Omari M. OdongoGeorge A. OmbakhoNjoka, Fredrick M.Kiplangat, Charles Melil2012-11-052012-11-052012-11-05http://ir-library.ku.ac.ke/handle/123456789/5949Department of Plants and Microbial Sciences:77p. The SB 191 .M2M4 2012Maize is the most important staple food crop and single agricultural commodity in Kenya. It is grown on 1,600,000 hectares in the country due to its adaptability and productivity. Maize is used as food and livestock feed as well as raw material for industry for making starch and oil extraction. Major constraints to maize production in Kenya include moisture stress, low soil fertility and lack of access to suitable varieties. This study was conducted to select hybrids with improved yield and identify testers for grain yield, determine the magnitude of genetic variability in maize hybrids for yield and its elements and to compare adaptability of top cross and double cross hybrids under varying agro-climatic conditions of parts of Western Kenya. The experiment was conducted in three sites, Kitale in Trans-Nzoia County, Kakamega in Kakamega County and Lurende in Bungoma County, representing the highlands, moist transitional and midlands, respectively. This was done during the onset of the long rains of 2009. Thirty four experimental hybrids and two commercial maize varieties (standards) were planted in a 6x6 balanced lattice design with three replicates in each of the three sites. Morphological and phenotypic traits were observed and recorded. Collected data was subjected to analysis of variance (ANOV A) using the general linear model SAS. The results showed that there was significant difference in grain yield in Kitale double cross (F= 0.83, df = 16, P<0.05) and also Kitale top cross (F = 0.66, df = 18, P<0.05. In Kakamega both the double cross (F = 0.84, df= 16, P<0.05) and top cross (F = 1.14, df= 18, P<0.05) showed significant differences, while in Lurende the double cross (F = 0.64, df = 16, P>0.05) showed no significant difference however top cross (F = 0.56, df = 18, P<0.05) showed significant difference. The best yielding hybrid across the three sites was entry 12 (64xF)x R'2C" a top cross hybrids which produced 14.25 ± 0.13 t ha-' from Kakamega, Entry 3 in Kakamega also produced promising yield ,while Lurende had entry 3 and 1 0 as the best yielders, in Kitale the best yielders was recorded from entry 3 and 12. The low yielders were mostly double cross hybrids. The lower grain yields in Kitale can be associated with insufficient rainfall during the critical stages of kernel formation. The grain yields for top cross hybrids for the three sites were higher than double cross grain yield in the three sites. Results showed that most tall hybrids in Kitale and Kakamega had more grain yield than the shorter ones. The study showed that, 1000-grain weight had positive and significant correlation with grain yield in all the three sites. The highest correlation (r = 0.7361), was observed between ear height and plant height in Lurende. Four hybrids (AxF)xR'2C", (95xF)x82x93, (64xFxR'2C" and 56x44x R'2CIO indicated promise in all the sites, while entry 22 (44xF)x(82x93) was particularly good in Kitale, entry 14 (AxF)x(50x82) and 30 (8xF)x R'2C" was specific for Kakamega and entry 27 (64x8)x(82x93) did well only in Lurende and are recommended for advanced evaluation and eventual release to farmers. It was also found that AxF and R'2C" are good testers for grain yield, and that the combinations where one of the parental crosses had one of the parents with "F" recorded promising yields. The results will be used to advise both the farmers and researchers on the hybrids they cultivate or study and their relationship to agronomic value. This information will also be used in maize breeding programmes. The present study revealed considerable amount of diversity among the tested hybrids which could be manipulated for further improvement in maize breeding.enHybrid cornKenyaCornBreedingThesis