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

Authors: SWARTS, KELLY - Austrian Academy Of Sciences; BAUER, EVA - Technical University Of Munich; GLAUBITZ, JEFFREY - Cornell University; HO, TIFFANY - Cornell University; JOHNSON, LYNN - Cornell University; LI, YONGXIANG - Chinese Academy Of Agricultural Sciences; LI, YU - Chinese Academy Of Agricultural Sciences; MILLER, ZACHARY - Cornell University; SCHON, CHRIS-CAROLIN - Technical University Of Munich; WANG, TIANYU - Chinese Academy Of Agricultural Sciences; ZHANG, ZHIWU - Washington State University; Buckler, Edward - Ed; Bradbury, Peter

Submitted to: Heredity
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/25/2021
Publication Date: 4/22/2021
Citation: Swarts, K., Bauer, E., Glaubitz, J.C., Ho, T., Johnson, L., Li, Y., Li, Y., Miller, Z., Schon, C., Wang, T., Zhang, Z., Buckler Iv, E.S., Bradbury, P. 2021. Joint analysis of days to flowering reveals independent temperate adaptations in maize. Heredity. 126:929-941. https://doi.org/10.1038/s41437-021-00422-z.
DOI: https://doi.org/10.1038/s41437-021-00422-z

Interpretive Summary: Controlling the time to flowering of maize is the key element of making it adapted to its local environment. The genetics of flowering have been studied in maize adapted all different locations to the world from the tropics, China, Europe, and the US. In this study, the knowledge of DNA sequence variation from across all the world maize varieties was integrated, and then genetic controllers were flowering time mapped. The genetic controllers of flowering in the Americas and Europe were highly overlapping, while Chinese germplasm appears to have substantially independent genetics controlling its flowering time variation. Chinese germplasm likely represents an independent temperate adaptation. Prediction accuracy of flowering time across the Americas was as high as 90% and was high across the non-Chinese germplasm. These models can be used move allelic variation across germplasm pools and increase adaptation, and they suggest there may be opportunities to improve temperate adaptation by combining American and Chinese germplasm.

Technical Abstract: Domesticates are an excellent model for understanding biological consequences of rapid climate change. Maize (Zea mays ssp. mays) 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.