INTRODUCTIONSubmarine canyons continued to be a major area of interest for Francis P. Shepard throughout his career. His Submarine Geology textbooks, which were revised over the years, devoted substantial space to a review of the world's canyons that were known at the time; these reviews were a standard resource for several decades of marine geologists (e.g., Shepard, 1973). Shepard and his coworkers observed that canyon size was not directly related to the size of the rivers that fed them. In fact, many major canyons were not fed directly from rivers but fed from littoral drift transport of beach sediment, for example, the La Jolla Canyon (Shepard, 1973;Shepard and Dill, 1966). As our ability to map the deep-sea floor has improved since Shepard's pioneering work, it has become clear that the size of submarine canyons also shows no simple relation to the size of the turbidite systemssubmarine fans, abyssal plains, slope basin deposits, etc.-fed through the canyons.
ABSTRACTSubmarine canyons are the most important conduits for funneling sediment from continents to oceans. Submarine canyons, however, are zones of sediment bypassing, and little sediment accumulates in the canyon until it ceases to be an active conduit. To understand the potential importance in the rock record of any given submarine canyon, it is necessary to understand sediment-transport processes in, as well as knowledge of, deep-sea turbidite and related deposits that moved through the canyons.
There is no straightforward correlation between the final volume of the sedimentary deposits and size of the associated submarine canyons. Comparison of selected modern submarine canyons together with their deposits emphasizes the wide range of scale differences between canyons and their impact on the rock record.Three of the largest submarine canyons in the world are incised into the Beringian (North American) margin of the Bering Sea. Zhemchug Canyon has the largest cross-section at the shelf break and greatest volume of incision of slope and shelf. The Bering Canyon, which is farther south in the Bering Sea, is first in length and total area. In contrast, the largest submarine fans-e.g., Bengal, Indus, and Amazon-have substantially smaller, delta-front submarine canyons that feed them; their submarine drainage areas are one-third to less than one-tenth the area of Bering Canyon. Some very large deep-sea channels and turbidite deposits are not even associated with a significant submarine canyon; examples include Horizon Channel in the northeast Pacific and Laurentian Fan Valley in the North Atlantic. Available data suggest that the size of turbidity currents (as determined by volume of sediment transported to the basins) is also not a reliable indicator of submarine canyon size.