PRELIMINARY SCIENTIFIC ASSESSMENT

IODP Expedition 305 was planned in conjunction with Expedition 304, with the overarching goals as noted in "Scientific Objectives." Our overall objectives encompass investigating the nature, evolution, and geophysical signature of the oceanic lithosphere accreted at slow-spreading ridges. More specifically, the aims were (1) to address the formation of OCCs and (2) to transect a section, possibly an alteration front, corresponding to the transition to rocks with a seismic velocity of 8 km/s—commonly interpreted to represent fresh residual mantle peridotite. Objectives related to drilling the hanging wall were attempted during Expedition 304, and the footwall objectives were pursued during both Expeditions 304 and 305.

Clearly, the set of working hypotheses related to drilling was only partly addressed because of logistical failures at the hanging wall site. In brief, all attempts to start/case a hole in these young basalts with the equipment on hand were unsuccessful. However, the coring results at the footwall site directly document some aspects of the detachment-controlled tectonics. The detachment system at Atlantis Massif must have developed under low-temperature (greenschist facies) conditions, and the fault zone(s) must be strongly localized. The existence of a capping fault is supported by the fragments of fault rocks and metabasalts recovered during the series of shallow cores drilled at the end of Expedition 304.

The initiation of the deep hole in the footwall during Expedition 304 was clearly a success, with ~400 m cored and a high-quality series of downhole geophysics logs. The 1415.5 m depth reached in Hole U1309D during Expeditions 304 and 305 provides a unique set of geological data to address fundamental questions related to the formation and evolution of core complexes (working hypotheses 2, 3, and 4) and, more generally, to the accretion of oceanic lithosphere. The most surprising finding in the 1043 m of recovered rocks from Site U1309 is the dominantly gabbroic section from a slow-spreading ridge hypothesized to be magma-starved. It is intriguing that all holes drilled in OCCs have recovered dominantly gabbroic rocks (Robinson, Von Herzen, et al., 1989; Dick, Natland, Miller, et al., 1999; Pettigrew, Casey, Miller, et al., 1999; Kelemen, Kikawa, Miller, et al., 2004; Cannat, Karson, Miller, et al., 1995). A simple lack of magma production cannot explain the development of low-angle detachment faults at slow- and ultraslow-spreading ridges. The lack of significant moderate- to low-temperature (i.e., <520°C) tectonic rotation inferred from the shipboard paleomagnetic measurements suggests that the rolling hinge model must be reconsidered.

Core from Hole U1309D offers a unique opportunity to address the evolution of intrusive sequences at slow-spreading centers. This section is unique in that it represents the most primitive interval of lower oceanic crust ever recorded, opening a window into lowermost crustal accretion processes. Among the many fresh and beautiful rocks recovered in Hole U1309D, three meter-scale intervals of dunitic troctolite were recovered below ~1000 mbsf. These ultramafic rocks, presenting cumulate textures, are locally very fresh (as low as <1% alteration) and, therefore, unique in ocean drilling records. These rocks likely represent the primitive end-member compositions of igneous oceanic crust. Alternatively, the ultramafics could be relics of mantle peridotite from a crust–mantle transition zone, where it reacted with large volumes of percolating melt. The origin of these rocks will be the focus of detailed postcruise studies. The occurrence of similar olivine-rich rocks at mid-ocean ridges is rare; in Hole U1309D their abundance reaches 5% of the core.

One rationale for drilling a deep hole in the core of Atlantis Massif was a convergent set of geophysical data (NOBEL refraction data; air gun refraction data; multichannel reflection data, and gravimetry data) that indicated the presence of fresh mantle at shallow (~800 mbsf) depth and of a gradient in V_P, presumably related to gradational serpentinization of peridotite. Site U1309, based on the available survey data (analysis of reflection and refraction seismic data, gravimetry data, swath mapping, seafloor sampling by dredges, and submersible dives), was located a few hundred meters south of one of the NOBEL experiment lines (Fig. F2). It was assumed that the high-velocity material extended beyond the local area of the NOBEL seismic lines. We know now that this assumption was incorrect. The olivine-rich troctolites recovered from Hole U1309D, if fresh, are potential candidates for propagating seismic compressional waves at >8 km/s. However, the intervals of fresh rock found are not thick enough to account for the NOBEL results. We did not drill through a serpentinization front, nor did we reach fresh mantle peridotite. We were, therefore, unable to directly test working hypotheses 5, 6, and 7.

Our findings dictate that more complex analysis of existing (and any future) seismic data is required. Two-dimensional variability had been considered as part of the prior refraction analysis, but inversion results were strongly dependent on starting model geometry. A first step in reanalysis of existing refraction and reflection data is to include physical property and logging information from Hole U1309D. Depending on the results of the better-constrained modeling, additional seismic acquisition may well be quite worthwhile. The rather continuous nature of key seismic reflectors across Atlantis Massif gives the impression that broader scale seismic structure is relatively continuous. Again, now we have impetus to look in more detail than the work done by Canales et al. (2004).

The downhole measurement program during Expeditions 304 and 305 was generally successful. For the first time, we were able to log in lower crustal rocks, to ~1400 mbsf. We obtained a complete set of very high quality triple combination (triple combo) and Formation MicroScanner (FMS) tool string data, certainly the best ever recorded in igneous rocks. This, together with the high recovery, offers a unique opportunity to establish the core–logging structural integration. Unfortunately, the failure of the sonic and VSP tools, and, eventually, the rough sea conditions, did not allow us to acquire downhole seismic data during the last logging operations (below 850 mbsf).

The drilling objective of Expedition 305 was to continue drilling and coring Hole U1309D, to the targeted depth of >700 mbsf, in the presumed high seismic velocity zone. We reached 1415.5 mbsf (i.e., ~200%) of the expected final depth. The overall recovery in Hole U1309D was 75%, exceeding by far the standard average recovery (~30%) in hard rock boreholes (Fig. F5), with the exception of ODP Hole 735B, in which the recovery of gabbroic rocks was exceptionally high (~86%).

We fulfilled the objective of deep drilling into an OCC to study its formation and evolution. We left IODP Site U1309 with a 1415 m long crustal section from Atlantis Massif, which will be the subject of many postcruise research projects. Hole U1309D is a unique, open, deep hole into the oceanic crust, ready for drilling deeper in the future and/or for in situ experiments, and we view it as an important legacy to IODP and to the marine geosciences community.