Covariation between line and testcross performance for reduced mycotoxin concentrations in European maize after silk channel inoculation of two Fusarium species

Fusarium spp. in maize can contaminate grain with mycotoxins harmful to humans and animals. Breeding and growing resistant varieties is one alternative to reduce contamination by mycotoxins. Little is known about the population parameters relevant to resistance breeding. The objectives of this study were to draw conclusions on breeding of reduced mycotoxin concentrations of deoxynivalenol, zearalenone and fumonisins, and resistance to ear rot after silk channel inoculation with F. graminearum or F. verticillioides, respectively. For that, variation and covariation of line and testcross performance and correlations between both species and between mycotoxin concentrations and ear rot resistance were calculated. Means of ear rot after infection with F. graminearum were higher than with F. verticillioides. Moderate phenotypic correlations (r = 0.46–0.65) between resistances to both Fusarium spp. implicate the need of separate testing. Analyses of variance revealed significant (P < 0.01) differences among lines in line and testcross performance for 30–60 entries per maturity group. Multi-environmental trials for accurate selection are necessary due to significant (P < 0.1) genotype × environment interactions. High genotypic correlations between ear rots and mycotoxins (r ≥ 0.90), and similar heritabilities of both traits, revealed the effectiveness of indirect selection for mycotoxin concentrations based on ear rot rating after inoculation. Moderate genotypic correlations between line and testcross performance were found (r = 0.64–0.83). The use of one moderately to highly susceptible tester is sufficient since genotypic correlations between testcrosses of different testers were high (r = 0.80–0.94). Indirect selection for testcross performance based on line performance is less effective than selection based on mycotoxin concentrations. Consequently, selection for resistance to ear rot and mycotoxin accumulation should be started among testcrosses tested first for general combining ability based on ear rot data in parallel with a negative selection for line per se performance.