This chapter is a synthesis of geochemical information now available for the gabbroic rocks of Ocean Drilling Program (ODP) Hole 735B. The rocks were unroofed along a detachment fault near a ridge-transform intersection at this slowly spreading ridge (Fig. F1) some 9-13 m.y. ago (Dick et al., 1991b). They comprise the only long section of the lower ocean crust so far obtained in the ocean basins. Average recovery was 87% during both Legs 118 and 176 and ranged from 90% to 95% for most of the core.
A short history of the planning and drilling of Hole 735B and a summary of drilling objectives are presented in Natland and Dick (History Chap., this volume). The core and its setting are described in several places (Shipboard Scientific Party, 1989; 1999c; Dick et al., 1991a, 2000; Robinson et al., 2001; Natland and Dick, 2001). Briefly, Hole 735B was drilled atop the flat summit of Atlantis Bank, which is a portion of an uplifted transverse ridge on the eastern side of the Atlantis II Fracture Zone near the Southwest Indian Ridge. The summit was flattened by waves when the massif was lifted to sea level from beneath a rift valley floor, currently about 110 km to the north. Since then, seafloor spreading has carried Atlantis Bank to the south, and it has subsided to a depth of 719 m at the site of Hole 735B. The core was first occupied in 1987 during Leg 118 and drilled to a depth of 504 meters below seafloor (mbsf). It was reoccupied in 1997, and the drilling was carried to a depth of 1508 mbsf, almost all of it in gabbro.
Core stratigraphy in prior reports is based mainly on shipboard data, including whole-rock and trace element measurements obtained by X-ray fluorescence (XRF). Such data have traditionally provided the basis for defining the chemical stratigraphy of basalts obtained by ODP and its predecessor, the Deep Sea Drilling Project. During Legs 118 and 176, more than 330 chemical analyses of gabbros were obtained using the shipboard XRF unit. Postcruise analyses have expanded this total to 862, plus 3 analyses of the single diabase dike that intruded the core. Many petrologists would likely prefer using mineral analyses rather than bulk compositions to define gabbro stratigraphy. However, such a strategy at the outset would have deferred definition of any stratigraphy for Hole 735B until years after Leg 176, and mineral data sets are still not sufficiently integrated for us to attempt this just yet. For the present, then, we retain bulk compositions as the principal basis of defining the gabbro stratigraphy, realizing that it is likely to be revised when mineral data can be more fully integrated into a general scheme and other synoptic studies incorporated. Nevertheless, we use some mineral data to augment our discussion and to consider particular problems of petrogenesis.
Hole 735B gabbros record extended differentiation of abyssal tholeiite magma, including formation of oxide-rich ferrogabbros and felsic, or more generally granitic, veinlets that together comprise some 24% of the core. The formation of these rocks and their relationship to zones of deformation in the core was a major concern of the igneous petrologists of Leg 118 and also of some of those of Leg 176. In the following, however, we focus instead on the 76% of core that is more primitive olivine gabbro, troctolitic gabbro, and troctolite, which we term the olivine gabbro suite, which likely was emplaced beneath the rift valley floor before much deformation and extended differentiation took place. These rocks, more than the oxide gabbros, are central to the problem of accretion of the lower ocean crust at a slowly spreading ridge.
Until now, views on this have largely been structured by the results of explosion seismology. Not even ophiolites have provided much insight on the deep structure of slowly spreading ocean crust. Hole 735B currently provides the only detailed lithologic perspective on gabbros at any type of ocean crust obtained thus far by drilling. Part of our task will therefore be to disentangle some of the overlapping threads of late-stage deformation and deformation and reconstruct the section as it originally was. We shall try to imagine what the rocks tell us about the process of accretion beneath the rift valley before or at least while only the earliest stages of deformation were taking place.