1114, 1117, and 1118 normalized to the average for Site 1109, which, on petrographic evidence, is the freshest of the sites. The patterns are remarkably consistent, showing very little variation except for K₂O, Rb, and Ba. Because these are all well-known mobile elements, it is tempting to ascribe this variation to the effects of alteration, either by weathering or hydrothermal activity. The rocks from Site 1118 are the most affected by weathering, on the basis of the core descriptions that record extensive formation of red iron oxides, but may also have suffered some hydrothermal alteration. As this site diverges least from the values from Site 1109, we conclude that the variations shown are most likely caused by metamorphism and hydrothermal alteration, especially as these variations are most marked at Site 1114. Small variations, especially in Ni, Cr, and V can probably be ascribed to different amounts of igneous fractionation as seen at Site 1118.

Figure F31 shows plots of major and trace elements of variably altered rocks at Sites 1108, 1111, and 1114 normalized to values for Site 1109 (i.e., the least altered samples). The samples plotted here are specifically chosen to reflect varying degrees of alteration. The samples were arranged in order of decreasing loss on ignition (LOI), as a first-order index of alteration. Encouragingly, Site 1109, judged to be the least altered, had also the lowest loss on ignition. This method of plotting the results shows unequivocally that these samples have gained K₂O, Rb, and Ba relative to Site 1109. Also, the order of increase for all the elements is consistent. Thus, the foliated metadolerite from Site 1114 is the most enriched, followed by the foliated epidote-rich schist, then by the nonfoliated dolerite and the nonfoliated metadolerite, and finally by the metadolerite pebble from Site 1108. This is not the order of LOI, but this may not be too significant. Petrographic evidence shows that the metadolerite from Site 1108 has suffered low-temperature alteration, which has altered the feldspars, while other samples show epidote growth indicative of more drastic mineralogical changes. As an increasing degree of enrichment is related to increasing signs of metamorphism on the basis of, among other things, epidote growth, we conclude that the enrichments shown in Figures F30 and F31 are caused by hydrothermal effects associated with greenschist facies metamorphism. On the evidence from Site 1117, we propose that this metamorphism and chemical alteration have taken place adjacent to fault zones in the presence of fluids channeled by these faults, which have facilitated the metamorphism and hydrothermal alteration.

Data shown in Figure F32 suggest that the Leg 180 dolerites and gabbros are remarkably similar to enriched mid-ocean ridge basalt (E-MORB), with the exception of the elements noted as deviating in the two previous figures, and Ba and Ce are rather surprisingly below the values for E-MORB at all localities. The simplest explanation for this is that Ba is intrinsically low in these samples, as is Ce, suggesting that a better comparison may be with normal- (N-) MORB. This is especially true for Site 1109, which is the least altered of the Leg 180 sites. The other sites have higher Ba because of alteration, as noted above, but never quite reach E-MORB levels. Barium is strongly enriched over N-MORB levels in island arc related lavas and is one of the elements that distinguishes the two (e.g., in the Lau Basin; Hergt and Farley, 1994). Because the Leg 180 dolerites and gabbros have been subjected to varying degrees of Ba enrichment, but have still not reached E-MORB levels, the shipboard geochemical data do not show evidence of a subduction-related signature.

The most spectacular tectonic structure encountered during Leg 180 is the Moresby detachment fault, which dips 25º-30º toward the north-northeast (Fig. F4). At Site 1117, where the fault plane crops out on the northern flank of Moresby Seamount (Fig. F6), we drilled an ~100-m-thick succession of deformed rocks above a basement of undeformed gabbro (Fig. F21). From bottom to top, the gabbro ranges from undeformed (Fig. F18) to brecciated to mylonitic with a well-developed foliation with S-C structures (Fig. F20). Epidote and quartz dominate the secondary minerals, indicating syntectonic greenschist facies conditions. Above this, a fault gouge (Fig. F19) several meters thick crops out on the seafloor and represents the most advanced stage of deformation with evidence for fluid-assisted alteration to produce serpentinite, chlorite, talc, calcite, ankerite, and fibrous amphibole.

By contrast, at Site 1114 a south-southwest-facing normal fault offsets the basement by ~2 km near the crest of the seamount. At Site 1114 a 6-m-thick tectonic breccia lies above a basement of metadolerite. We thus infer a much larger displacement and more extreme P-T conditions along the Moresby detachment fault. One other large normal fault was penetrated at 165 mbsf at Site 1108 within the rift basin. This fault, for which the dip direction could not be determined aboard ship, is revealed mainly by well-developed scaly fabric. It has a vertical offset of ~200 m, which is based on offsets in sediment age, porosity, pore-water chemistry, and temperature vs. depth curves. Otherwise, the observed normal faults have minor offsets and correspond to high level brittle faulting.

At all sites, dip-slip normal faults are predominant but usually coexist with both oblique- and strike-slip faults. The proportion of strike-slip faults markedly increases from the northern sites toward the Moresby Seamount (Fig. F33), in agreement with probable oblique motion on west-northwest-trending normal faults that affect the seamount. This oblique motion, inferred to be left lateral, is in agreement with the north-south extension deduced from earthquake fault plane solutions (Abers, 1991; Abers et al., 1997) and GPS measurements (Tregoning et al., 1998).

Estimates of thermal gradient were obtained at five sites during Leg 180 (Fig. F34). The most reliable estimates were from Sites 1109 and 1115 (31º and 28ºC·km^-1, respectively), where in situ data were collected from multiple Adara temperature tool and Davis-Villinger temperature probe (DVTP) deployments. At Site 1111, two in situ DVTP deployments were made, which were supplemented by an open-hole measurement that was used to extrapolate in situ temperature. Results suggested natural or drilling-induced influx of seawater at shallow depths, with a large-scale thermal gradient of 95ºC·km^-1, based on the mudline temperature and a reliable DVTP estimate at 136.5 mbsf.

Thermal estimates from Sites 1108 and 1118 were obtained solely from open-hole temperature surveys in which data were collected at several depth stations. Measurements from Site 1108 were obtained with the Adara tool on the wireline, whereas Site 1118 data were collected with the temperature tool (TLT) on the logging string. Results from Site 1108 suggest a thermal gradient of 100ºC·km^-1, but the profile suggests possible disruption because of recent faulting or fault-related fluid flow. Results from measurements at Site 1118 yield an approximate thermal gradient of 60ºC·km^-1. However, this estimate was based on only the mudline temperature and one extrapolation at 834 mbsf, and evidence suggests influx of warm fluid at ~700 mbsf. Overall, temperature measurements indicate a trend of increasing thermal gradients toward the graben to the north of Moresby Seamount.

The inorganic geochemistry sampling program during Leg 180 focused in large part on the acquisition of high-resolution pore-water profiles at Sites 1109, 1115, and 1118. Data from these three northern sites provide a basis for evaluating sediment diagenesis in the Woodlark Rise on a regional basis.

Although notable differences exist between the three sites, similar trends in various portions of the profiles of interstitial water (IW) constituents at Sites 1109, 1115, and 1118 indicate that many of the same diagenetic processes mediate the pore-water composition throughout sediments of the Woodlark Rise. The IW chemistry is most similar at Sites 1109 and 1115, with these two sites displaying greater differences with respect to the southernmost Site 1118.

Concentrations of IW constituents in the upper portions of all holes reflect the oxidation of organic matter mediated by microbial activity and the concomitant early diagenesis of biogenic carbonates. This is reflected in extensive SO₄^2- depletion and NH₄⁺ production (see below) in the younger sedimentary sequences section of each site. The upper 500 m at Site 1109 is most similar to the upper 700 m at Site 1118 in this regard, whereas the sediments of the thick onlap sequence at Site 1118 represent an expanded version of part of the section cored at Site 1109. Specific reactions thought to occur include dissolution of aragonite and recrystallization into low-magnesian calcite. The dominant lithologies and mineralogies at all three sites support this inference. The alteration of volcanic matter and clay-mineral diagenesis also occurs at all three sites, albeit to a varying extent. The profiles of dissolved Ca²⁺, Mg²⁺, and SiO₂ shown in Figure F15 reflect a combination of carbonate diagenesis and alteration of volcanic matter. Volcanic alteration and clay-mineral authigenesis typically increase in importance deeper in the sedimentary sequences of each hole. At each site, volcanic ash layers as well as volcaniclastic sands disseminated in carbonates are altered, leading to enrichment of SiO₂ and Ca²⁺ in the pore fluids, whereas the precipitation of various authigenic minerals (e.g., chlorite, smectite, and zeolites) leads to a decrease in the concentration of Mg²⁺ and other pore-water constituents such as K⁺ and Na⁺. Conversion of pre-existing detrital clays, such as illite interlayering into smectite, also contributes to depletion of K⁺ downhole.

The differences between the three sites are often attributable to differences in the thickness and/or the presence of different lithologies at one site relative to the other. Paramount among these are the existence of a lagoonal/brackish water to freshwater transitional sediment sequence overlying dolerite at the bottom of Site 1109, a connectivity between the forearc sediment sequence and the synrift sediments at Site 1115, and the absence of this transitional lagoonal/brackish water to freshwater sequence at Site 1118. A marked limestone/coarse sandstone neritic sediment sequence at Site 1118 imposes important constraints on changes in the pore-water chemistry at this site. The elevated dissolved SiO₂ (and also Li⁺ and Sr²⁺) concentrations deep at this site likely reflect the alteration of the volcanic matter under a higher temperature regime than existed at Sites 1109 and 1115.

The chemical composition of the IW in the sediments of the Woodlark Rise is influenced by a series of sedimentary diagenesis reactions. The alteration of volcanic matter whether as ash layers or dispersed throughout the sediments, carbonate recrystallization reactions mediated by the microbially driven oxidation of organic matter, as well as silicification reactions, all contribute to the observed profiles of pore-water constituents.

Bacteria play a dominant role in the degradation of organic matter in sediments and, as a consequence, drive chemical changes and diagenesis. Although the existence of a deep bacterial biosphere in marine sediments has only recently been established (Parkes et al., 1994), the activity of bacteria at depths to 750 mbsf and their direct involvement in geochemical changes have been demonstrated.

Bacteria were present in all samples analyzed at all three of the deep northern sites drilled during Leg 180 (Sites 1109, 1115, and 1118 [Fig. F14]). Near-surface bacterial populations are similar to those at other sites with similar overlying water depths and near-surface organic carbon concentrations. Population numbers decrease rapidly with increasing depth and conform to the general model for bacterial distributions in marine sediments of Parkes et al. (1994), although in the deeper, more indurated sediments from Leg 180 there is an indication that numbers are decreasing more rapidly than the model predicts, resulting in a sigmoidal depth distribution (Fig. F14A).

The activity of deep subsurface microbial populations is evident in geochemical data from these sites (Fig. F14B-D). Pore-water sulfate concentrations are depleted in the uppermost sediments, below which methane concentrations increase rapidly as methanogenic bacteria gain a competitive advantage over sulfate-reducing bacteria for common organic substrates. Biological decomposition of organic matter is also evident from the accumulation of ammonia in pore waters.

At Site 1118, below ~700 mbsf, an increase in pore-water sulfate concentrations, probably associated with lateral fluid flow, is associated with a rapid drop in methane concentrations (two orders of magnitude in one core; Fig. F14B). This constitutes compelling evidence for continuing microbiological activity in the deep subseafloor environment.

In contrast, at Site 1109 sulfate concentrations were depleted to 0 by 107 mbsf, and remained very low throughout the rest of the depth profile (Fig. F14C). The pore-water ammonia profile shows two distinct peaks, the uppermost of which is associated with a peak in bacterial numbers (Fig. F14A) and is clear evidence of bacterial organic matter degradation. The second peak in ammonia was associated with an increase in alkalinity (data not shown) reflecting continued microbiological activity at depth.

At Site 1115, there are distinct minima in both methane and ammonia depth profiles around ~520-580 mbsf (Fig. F14D). These suggest two separate zones of peak bacterial activity that are separated by the regional unconformity at 572 mbsf.

At Sites 1109, 1115, and 1118, C₁/C₂ ratios remain between 10³ and 10⁴ throughout the holes, reflecting the biological origin of methane (Fig. F14E). At Site 1108, methanogenesis was biological in origin in the uppermost sediments, with a dramatic increase in methane concentrations below the depth of pore-water sulfate depletion. However, increasing quantities of ethane with increasing depth cause a decrease in the C₁/C₂ ratio (Fig. F14E). Below ~100 mbsf small quantities of ethane and propane were present in the headspace gas samples; their concentrations increased with depth. Furthermore, below ~400 mbsf traces of branched and straight-chain C₄ and C₅ components were also detected. The accumulation of >C₁ hydrocarbons and the concomitant decrease in C₁/C₂ ratio confirm a thermogenic gas input to the sediments. The presence of thermogenic gas at shallow depths is consistent with the high thermal gradient at Site 1108 (~100ºC·km^-1) and the thick sedimentary section in the rift basin. Pollution and safety concerns, therefore, prevented us from deepening this site.

The continued presence of bacterial populations at more than 800 mbsf is of fundamental significance. The persistence of microbial life into indurated sedimentary rock adds to a steadily growing body of evidence for a more extensive biosphere than previously imagined. Microbes live in high-temperature regions of the lithosphere (e.g., oil reservoirs, aquifers, iron-rich ores, ocean rifts, and hot springs; Brock, 1985) and even into basalt (Furnes et al., 1996; Giovannoni et al., 1996), and granites (Pedersen and Ekendahl, 1990). Bacteria are known to grow at temperatures to ~120ºC (Stetter et al., 1993) with indirect evidence for the existence of cells approaching 400ºC at abyssal depths (Deming and Baross, 1993) and at pressures of over 1000 atmospheres. The bacterial biosphere in deep-marine sediments may be conservatively equivalent to about 10% of the surface biosphere (Parkes et al., 1994). Recent advances have made significant inroads into understanding both energy sources (e.g., Wellsbury et al., 1997) and electron acceptors (Lovley et al., 1996; Raiswell and Canfield, 1996) for life at "extremes" along with significant improvements in understanding the diversity and physiology of microbial life in deep-sea sediments. The discovery of the deep bacterial biosphere has extended our perception of life from merely a surface phenomenon, and has profound implications for the biodiversity of our planet, fossil fuel formation, the origins of life on Earth, and the potential for life on other planets.

The limits of life on Earth remain unknown: ODP Leg 180 has provided samples and data that considerably extend our knowledge of the deep-marine component of our biosphere.

During Leg 180, four holes were logged by the triple-combo geophysical and the FMS-sonic tool strings, and among these, three were logged for more than 750 m (Table T3). This yields ~2.4 km of FMS and classical logs.

Despite rather uniform lithologies for most of the holes, the log response shows significant variability and will therefore prove very useful to correlate the different drilled sequences (Figs. F35, F36, F37, F38). For instance, the 200-250 mbsf interval in Hole 1118A shows low gamma ray (Fig. F35), low porosity (Fig. F36), and high sonic velocities (Fig. F38). The same features are observed in Hole 1109D in a narrower interval around 230 mbsf. In both cases, this interval can be correlated with a prominent seismic reflector. However, the closer convergence of the two porosity estimates in Hole 1109D indicates that the corresponding sedimentary unit contains less clay than in Hole 1118A. The brackish environment encountered above the major regional unconformity at 572 mbsf in Hole 1115C and at 773 mbsf in Hole 1109D also displays a characteristic log response in both holes. The sonic velocity logs will be used to interpolate the regional velocity structure and better migrate existing multichannel seismic data.

The FMS images acquired during Leg 180 provided critical information about the detailed stratigraphy and structure of both the hanging wall and footwall drill sites around Moresby Seamount (Figs. F39, F40). Excellent quality FMS images were acquired at the three sites on the northern margin (Sites 1109, 1115, and 1118), and at one site near the crest of Moresby Seamount, which penetrated an antithetic normal fault (Site 1114). The FMS images from the northern sites reveal conductive clay and resistive sandy or carbonate-rich interbedded units; centimeter-scale depositional features such as parallel laminae, foresets, and bioturbation; and post-depositional faults and fractures. In particular, FMS images were instrumental in determining the precise depths of important stratigraphic boundaries and the structural orientations of beds and fractures, which could not be precisely determined from core samples alone. Analysis of bed and fracture dips in the FMS images from Site 1114 revealed two distinct dip populations that vary vertically with distance from the south-southwest-dipping antithetic normal fault (Fig. F40). Histogram plots of the dip directions and stereonet plots of the strikes were particularly useful in determining the major bed and fracture dip distributions (Fig. F22).

The VSP experiments were carried out at Sites 1109, 1115, and 1118. The depth range of the recording locations were limited at Sites 1109 and 1115 (effectively to "check shots") but covered 565 m at Site 1118. In conjunction with the VSP experiments, we compiled data of velocity variation with depth using laboratory measurements when sonic velocity data from logging were absent, erroneous (because of washouts), or failed to characterize very thin, high-velocity horizons. The resultant velocity-depth functions were used to depth convert the MCS data immediately adjacent to the boreholes. Check shot information derived from the VSP experiment was used to refine these depth conversions to errors within about 10 m. This will be further refined postcruise.

Depth conversion of the seismic data allows direct correlation of the Site 1109, 1115, and 1118 results with the regional stratigraphy. Figure F25 shows the depth conversions at Sites 1109 and 1118 along with the lithostratigraphic columns. The MCS data in time correlated with depth are shown linking the two sites, which are 8.7 km apart. The limestones and calcareous sandstones close to the base of Site 1118 are thin in comparison to those at Site 1109. These two units, both formed in a neritic environment, fall within a continuous seismic horizon, except that Site 1118 contains only the upper part of the sequence. The interval between 430 and 715 mbsf at Site 1118, corresponding to claystones, siltstones, and sandstones, rapidly thins between common midpoints (CMP) 4075 and 4200, maintaining a conformable relationship with the underlying beds. This interval corresponds to 305-390 mbsf at Site 1109. Volcaniclastic layers are found toward the bottom of the interval at both sites. The reddish claystone and siltstone found between 380 and 430 mbsf at Site 1118 was not seen at Site 1109. Following the corresponding reflectors from Site 1118 toward Site 1109 reveals that this unit pinches out northward between CMP 3950 and 4100 and does not reach Site 1109. The overlying reflectors indicate the possibility of local channel fill above this horizon. Above 385 mbsf at Site 1118 and above ~300 mbsf at Site 1109, the lithostratigraphy is dominated by monotonous silty clays and clayey silts with a middle bathyal origin. However, these horizons correlate well between the two sites.