Ocean Drilling Program Site 906 was drilled through a 5-km-wide, buried slope canyon to learn when and how this feature formed, was maintained, and subsequently filled. Seismic profiles show that it is the oldest Neogene canyon in the region, and that it marks the beginning of major slope bypass that continued intermittently to Pleistocene time. Sampling the oldest material within the canyon (pre-reflector m2) to determine the minimum canyon age, and tracing the youngest surface cut by the canyon to a nearby site (reflector m3) to determine the maximum age, shows the canyon formed between 13.5 ± 0.5, and 12.4 ± 0.5 m.y.
  The first sediment that accumulated within the canyon is a 57-m breccia deposit shed from the adjacent walls; matrix-supported debris flows of similar age and lithology were sampled at Site 905, 52 km southeast at 2800-m water depth on the continental rise. At both sites this material lacks shallow-water sand, and together with the flat (not V-shaped) canyon cross-section cutting into reflector m3 suggests that the canyon formed sometime after 13.5 ± 0.5 m.y. as a result of sediment failure that widened and lengthened the scar during headward erosion. By this process the canyon connected with a sediment source on the adjacent shelf, probably in less than ~0.5 Ma, and 60 m of mostly quartz sand turbidites accumulated above the Site 906 breccia; no equivalent turbidites have been found outside the canyon.
  Reflector m2 (12.4 ± 0.5 m.y.) is the oldest seismic reflector draping both the canyon and the adjacent slope, and it matches the top of an 82-m unit of laminated siltstone at Site 906. Seismic mapping beneath the upper slope shows this laminated unit completely filled the canyon 5 km landward of Site 906; furthermore, this unit lapped out against the basal breccias and turbidites just 7 km seaward of Site 906, demonstrating the head of the canyon filled first. This left a 250-m high, headless canyon on the middle and lower slope that has subsequently been utilized and completely filled by overlying canyon systems. Younger Miocene canyons seen in nearby profiles reveal a similar history.
  This evidence emphasizes that sediment failure and headward erosion are important precursors to shelf-edge sands finding a ready conduit to the deep sea. Sea-level fall can thus be a secondary process in slope canyon formation. Because canyons are buried on the upper slope before those on the lower slope, canyon piracy is more common in the latter setting. Consequently, the stratigraphic record of lower slope canyons can be especially complex.

Date of initial submission: 7 March 1995
Date if acceptance: 31 October 1995

Return to Contents of Leg 150
Return to Contents of Scientific Results