ESTIMATION OF BOTANICAL COMPOSITION OF ESOPHAGEAL EXTRUSA SAMPLES USING NEAR-INFRARED REFLECTANCE SPECTROSCOPY

The objective of this study was to evaluate the potential of near infr ared reflectance spectroscopy (NIRS) for estimating botanical composit ion of esophageal extrusa samples, Spectral data were collected on 361 samples from fistulated sheep and cattle grazing native tallgrass ran ge, Principal components analysis was used to identify a subset of 73 samples with spectral dissimilarity, These samples were microhistologi cally analyzed to determine botanical composition and were considered 'actual' for regression and calibration purposes, Thirty-six species ( 12 grasses, 22 forbs, and 2 sedges) were identified in the microhistol ogically analyzed samples, However, most accounted for less than 5% of the total diet, Additional pure calibration samples were obtained by feeding individual species to confined fistulated sheep, Initial regre ssion analyses and predictions were made on 13 major species or specie s groups, Satisfactory prediction equations could only be developed fo r big bluestem (Andropogon gerardii Vitman) (r(2) = 0.61), and the tot al grasses (r(2) = 0.79) and total forbs (r(2) = 0.79) groups, Additio n of spectra from pure samples into the calibration set was beneficial , In general, valid predictions could not be made for individual speci es that constituted less than 10% of the sample and/or had a low frequ ency of occurrence in the calibration samples, The NIRS method offered acceptable precision and accuracy in the prediction of major botanica l components and it would be practical and efficient because it reduce s the number of samples that would have to be microhistologically anal yzed.