A Field-based High-throughput Phenotyping Platform to Discover the Genetic Architecture of Sorghum Biomass Yield Components Over Time

High-throughput phenotyping technologies have emerged as a consequence of the need to obtain data at large scale, to increase accuracy and repeatability, and to phenotype plants over time for complex traits that could not be characterized by hand. The first succesful attempts to create high-throughput phenotyping systems were reported for laboratory or greenhouse settings or for field crops with small plant architecture. We have created a novel field-based self-propelled platform equiped with high resolution cameras that was specifically designed for a high biomass crop such as sorghum. This technology has been used to collect digital images during the entire season for 700 diverse sorghum lines in multiple environments. Data could be collected for a single location in a few hours and images were subsequently processed with appropriate algorithms to obtain plant architecture parameters such as plant height, stem diameter, panicle size and leaf angle. The accuracy of image-based algorithmically-derived data was demonstrated when compared with ground-truth measurements. These validated phenotypes will provide novel information about the dynamic changes in biomass production, growth rate and responses to variable environmental conditions during the growing season. 

The phenotypic data generated in this project will be used to discover genes/SNPs associated with variation in plant architecture traits over time, could be further utilized for the genetic improvement of sorghum and represents a significant contribution to the emerging field of predictive phenomics.