Estimating End Biomass in Sorghum Using Aerial and Ground-Based Phenotypes.

Above-ground plant biomass is considered the measure of yield for a crop like bioenergy or forage sorghum.  However, waiting until the end of the year to harvest destructively and process biomass samples with a biomass harvester does not allow selections in a breeding program before flowering and crossing, thereby delaying the breeding cycle.  Early-season destructive sampling is one option that allows selection before the sorghum flowers; however, removing plants decreases the accuracy of the biomass estimate as well as the number of plants available for breeding (or increases the total experiment size needed to account for these losses), and removes the option of later phenotyping on the removed material.  Less direct measurements of plant biomass can be calculated non-destructively, such as the approximation of stem volume as a cone, derived from plant height and diameter.  Presented here across multiple times in the growing season, this approach can provide a fairly accurate prediction. Taken by hand, these measurements can be time-consuming and labor-intensive, but advances in remote sensing technologies such as remotely sensed plant height (derived from RGB point clouds), and hyperspectral features (including Extended Multi-Attribute Profiles, or EMAPs) help solve these issues.  Overall, this study highlights the usefulness of a combination of platforms in predicting end-of-season biomass in forage or bioenergy crops.