A new near-infrared reflectance spectroscopy method for high-throughput analysis of oleic acid and linolenic acid content of single seeds in oilseed rape (Brassica napus L.).

The development of oilseed rape cultivars with a high content of oleic acid (18:1) and a low content of linolenic acid (18:3) in the seed oil is an important breeding aim. Oil of this quality is increasingly being sought by the food and the oleochemical industry. Since the oil quality is determined by the genotype of the seed, a selection can be performed among single seeds of segregating populations. For this purpose a high-throughput Near-Infrared Reflectance Spectroscopy (NIRS) method using an automated sample presentation unit for single seeds of oilseed rape and a spectrometer equipped with a photodiode array detector was developed. Single-seed analyses have been accomplished with a throughput of up to 800 seeds per hour. Seeds from segregating populations of different origin were analyzed by NIRS and gas chromatography. Calibration equations were developed and validated applying the Modified Partial Least Square regression (MPLS) and LOCAL procedure. In three independent validations, standard errors of prediction corrected for bias between 2.7% and 3.7% for oleic acid and 1.2% and 1.8% for linolenic acid were determined using MPLS. Similar results were obtained applying the LOCAL procedure. The results show that the new high-throughput method can be applied to predict the oleic acid and linolenic acid content of single seeds of oilseed rape.