Unoccupied aerial systems discovered overlooked loci capturing the variation of entire growing period in maize. Issue 2 (19th May 2021)

• Record Type: Journal Article
• Authors: Adak, Alper; Murray, Seth C.; Anderson, Steven L.; Popescu, Sorin C.; Malambo, Lonesome; Romay, M. Cinta; de Leon, Natalia
• Is Part Of: œplant genome. Volume 14:Issue 2(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-05-19
• Subjects: Plant genomes -- Periodicals; Plant genome mapping -- Periodicals; 572.862
• DOI: 10.1002/tpg2.20102
• Languages: English
• ISSNs: 1940-3372

Abstract

Abstract: Traditional phenotyping methods, coupled with genetic mapping in segregating populations, have identified loci governing complex traits in many crops. Unoccupied aerial systems (UAS)‐based phenotyping has helped to reveal a more novel and dynamic relationship between time‐specific associated loci with complex traits previously unable to be evaluated. Over 1, 500 maize ( Zea mays L.) hybrid row plots containing 280 different replicated maize hybrids from the Genomes to Fields (G2F) project were evaluated agronomically and using UAS in 2017. Weekly UAS flights captured variation in plant heights during the growing season under three different management conditions each year: optimal planting with irrigation (G2FI), optimal dryland planting without irrigation (G2FD), and a stressed late planting (G2LA). Plant height of different flights were ranked based on importance for yield using a random forest (RF) algorithm. Plant heights captured by early flights in G2FI trials had higher importance (based on Gini scores) for predicting maize grain yield (GY) but also higher accuracies in genomic predictions which fluctuated for G2FD (−0.06∼0.73), G2FI (0.33∼0.76), and G2LA (0.26∼0.78) trials. A genome‐wide association analysis discovered 52 significant single nucleotide polymorphisms (SNPs), seven were found consistently in more than one flights or trial; 45 were flight or trial specific. Total cumulative marker effects for each chromosome's contributions to plant height also changed depending on flight. Using UAS phenotyping, this study showed that many candidate genes putatively play a role in the regulation of plant architecture even in relatively early stages of maize growth and development. Core Ideas: Temporal plant heights can predict maize hybrids via machine learning regression. Early plant heights were more important in yield prediction under optimal conditions. Segregating loci affecting maize plant height varied temporally and were determined by stress. Genomic predictions varied depending on plant heights belonging to different flight dates. Genomic prediction accuracy can be increased with using temporal phenotype data.