A 3-yr field experiment was initiated in 2001 to evaluate weed suppression and sweetpotato productivity in three organic sweetpotato production systems. Organic systems were (1) compost and no cover crop with tillage (Org-NC), (2) compost and a cover crop mixture of hairy vetch and rye incorporated before transplanting (Org-CI), and (3) compost and the same cover crop mixture with reduced tillage (Org-RT). A conventional system with tillage and chemical controls (Conv) was included for comparison. Suppression of monocot and dicot weed density and biomass was similar between Org-NC and Org-CI each year, and were frequently similar to Conv. Org-RT was as effective as Org-NC in controlling dicot weed density and biomass each year, but did not suppress monocot weeds. At sweetpotato harvest, an increase in cover crop surface residue biomass in Org-RT was associated with a decrease in cumulative total weed density (R 2 = 0.43, P = 0.0001); however, the amount of that residue was insufficient to suppress late-emerging monocot weeds. Total sweetpotato yield in Org-RT was at least 45% lower than other systems in 2002 and was most likely due to an increase in monocot weed density and biomass concurrent with a decrease in sweetpotato vine biomass. Total sweetpotato yield was similar among all systems in 2001 and 2004; the exception was lowest yields obtained in the Org-RT system in 2002. Organically managed sweetpotato with or without an incorporated cover crop produced sweetpotato yields equal to conventionally managed systems despite difficulties controlling weeds that emerged later in the season.

Arrowleaf sida (Sida rhombifolia) is a tropical weed that grows in disturbed areas. Its control by hand pulling and mowing are only partially effective because arrowleaf sida is difficult to pull and quickly sprouts after cutting. We studied the seed-bank dynamics of S. rhombifolia in the region of Campinas, São Paulo State, Brazil for 2 years, following a known number of seeds placed artificially at five burial depths. Predators and pathogens were responsible for the loss of c. 40% of the seeds during the first year of the experiment. The importance of predation declined with depth. Mortality due to fungal attack did not differ among the burial depths, except for seeds placed at the surface (where it was negligible). Seedling emergence only occurred during the first year, after the beginning of the rainy period and when the study area was still free of a dense vegetation cover. Seedlings did not emerge from depths greater than 5 cm. We conclude that seedling emergence and death are significant causes of seed losses from the seed bank of S. rhombifolia but their relative importance is highly dependent on depth of burial.