[en] Alfalfa (Medicago sativa L.) is a high quality forage which has been used worldwide. The superiority of alfalfa lies in its high yield, high protein content and high digestibility. The aim of this study was to develop a simple, fast and non-destructive method, named Near Infrared Spectroscopy (NIRS) to determinate alfalfa quality. To realize this study, alfalfa samples were obtained from Mănăştur Experimental Station – Farm Cojocna in 2008–2009, in one experiment carried out using randomization blocks design with two experimental factors (mineral fertilization and period of harvest). Alfalfa quality was first determinated on 48 samples by classical analyses: crude protein CP (AOAC, 1990), crude fiber CF (Fiber Cap, FOSS, DK) and in vitro organic matter digestibility OMDrt (DeBoever, 1986). Then the samples were scanned by NIRS. Calibration models were performed by PerkinElmer Spectrum Quant + 4.21 program (USA) on the 48 samples determinated by classical analysis. The results showed fully confirmed by acceptable coefficients of determination and standard error of cross-validation (R2=0.96 for CP, 0.94 for CF, 0.98 for OMDrt and SECV=0.77 for CP, 1.35 for CF, 1.13 for OMDrt). Successful results for prediction of other 176 alfalfa samples were then obtained using these calibration models: SEP=0.869 for CP, 1.058 for CF, 1.058 for OMDrt). The highest CP and OMDrt values of alfalfa were obtained in bud stage (22.0% and 66.0% respectively). While for CF, the highest content was registered in the seed formation stage (46.0%). The NIRS technique offers us the possibility to determinate rapidly and easily Romanian alfalfa important parameters, but the system could also be used for the determination of other constituents.
Research Center/Unit :
Animal Science Unit
Disciplines :
Animal production & animal husbandry
Author, co-author :
Dale, Laura ; Université de Liège - ULiège > Doct. sc. agro. & ingé. biol.
Thewis, André ; Université de Liège - ULiège > Sciences agronomiques > Zootechnie