Abstract

Low soil nitrogen (N) limits maize production in Zambia. S₁ selection was used to select for tolerance to low N among ninety-six maize landraces during 2004 - 2007 in Zambia. The landraces were evaluated under low N, drought and optimal conditions; and selfed in a nursery, under optimal conditions. Data on grain yield (GY), number of ears per plant, leaf senescence and anthesis-silking interval were used to calculate selection indices. Fourteen S₁ lines, from each of the best four landraces under each environment and across all environments were evaluated under the three environments, and at the same time crossed to a tester. Twenty-two best S₁ lines under each environment and across were identified and also their testcrosses were evaluation under the three environments. Significant GCA effects for GY under low soil N were found, suggesting that population improvement under soil N stress was effective. Heritability for GY under low soil N conditions was low (0.38) implying that selection based on GY was ineffective. Ther_G for GY under low soil N and optimal environments was moderate (0.458), suggesting that selection for GY in one environment was not as effective as in the other. Low soil N tolerant landraces were identified and should be used to breed for the low soil N conditions.

Key words: Maize, landrace, heritability, correlation, nitrogen, tolerance, stress.