Plant and soil influences on estimating biomass of wheat in plant breeding plots using field spectral radiometers

Abstract

Field spectral radiometers were used to estimate the biomass of wheat at early growth stages, as wheat breeders require a rapid, non-destructive technique to rank wheat genotypes for early vigour. Under experimental conditions, good relationships were obtained between reflectance and biomass prior to the wheat crop achieving a green area index of 1.5. When used above different soil types, good results were achieved on very uniform dark and light soils under experimental conditions, but greater differentiation between plots differing in biomass was achieved on darker soils. Similarly, under operational conditions in wheat breeders' plots, the best results were achieved against a dark soil background. Structural differences between plants also influenced solar radiation reflectance. At the Merredin site with the dark soil background, where the best correlation between reflectance and biomass was achieved, the relationship was much stronger for the more uniform genotypes at the second stage of selection than for the more heterogeneous genotypes at the first stage of selection. On these plots, the vegetation spectral indices NDVI (normalised difference vegetation index) and TSAVI (transformed soil-adjusted vegetation index) had a coefficient of determination 90-95% as good as the best regression using two wavebands. To optimise the field spectroradiometry technique for estimating early biomass, it should be applied at a weed-free site, with a uniform dark soil background and on material that is relatively homogenous in structure. We conclude that, unless these precautions are taken, the technique will have limited utility in breeding programs.