The performance of near-infrared (NIR) spectroscopy as a rapid technique for the estimation of chlorophyll and protein contents in alfalfa (Medicago sativa L.) was investigated. A fiber-optic probe was employed directly on a total of 198 fresh leaves to measure spectra between 1100 and 2200 nm. Partial least squares (PLS) regression models were developed with a calibration set of 120 samples spanning a concentration range of 5.20-158.5 for the chlorophyll content index (CCI), 0.39-4.60 mg g-1 (fresh weight) for the chlorophyll extracted with dimethylsulfoxide (DMSO), and 9.92-45.32% (dry matter) for protein content. The models obtained were validated with 78 independent samples. Standard errors of prediction of 12.49 were obtained for the CCI, 0.24 mg g-1 for DMSO-extracted chlorophyll, and 3.27% for the protein content. These results support the use of NIRS equipped with a fiber-optic probe to monitor and assess the composition and quality of forages in a nondestructive way.

This work was supported by Spanish Ministry of Education and Science Project AGL2005‐02839/AGR. Thanks are also due to J. C. Estevez for technical assistance. C. Petisco acknowledges financial support from the University teacher training predoctoral grant awarded by the Spanish Ministry of Education and Science, Spain.