Phenotypic Characterization of a Maize Association Mapping Panel Developed for the Identification of Aspergillus Flavus and Aflatoxin Accumulation Resistance Genes.

Phenotypic characterization of a maize association mapping panel developed for the identification of Aspergillus flavus and aflatoxin accumulation resistance genes. Marilyn L. Warburton W. Paul Williams Gary Windham Seth C. Murray Wenwei Xu Leigh Hawkins Chris Daves Brien Henry The improvement of aflatoxin and Aspergillus flavus resistance in maize is urgent, but hampered by low heritability, high GxE, a highly quantitative nature, and difficulty in accurate phenotyping. New resistant germplasm and techniques for identifying genes or markers for resistance will give breeding efforts a boost. To this end, a panel of 300 diverse inbred lines has been developed for association mapping. Lines were testcrossed to Va35, a susceptible, southern adapted inbred line, and testcrosses were phenotyped in replicated field trials in four sites over two years. Traits measured include alflatoxin levels via Vicam AflaTest assay, fungal biomass levels via qPCR, infection estimation via NIRS, earworm damage ratings, husk coverage, and maturity. Not all traits were phenotyped at all sites. Germplasm known to be resistant to aflatoxin accumulation tended to perform well, as expected, and new lines, including diverse southern adapted US inbreds and CIMMYT derived inbreds, also displayed resistance. These may prove to be valuable sources of new breeding lines and/or resistance genes. Means of all traits were distributed over a wide range, and standard deviations associated with toxin and fungal biomass tended to be high. Analysis of variance showed that there was significant variation between replications within a field, between fields, and across years. Heritability for each trait was low, but there was sufficient genetic variation to assume that future association studies will successfully identify genetic factors related to resistance. Correlations between traits and between environments varied. Over this wide range of genetic backgrounds, the correlation between NIRS and Vicam values was near zero. The correlation between qPCR and Vicam values was high, with notable exceptions. Information from these lines and these measurement techniques are immediately applicable to genetic improvement of this trait, and future association mapping using this panel should yield new resistance factors.