Field experiments were conducted in the Nigerian savanna during the 1994 an
d 1995 wet season to assess the influence of nitrogen application on weed i
nterference, and determine the critical period of weed competition in a mai
ze/rice intercrop. Treatments were laid out in a split-plot design with thr
ee replications. Three levels of nitrogen (0, 60 and 120 kg/ha) were assign
ed to the main plot, and ten periods of weed interference to the sub-plots.
In both seasons, N-application resulted in significant increases in grain
yields of maize and rice in the plots that were kept weed-free for 6 weeks
and more. Uncontrolled weed growth resulted in 83.03% and 93.85% reductions
in average grain yield of maize and rice respectively. Weed control for 9
weeks after sowing at 120 kgN/ha gave the best crop yields. However, it is
recommended that for resource-poor farmers, rather than keeping plots weed-
free for long periods, n-application of 60 kg/ha in combination with a weed
-free regime of 6 weeks after sowing will be adequate to provide a good cro
p yield for the intercrop.