NIR calibration models to estimate the bioethanol production potential in fresh grains of wheat, rye and triticale

Ethanol yield is among other agronomic measures one of the most important traits in breeding and processing of cereals for ethanol production. However, the laboratory-analysis of the ethanol yield is very laborious and therefore the application of near infrared spectroscopy (NIRS) was tested for the purpose of the fast screening of this measure.

The spectrometer NIRS5000 (FOSS, ground samples) and PSS 1720 (Polytec, fresh samples) were used and the calibrations were developed using partial least squares regression.

Calibration experiments have been carried out using samples from several field trials of the German Federal Plant Variety Office and breeding material from plant breeders. The bioethanol production potential for whole grains of wheat, rye and triticale can be estimated with an error of about 0.61 L ethanol/100 kg dry matter (SEP) with a coefficient of determination of 0.81.

These calibration models already allow the rough estimation of the bioethanol production potential within breeding programs. The possibility to further improve the developed calibration models is at time restricted by the low variability of the trait bioethanol production potential in existing cereal varieties. So far breeding programs do not focus on ethanol production. However, with this NIR based method the ethanol production potential can be estimated more precisely than with any existing batch tests.