Appreciable genetic correlation between inbred lines and testcrosses facilitates breeding for resistance to Fusarium head blight in hybrid rye (Secale cereale)

Abstract

Hybrid breeding is based on selection of inbred lines in early generations. A sufficient genetic correlation between inbred lines and testcrosses is, however, the prerequisite for a higher selection gain in the hybrids. Therefore, we investigated this crucial parameter for Fusarium head blight (FHB) resistance on 60 inbred lines each of the heterotic groups Petkus and Carsten and their corresponding crosses with two testers each at four (inbred lines) or six (testcrosses) environments (location × year combinations). FHB severity measured as percentage of infected spikelets per plot was used as resistance trait, and a correction was made by using flowering time as covariate. Variances for genotype and genotype–environment interaction were high, and the heritability was estimated .84 and .74 for the line per se performance (Petkus and Carsten) and .71 and .78 for the general combining ability (GCA). For both inbred lines and testcrosses, negative correlations with plant height were found ranging from −.20 to −.51. FHB severity was considerably higher for the inbred lines than for the testcrosses. Genetic variance was lower for testcrosses, and dominance effects of the testers and heterosis may be key factors reducing the general infection level of three‐way hybrids. The effect of the tester (i.e., difference between testers) was very small and only significant for the Carsten group. The variance for specific combining ability (SCA) was very small, too. This, and high genetic correlations between line per se and GCA (r = .82 for Petkus and r = .72 for Carsten), does allow or even favour the selection based on line per se performance. Still, in the last generation before the official trials, testcrosses should also be tested for FHB resistance and mycotoxin contents to reliably improve food and feed quality in rye.