Joint analysis of days to flowering reveals independent temperate adaptations in maize

Kelly Swarts,Eva Bauer,Jeffrey C. Glaubitz,Tiffany Ho,Lynn Johnson,Yongxiang Li,Yu Li,Zachary Miller,Cinta Romay,Chris-Carolin Schön,Tianyu Wang,Zhiwu Zhang,Edward S. Buckler,Peter Bradbury

DOI: https://doi.org/10.1038/s41437-021-00422-z

IF: 3.832

2021-04-22

Heredity

Abstract:Domesticates are an excellent model for understanding biological consequences of rapid climate change. Maize (<i>Zea mays</i> ssp. <i>mays</i>) was domesticated from a tropical grass yet is widespread across temperate regions today. We investigate the biological basis of temperate adaptation in diverse structured nested association mapping (NAM) populations from China, Europe (Dent and Flint) and the United States as well as in the Ames inbred diversity panel, using days to flowering as a proxy. Using cross-population prediction, where high prediction accuracy derives from overall genomic relatedness, shared genetic architecture, and sufficient diversity in the training population, we identify patterns in predictive ability across the five populations. To identify the source of temperate adapted alleles in these populations, we predict top associated genome-wide association study (GWAS) identified loci in a Random Forest Classifier using independent temperate–tropical North American populations based on lines selected from Hapmap3 as predictors. We find that North American populations are well predicted (AUC equals 0.89 and 0.85 for Ames and USNAM, respectively), European populations somewhat well predicted (AUC equals 0.59 and 0.67 for the Dent and Flint panels, respectively) and that the Chinese population is not predicted well at all (AUC is 0.47), suggesting an independent adaptation process for early flowering in China. Multiple adaptations for the complex trait days to flowering in maize provide hope for similar natural systems under climate change.

genetics & heredity,ecology,evolutionary biology