Breeding for Increased Water Use Efficiency in Corn (Maize) Using a Low-altitude Unmanned Aircraft System

Low-altitude aerial imagery collected by unmanned aircraft systems (UAS) at centimeter-level spatial resolution provides great potential to collect high throughput plant phenotyping (HTP) data and accelerate plant breeding. This study is focused on UAS-based HTP for breeding increased water use efficiency in corn in eastern Nebraska. The field trail is part of an effort by the Genomes to Fields consortium effort to grow and phenotype many of the same corn (maize) hybrids at approximately 40 locations across the United States and Canada in order to stimulate new research in crop modeling, the development of new plant phenotyping technologies and the identification of genetic loci that control the adaptation of specific corn (maize) lines to specific environments. It included approximately 250 maize hybrids primary generated using recently off patent material from major seed companies. These lines are the closest material to what farmers are growing today which can be legally used for research purposes and genotyped by the public sector. During the growing season, a hexacopter equipped with a multispectral and a RGB cameras was flown and used to image this 1-hectare field trial near Mead, NE. Sensor data from the UAS were correlated directly with grain yield, measured at the end of the growing season, and were also be used to quantify other traits of interest to breeders including flowering date, plant height, leaf orientation, canopy spectral, and stand count. The existing challenges of field data acquisition (to ensure data quality) and development of effective image processing algorithms (such as detecting corn tassels) will be discussed. The success of this study and others like it will speed up the process of phenotypic data collection, and provide more accurate and detailed trait data for plant biologists, plant breeders, and other agricultural scientists. Employing advanced UAS-based machine vision technologies in agricultural applications have the potential to increase the rate of genetic gain in plant breeding applications, as well as guide the optimization of management practices in precision agriculture.