GEOLOGICAL IMPLICATIONS

The obvious question arising from the above interpretations of the test results is why such an elevated pore pressure exists near the base of the sediment column at Site 897. A possibly related question is why this clay has remained in an uncemented condition, especially considering its relatively high carbonate content.

The test core was recovered from the base of Subunit IIC, a calcareous turbidite and pelagic sequence, overlying about 20 m of noncalcareous terrigenous claystone (Subunit IIIA), which in turn overlies mass flow and polymict "breccia" deposits (Subunits IIIB and Unit IV). It is clear that porosities in the basal section of Subunit IIC and in Unit III are significantly higher than in the overlying sediments despite a wide scatter in the data within Hole 897D and between Holes 897C and 897D (Fig. 10). Some of this scatter may result from small-scale lithologic variations, but the general trend must reflect stress conditions. The test core lay within the zone of downward increasing porosity, which appears to increase smoothly across the lithologic unit boundary between Units II and III.

This zone of downward-increasing porosity is consistent with the low yield strength and high pore-fluid pressure interpreted from the tests reported here. Moreover, the location of the section of downward increasing porosity at the base of the post-rift sediments strongly suggests that both the elevated and implied high P are associated with fluid expulsion from the basement and/or the overlying massive breccia unit. Such elevated and P are not expected from consolidation-induced permeability reduction, assuming a typical permeability on the order of 10^-17 m² for a clay with a porosity of 32% (Lambe and Whitman, 1969) and with 45 m.y. of consolidation under a relatively thin sediment cover. Although the data are insufficient for quantification, it seems most reasonable that such a period of elevated P must have begun early in the postrift depositional history of Site 897 to arrest consolidation at a relatively early stage. That the tested sediment has a low yield stress implies that the source of the high-pressure fluid must have persisted until recently, or else these uncemented basal sediments should have resumed consolidation.

Additional corroboration of the proposed fluid expulsion from the basement is derived from the intense fracturing and veining described by the Shipboard Scientific Party (1994a). Of particular interest is the reported upward increase in brecciation and calcification in the basement. These observations support not only a high degree of upward fluid transport but also a very low effective stress state. Because Site 897 lies on a ridge crest there is probably a component of lateral flow through a more permeable basement toward a relatively lower pressure environment at the sediment surface. The relatively thin sediment cover over the ridge crest should induce a steeper pressure gradient toward the free-water surface there and lead to concentrated flow (Fig. 11). The source of this fluid remains unknown; two speculations are that meteoric water might have migrated down a structurally deeper décollement, or conversely the water might have been generated in the mantle beneath the rifted crust through pressure-release mechanisms. The large flow of water through the system could be responsible for the lack of cementation in the sediments, but the chemistry associated with this flow is not apparent.

High pore pressures near the base of the postrift sediment section should also engender a zone of structural weakness with a disposition toward gravity sliding. This may not be likely on a ridge top as at Site 897, but might be important elsewhere along rifted margins.