Premium
Ash in Forage, Esophageal, and Fecal Samples Analyzed Using Near‐Infrared Reflectance Spectroscopy
Author(s) -
Windham W. R.,
Hill N. S.,
Stuedemann J. A.
Publication year - 1991
Publication title -
crop science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.76
H-Index - 147
eISSN - 1435-0653
pISSN - 0011-183X
DOI - 10.2135/cropsci1991.0011183x003100050053x
Subject(s) - pasture , calibration , wavelength , forage , grazing , near infrared reflectance spectroscopy , ashing , zoology , analytical chemistry (journal) , biology , soil science , chemistry , environmental science , agronomy , mathematics , environmental chemistry , near infrared spectroscopy , physics , optics , statistics , quantum mechanics , neuroscience
In study of the plant‐animal interface, knowledge of the organic fraction of pasture, esophageal, and fecal samples can help predict energy intake and digestion. Ashing of the large sample volumes necessary to conduct interface studies, however, is cumbersome and impractical. This study was initiated to examine the use of near‐infrared reflectance spectroscopy (NIRS) as a means of predicting ash concentration in pasture, esophageal, and fecal samples. Pasture mixtures of alfalfa ( Medicago sativa L.) and tall fescue ( Festuca arundinacea Schreb.) were defoliated by grazing animals. Pasture, esophageal, and fecal samples were collected and scanned by NIRS. A calibration subset was selected, samples were ashed, and the data used for equation development. The optimum equation was developed using log 1/reflectance ( R ) spectral data and produced standard errors of calibration (SEC) and validation (SEP) of 14.4 and 14.2 g kg −1 , respectively, regardless of whether ash or organic matter data were used. Equations selected during calibration contained three wavelength pairs within 40 nm of one another. New calibration equations were generated by forcing in one wavelength in the neighboring pair ± 20 nm from the selected band. Validation error of calibration equations with forced wavelengths was not different from that of the original calibration equation. It was concluded that wavelength pairs were necessary to calculate slope differences of spectra from samples varying in ash concentration. Recalibration of the instrument with a subset of pasture and esophageal samples unproved SEC and SEP to 9.5 and 9.8 g kg −1 , respectively; recalibration with a subset of fecal samples improved SEC and SEP to 10.6 and 12.7 g kg −1 , respectively.