Sediments travel, make waves in gravel.

When a magnitude 7.2 earthquake rattled the Grand Banks (Newfoundland) coast in 1929, a number of underwater telephone and telegraph cables lying within 100 kilometers of the epicenter suddenly snapped. Over the next 13 hours, other cable breaks occurred, progressing farther down the continental slope. Scientists attributed the breaks to a turbidity current, a heavy slurry that raced down the slope at speeds up to 55 km per hour and extended 700 km after a large underwater landslide threw muds into suspension. In fact, the Grand Banks cable breaks offered one of the first compelling demonstrations that turbidity currents exist in deep waters.

Now, the Grand Banks slope has produced another scientific first. Over the last two years, a team of Canadian and U.S. researchers has been using the Sea MARC I (sidescan sonar) and a submersible to survey the underwater effects of the 1929 quake. They have discovered a field of "gravel dunes"--2-to 3-meter-high ridges 50 to 100 m apart, standing below 1,500 to 4,500 m of water. "this is a new type of bedform which we associated with quite fast flows of the [1929] turbidity current," says Alexander Shor at Lamont-Doherty Geological Observatory (LDGO) in Palisades, N.Y., who has coauthored a paper with David Piper at the Geological Survey of Canada in Dartmouth, Nova Scotia, and John Hughes Clark at Dalhousie University in Halifax, Nova Scotia.

According to Shor, "smaller bedforms develop from fast tidal currents in shallow water, but normally the flows in deep water aren't strong or persistent enough to form [such large features in gravel]." But as more surveys are conducted with sonar systems along coastlines where strong turbidity currents might have flowed, it's likely that more gravel dunes will turn up. Already Albert to Malinverno at LDGO and co-workers report finding similar beforms off the coast of Nice, France.