The classic studies of Brenchley and Warington (1930) and numerous other investigations have demonstrated that large reserves (pools) of weed seeds accumulate in arable soils. Work reviewed by Kropad (1966) and Roberts (1981) shows that the number of seeds per m2 in the upper 15-25 cm of soil in cultivated fields may be as high as 70,000-90,000, and up to 95% of these may be contributed by annuals. Soil samples collected under vegetation of known age and history suggest that buried seeds of some weedy species may remain viable for 50100 years or more (e.g., Livingston and Allessio 1968, Oosting and Humphreys 1940). Seeds of 3 of 21 species buried in 1879 by W. J. Beal were viable after 100 years (Kivilaan and Bandurski 1981). In another experiment started by J. W. T. Duvel in 1902, buried seeds of 36 of 107 species were viable after 39 years (Toole and Brown 1946). Researchers have known for decades that many buried seeds need to be brought to the soil surface before they will germinate. To understand how germination of buried seeds is controlled, plant biologists have studied the re-