Deriving radiation use efficiency from hyperspectral sensing for enhanced sorghum production

Abstract

The efficiency with which sorghum crops use intercepted radiation (RUE) affects their maximum productivity. RUE in turn, is affected by crop canopy characteristics, as well as leaf biochemical parameters determining maximum leaf photosynthetic rates. These traits are notoriously slow to measure and have previously not been amenable to selection in breeding programs.

We used hyperspectral sensors attached to a tractor to estimate leaf and canopy traits associated with RUE in large sorghum field trials. Using Partial Least Square Regression we developed models to predict leaf biochemical photosynthetic parameters known to limit C4 photosynthesis, such as maximum Rubisco and PEP carboxylation rates (Vcmax and Vpmax, respectively) and maximum electron transport rate (Jmax) from the canopy hyperspectral reflectance. We also used various vegetation indices to predict canopy characteristics such as light interception and biomass which allowed us to calculate canopy radiation use efficiency for each trial plot.

This is the first time that parameters related to photosynthetic efficiency in sorghum have been estimated for so many different genotypes, showing that significant and heritable variation exists. The predicted values were also used for Genome Wide Association studies to identify the genomic regions associated with canopy RUE and hence improved productivity. Sorghum being such a great genetic model, this may also help elucidate the genetic mechanisms underpinning photosynthetic efficiency and RUE in other cereals.