Analysis of resistance to Striga hermonthica in diallel crosses of sorghumB I GHaussmannauthorD EHessauthorB V SReddyauthorH GWelzauthorH HGeigerauthorParasitic flowering weeds of the genus Striga are major biotic constraints to sorghum (Sorghum bicolor (L.) Moench) production in sub-Saharan Africa. The agar-gel assay was used to evaluate stimulation of Striga hermonthica (Del.) Benth. seed germination by a complete F1 diallel involving nine sorghum cultivars and inbred lines. Striga populations from Mali and Niger were employed. The same genetic materials were planted in pot trials in both countries to observe striga plant emergence. Variation in hybrid performance was determined by general (GCA) and specific combining ability (SCA) effects, with preponderance of GCA, for both germination distance in the agar-gel assay and number of emerged striga. Reciprocal effects were significant only in the agar-gel assay and were unstable across striga populations. For lines and hybrids, estimates of broad-sense heritabilities were 0.97 and 0.91 for germination distance, and 0.38 and 0.58 for emerged striga, respectively. Only a weak positive relationship existed between in vitro germination distance and emerged striga number in the pot trial. Although selection for low germination distance has merit, valuable material with resistance mechanisms other than low stimulant production may be lost if these traits are not additionally assessed. Laboratory assays which allow a non-destructive, quick and economical screening for resistance mechanisms other than the low stimulant character are likely to increase the efficiency of breeding programs for striga resistance. The significant contribution of SCA effects indicates that thorough screening of testcrosses is indispensable for selection in hybrid sorghum breeding programs.Sorghum2000KluwerArticle