SYNTHESIS

Along the way, we recorded a number of operational breakthroughs for ODP. These included the first deployment in hard rock of the ADCB and the first application of a hammer drill in a new strategy for hard-rock reentry as well as an innovative adaptation that allowed us to case an existing hole. We achieved the first free-fall deployment of a standard reentry cone, enabling a deep hole to be commenced exactly on a small target selected under drill stem video observation. Another innovation was the nesting of a FFF above the reentry cone, allowing dual casing strings to be set after the first one failed to seat properly. On the logging side, we achieved the first hard-rock application by ODP of LWD technology. We also performed the first direct comparison in hard rock of geophysical profiles obtained by two different tools—formation images obtained by the RAB tool during drilling, followed by wireline logging and formation images acquired with the FMS. These initiatives were vital to the scientific outcomes of the leg and will have long-term benefits for the program and its successor. Great credit is due to the drilling engineers and rig crew for their initiative and persistence.

Our scientific task was, in essence, to document the third dimension of the hydrothermal system responsible for development of massive sulfide chimneys and mounds at the PACMANUS site on felsic volcanic Pual Ridge. Thereby we would better understand the interplay between fluid pathways and fluid-wallrock interaction that governs the nature and location of mineral deposition within an environment representing a modern, actively forming analog of the settings of many orebodies in ancient geological sequences. Subsequent deformation and metamorphism commonly obscure genetic evidence pertaining to the latter. In addition, we would delineate the extent to which microbial life flourishes in the subsurface of such a hydrothermal system, contributing to ODP's "Deep Biosphere" initiative. The choice of location was governed by prior understanding of the tectonic and volcanological setting of most ancient massive sulfide ore deposits and by the detail with which the PACMANUS site had been surveyed by surface vessels and manned submersibles since its 1991 discovery.

Prior isotopic and geochemical research on the PACMANUS chimneys had indicated that magmatic fluids and metal sources represented a significant component of the system. Unraveling the relative importance of these and of seawater-dominated leaching of metals from country rocks was to be a main aim of drilling and the postcruise research to follow.

Our strategy was to drill as deeply as possible at four sites, including two along the crest of Pual Ridge representing hydrothermal outflow zones characterized by low-temperature diffuse venting (Site 1188) and high-temperature focused venting, respectively (Site 1189, augmented by Site 1191), plus a nearby background or reference site where no past or present hydrothermal activity was known (Site 1190). The fourth site was a presumed fault at the base of Pual Ridge, where major ingress of seawater into the lower parts of the hydrothermal system seemed likely. The very first cores returned showed that seawater influences were major even at the upper levels of the system, so plans to drill this fourth site were abandoned very early during the leg.

Several unsuccessful attempts were made to drill the reference site, which achieved a new relevance as a consequence of early drilling, namely assessing just how far the unexpected intense alteration extended under the newly revealed cap of fresh dacite lavas. Premature termination of the leg after an emergency port call to Rabaul prevented a final attempt to penetrate this difficult site. Fortunately its initial purpose, to assess the volcanic architecture of Pual Ridge and to provide samples for comparison with products of hydrothermal alteration, was nonetheless achieved because igneous fabrics were well preserved at the two main sites, especially within an intersection of relatively less altered dacitic lavas and clastic rocks in the lower part of Hole 1189B.

The proponents of Leg 193 expected that, under both outflow sites and especially at the one with higher temperature focused venting, there would be interleaving between altered and unaltered rocks, the former representing preferred fluid pathways and possibly being associated with mineralized layers and veins affiliated with the seafloor deposits. Marked changes were anticipated in the alteration patterns, both vertically and laterally, reflecting different temperature gradients and fluid chemistries. Several major outcomes of the drilling came as a surprise, although the unexpected results were, of course, welcome as a means of advancing our knowledge. The first surprise was the intensity and extent of subsurface hydrothermal alteration. Beneath a capping of unaltered dacite and rhyodacite lava equivalent to surficial dredged samples, ranging from < 10 to 40 m thick at the different hole sites, there was a rapid transition to pervasively and thoroughly altered derivatives. These continued without interruption to the bottom of all four holes cored to depths between 100 and nearly 400 mbsf. Some igneous plagioclase is preserved in the lower section of Hole 1189B and at several thin intervals in other holes, but even these rocks bear the clear imprint of hydrothermal alteration.

A second surprise was the predominance of clay minerals in altered rocks throughout all levels of the cored sequence in places accompanied by or substituted by chlorite. The assemblages differed little between the high- and low-temperature sites, although there was telescoping at the former (and also from the fringe to the center of this site) of a vertical progression believed to reflect thermal gradient overall although kinetic effects probably related to fluid:rock ratios also play a role. The progression, obscured somewhat by superimposed bleaching in which pyrophyllite appears a significant phase, is marked by a change with increasing depth from cristobalite to quartz as the accompanying silica mineral and by increasing silicification. The predominance of clays at the low-temperature site is broadly consistent with a borehole temperature of 312°C measured there at 360 mbsf after allowing 8 days for rebound after drilling. Their persistence in holes at the high-temperature site, especially in country rock fragments within a stockwork zone under a chimney mound at the high-temperature site, was less expected. Here we were unable to define ambient temperature at depth, although borehole temperatures at the base of this stockwork (near 125 mbsf) increased from ~45°C soon after drilling to 68°C in our final logging run < 12 hr later.

Yet another surprise was the measured porosity of the altered rocks and qualitative indications of their high permeability. Excepting some especially vesicular and brecciated samples, average porosity is near 25%, decreasing slightly downhole in the more silicified assemblages. Furthermore, intense fracturing is endemic, as indicated by semicoherent core returned by ADCB drilling, especially by borehole imaging, and in retrospect by the intervals of very poor or zero core recovery in both ADCB and RCB holes. These characteristics will require profound modifications to our original expectation of how mineralizing fluids might pass upward through the system. The volume of hydrothermal pseudoclastic rocks, and of closely related hydrothermal breccias recovered at Sites 1188 and 1189, is striking. They may represent creation and evolution of major fluid pathways in hydrothermal systems, so far poorly understood.

The frequency of anhydrite in veins and breccia matrices, as vesicle fillings, and particularly within bleached intervals as disseminations in altered wallrock, was also unexpected in the drilled portion of the PACMANUS hydrothermal system. The anhydrite abundance falls with depth in holes at the diffuse vent site (Site 1188) and under the fringe of the focused vent site (Site 1189), yet it remains a conspicuous vein mineral even in the deepest intersections. The presence of anhydrite implies a major role for seawater infiltrated or swept into the rising column of hot hydrothermal fluid. The aforesaid porosity, permeability, and fractured nature of the altered volcanic rocks provides a likely explanation for this exceptionally significant feature of the system.

Pyrite is a conspicuous disseminated mineral at the 1% to 5% level in the altered rocks at both main sites, except in the lower, somewhat less altered succession in Hole 1189B, which possibly represents something akin to what economic geologists might call an outer, propylitic zone less influenced by passage of mineralizing fluids. Pyrite is also present in thin veins and vein networks throughout the system and is especially prominent in the stockwork zone underlying the elevated mound portion of the high-temperature chimney site where, locally, it forms centimeter-scale pods and veins of massive pyritic sulfide. Chalcopyrite and sphalerite are virtually nonexistent in core from the diffuse venting site, trace to minor components in the fringe Hole 1188A under the focused venting site, and conspicuous only in an interval of semimassive sulfide (represented by a single specimen) cored immediately beneath casing placed to 36 mbsf at the chimney mound (Hole 1189B).

Although diffusive fluid flow through porous rocks and fractures might explain the low temperature venting characteristics at Site 1188, we failed to core any structure likely to represent a major conduit (or its fringing halo) enabling focused venting of 280°C fluids at the chimney site (Site 1189). The cored stockwork zone appears rather too subdued in this respect and too free of minerals other than pyrite. However, core recovery was particularly poor in this stockwork zone, so such a feature might not have been sampled. Further evaluation of logging data may resolve this discrepancy, although it must also be emphasized that a vertical drill hole could easily miss altogether a subvertical feeder conduit, even though our holes at Site 1189 were spudded near the base of active seafloor chimneys.

Concerning the architecture of Pual Ridge, delineation of which was a subsidiary objective of Leg 193, we have established by drilling that the upper two-thirds of this volcanic edifice is constituted of felsic lavas (dacite and rhyodacite). Altered equivalents of massive vesicular units, and flow-banded units that are commonly brecciated or show clear autoclastic structures, comprise the greater bulk. Less frequent intervals with palimpsest perlitic structure denote formerly vitreous rocks, possibly lava rinds. Unequivocal volcaniclastic breccia and former volcaniclastic sandstone layers define paleoseafloor positions. At least five and probably more such horizons were recognized in Hole 1189B, defining a maximum value of 50 m for the average thickness of flow units. This compares favorably with the 30-m average thickness deduced from terracing on the crest and eastern flank of Pual Ridge. No evidence was found for the presence of slowly cooled intrusive units, but possible occurrences of relatively thin dikes are not ruled out.

Data acquired on the physical properties of rocks cored at the PACMANUS site will constitute major aids to future modeling of future marine geophysical surveys in island arc and convergent margin terrains. In particular, we have measured the first compressional wave velocities for intensely altered felsic volcanic rocks, albeit at atmospheric pressure. We also found that whereas surficial unaltered dacite lavas will contribute to surface and deep-tow magnetic intensity anomalies, a greater effect may arise from deeply buried horizons of altered rocks with hydrothermal magnetite. At the PACMANUS site, these are characterized by high natural remanent magnetism as well as high magnetic susceptibility.

Another major success of Leg 193 was confirmation that microbes flourish in subsurface rocks of the PACMANUS hydrothermal system. Bacteria were confirmed by direct counting and by ATP analyses in cores recovered from depths to 80 m under the diffuse venting at Site 1188 and to 129 m under the chimney field at Site 1189. Cultivation experiments on board proved that these deep biota flourish under anaerobic conditions in seawater at temperatures as high as 90°C, and higher temperature experiments will be conducted on shore.

The results of Leg 193 and the postcruise research to follow will impact on our ability to correctly interpret ancient geological environments in which orebodies are found and to develop improved strategies for finding them. They will also clarify a number of issues regarding thermal and chemical fluxes between the seafloor and the oceans in volcanic arc environments at convergent plate margins. Some 2000 samples have been taken ashore by the scientific participants for a variety of studies employing frontier as well as conventional analytical techniques. The results of these endeavors will flow into the scientific literature over coming years. Unraveling the sources and mixing behaviors of metals and fluids and establishing whether the PACMANUS hydrothermal field represents a growing ore system or a minor manifestation now in decline will be key topics in these enquiries.

In the long term, two requirements apply to further PACMANUS seafloor research. One is the need for close-spaced shallow drilling to delineate more fully the main hydrothermal conduits at the high-temperature chimney site. Such a program using the portable remotely operated diamond drill was intended to precede Leg 193, but it was abandoned for technical reasons. It is by no means unnecessary now. The other is for deeper drilling. Leg 193, with its innovations such as nested casing and hammer-in casing, has shown that the technology is in hand to attempt deeper penetrations than the near 400 m we achieved.

Penetrating the deepest, very high-temperature levels of a seafloor hydrothermal system has been proposed as a priority for IODP, and the PACMANUS hydrothermal field must now be considered a prime candidate for that activity. If the thermal gradient measured in Hole 1188F accurately represents rock temperatures, then we might predict that the igneous intrusion evidently responsible for the heat energy needed to drive this system, and arguably a significant source of metals and hydrothermal fluid, might lie as shallow as 1.5 km below the crest of Pual Ridge.