Exploring the Genetic Variation in Maize Height Utilizing Unmanned Aerial Systems (UAS).

Technologies in agriculture are increasingly focused on high throughput phenotyping via advances in automation and data science. Implementation of these advances allows plant breeders to develop and utilize new tools in order to better characterize their crops, making informed decisions while maintaining low inputs (labor, cost, time, etc..). In recent decades plant breeders have seen an influx of high-quality genomic data while phenotypic data collection has generally remained static, this has caused a phenotyping bottleneck in plant quantitative genetics and breeding. Unmanned aerial systems (UAS) that include light-weight airborne sensors present new opportunities to rapidly phenotype breeding populations, aiding in dissection of genetics underlying quantitative traits. Three bi-parental inbred line (IBL) maize (Zea mays L.) populations were imaged utilizing a DJI Phantom 4 quadcopter at weekly intervals throughout the growing season in a Central Texas environment (College Station, TX), switching to bi-weekly flights during the flowering period. Orthomosaic images and three-dimensional point clouds were constructed using structure from motion (SfM) algorithms in Pix4D software and height estimates were extracted using open source LiDAR software. The interest of this research is to assess the ability of UAS imaging to accurately capture inbred height estimates and genetic variation within a segregating maize trial.