Tapping Native Diversity: Characterizing the Genetic Architecture of Complex Traits in Maize Landraces

Landraces represent a good source for useful alleles to improve yield, especially under conditions of biotic and abiotic stress. Until recently, the identification and deployment of such alleles has been slowed by the lack of phenotypic association with genetic markers and their linkage with numerous undesirable alleles. In addition, because of historic recombination, landraces offer the potential for resolving at high resolution regions associated with quantitative variation.

Here, we present the results for performing genome wide association for flowering time and plant height in a panel of 3,500 landraces from Latin America. Using Genotyping by Sequencing, we generated around 1 million SNPs for the accessions sampled. We performed phenotypic evaluation on 23 locations across Mexico for 2 years. With these data, we have identified association between flowering time and genes known to be part of its genetic architecture. Furthermore, we report association at hundreds of novel genes, as well as association at large structural variation like inversions and centromeres. For plant height, we show significant association at various genic regions. For both traits, we explore the predictive ability of the most significant markers and contrast the differences in predictive ability between both traits according to their genetic architecture.

This study provides better knowledge about the genetic architecture of two agronomically important traits, highlights the effect of genome structure variation on quantitative traits, and reveals the potential for uncovering useful alleles from very diverse populations