Title: Mid-Infrared and Near Infrared Calibrations for Nutritional Parameters of Triticale (Triticosecale) and Pea (Pisum sativum) Authors
Submitted to: Journal of Agriculture and Food Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: March 24, 2009
Publication Date: May 20, 2009
Repository URL: http://doi:10.1021/jf803936x
Citation: Calderon, F.J., Reeves III, J.B., Vigil, M.F., Poss, D.J. 2009. Mid-Infrared and Near Infrared Calibrations for Nutritional Parameters of Triticale (Triticosecale) and Pea (Pisum sativum). Journal of Agriculture and Food Chemistry. 57:5136-5142. doi:10.1021/jf803936x. Interpretive Summary: This manuscript is significant because we show for the first time how the mid-infrared spectra can be used effectively to develop calibrations for fiber and nitrogen properties on forages, especially triticale. The paper is also valuable because we compare two spectral methods (ie. Near infrared vs. Mid-Infrared) in their performance. Our data shows that the mid-infrared region performs adequately compared to near-infrared. We have shown how a new technology can be used to quantify forage fibers and total N, which are important components of ruminant diets, and can indirectly affect human nutrition.
Technical Abstract: Data is lacking about the relative effectiveness of diffuse reflectance Fourier transformed mid-infrared (MidIR) versus near infrared (NIR) spectroscopy for calibration development of forage constituents. The objective of this study was to develop MidIR and NIR calibrations for acid detergent fiber (ADF), neutral detergent fiber (NDF) and total nitrogen in triticale, peas, and triticale/pea mixtures. Heterogeneous calibration-validation combinations were also tested for calibration quality. The forage samples were collected from forage plots grown following millet or wheat. Other factors included population density, forage mixtures, and nitrogen fertilizer rate. Our results show that Total N always achieved a better validation R2 than ADF and NDF, regardless of the sample set or spectral range. The ADF and NDF could not be predicted well with heterogeneous calibration/validation sets, with the exception of ADF predicted by the pea/triticale mixture in the MidIR. Using the whole sample set in the calibration resulted in better predictive calibration for the fiber analytes for both the MidIR and NIR. This study shows that MidIR compares well with NIR for the development of ADF, and total N calibrations in forages. The NIR and MidIR are both useful as quick methods for measuring total N and they show promise for measuring ADF and total N in forage samples, but performance with NDF was less satisfactory.