Open Access

Estimating plant biomass under field conditions is a labor intensive and expensive activity that limits the capability to screen for this important trait in large wheat germplasm populations. Remote sensing represents a potential alternative to derive biomass and other relevant traits from spectral reflectance and thermal measurements. Using an unmanned aerial vehicle, we explored the capacity of aerial measurements of NDVI and canopy temperature (CT) to predict genetic variation for biomass in two spring wheat populations under yield potential conditions: the 2nd Wheat Yield Collaboration Yield Trial (2 WYCYT; 50 genotypes) and the High Biomass Association Panel (HiBAP; 150 genotypes). In each population the biomass was sampled at three different phenological stages: 40 days after emergence (E+40), 7 days after anthesis (A+7) and at physiological maturity (PM). Aerial data were collected six times during the growing season, covering the vegetative and grain filling stages of spring wheat. A correlation analysis was performed comparing aerial NDVI data with biomass estimated at the three phenological stages. A positive correlation was observed between NDVI and biomass during the vegetative period. No significant correlation was found at A+7 in both populations. At PM, only 2WYCYT showed a positive, but weak, correlation with final biomass. The relationship between cumulative CT and biomass was studied. In most cases, CT correlated negatively with biomass at a particular phenological stage. The biomass accumulated at A+7 and PM showed a strong negative relationship with CT measurements taken during grain filling. In general, no relationship was observed with CT measurements during the vegetative stage.

Text in English