ABSTRACT
Summary Objectives of this paper are to review the status of near-infrared reflectance spectroscopy (NIRS) in forage-quality evaluation and to indicate the approach being used in a national research effort to further test, validate, and develop the technology for use in forage breeding, production, and utilization programs. Use of NIRS assay of forage quality was first reported 5 years ago. Research has now been initiated at many locations.
529A national research project, involving personnel at six widely dispersed locations in the U.S., was established in 1978. Computerized, high-precision near-infrared reflectance spectrophotometers are employed. A set of 30 forage samples provided experimental materials to ascertain magnitudes of errors involved in chemical, in-vitro digestibility, and NIRS analyses at the six laboratories. The six instruments showed different average spectral curves, but normalization of the data resulted in curves essentially alike except for the bands affected by water, an effect caused by differences in moisture content of the samples at the different sites. Our studies indicate that the magnitude of the errors associated with NIRS analysis compares favorably with those found in routine chemical and in-vitro procedures. Examples of research findings involving different plant species, harvests, growth environments, and drying procedures are presented. The results show that for maximum accuracy the calibration samples should be representative of the forages for which quality will be predicted by NIRS.
