Near infrared spectroscopy of fecal samples (f.NIRS) has been shown to effectively predict botanical composition of diets of sheep and goats. However, because of the high absorbance in water bands and the effect of particle size feces have been dried and ground before collecting spectra. In contrast, Raman spectroscopy is an inelastic spectroscopy technique that is not affected by water and does not require sample preparation. The objective of this study was to compare f.NIRS to Raman for predicting the percentage of mesquite in cow diets and the discrimination of base rations by the two techniques. Fecal material was obtained from a feeding trial where known amounts of mesquite were introduced intraruminally into six ruminally fistulated cows with an average weight of 568 kg. Two animals were randomly assigned to one of three base rations, namely: timothy hay, Sudan hay or a 50:50 combination of Bermudagrass hay and beardless wheat hay. Fecal material was collected the last 2 days of four 7 day feeding periods where mesquite was fed at 0, 1, 3, and 5% of the base diet. NIR spectra of fecal material provided useful calibrations for predicting percentage mesquite in diets, but Raman spectra did not. Validation of f.NIR calibration using base diets containing Sudan or Timothy hays to predict percent mesquite in feces from base diets consisting of Bermuda and wheat hays had r2,slope and RMSE of 0.88,1.03 and 0.01, respectively. Cannonical discriminant analysis of wavelengths selected using a stepwise procedure showed that Raman and NIR spectra could classify the base hay ration for all the observations with 100% accuracy. The greater squared Mahalanobis distances between different base hay rations for the Raman spectra compared to the NIR spectra indicates that Raman spectra discriminated among the rations more efficiently than the NIR spectra.
Oral presentation and poster titles, abstracts, and authors from the Society for Range Management (SRM) Annual Meetings and Tradeshows, from 2013 forward.