Development of a calibration for predict total oil content in corn seeds using near infrared spectroscopy in breeding program
2010
Reunreudee Kaewcheenchai(Kasetsart University, Bangkok (Thailand). Faculty of Agriculture. Department of Agronomy) | Sumaporn Kasemsumran(Kasetsart University, Bangkok (Thailand). Kasetsart Agricultural and Agro-Industrial Product Improvement Institute) | Vichien Keeratinijakal(Kasetsart University, Bangkok (Thailand). Faculty of Agriculture. Department of Agronomy) | Warunee Thanapase(Kasetsart University, Bangkok (Thailand). Kasetsart Agricultural and Agro-Industrial Product Improvement Institute)
Determination of total oil content in corn seeds using NIRS with the PLSR method was used as a criterion to evaluate corn population. Pre-calibration was developed from 102 corn samples. The NIR absorptions were collected in the region of 1200 to 2400 nm and the total oil content was determined by soxhlet extraction. The total oil content of 102 corn samples ranged from 2.1792-10.8233 percent (w/w). Subsequently, the pre-calibration model was employed to predict the total oil contents in corn seeds from 4 breeding program, in which Suwan04L-336 and Suwan04L-337 were used as base population. From the NIR prediction results of oil content, 20 corn samples of each breeding program were selected for recalculation. The pre-calibration model was recalculated by adding NIR spectral and oil content data of 20 corn samples for each program, in which the pre-calibration model was updated following 1st, 2nd and 3rd re-calibration model, respectively. A calibration model that yielded the best statistic performance will be used for prediction total oil contents of corn samples from breeding program. The results showed high possibility of the use of NIRS method for determination of total oil content in corn seeds.
显示更多 [+] 显示较少 [-]