A preliminary evaluation of the use of mid infrared spectroscopy to develop calibration equations for determining faecal composition, intake and digestibility in sheep

• Published in: Animal feed science and technology Vol. 221; pp. 44 - 53
• Main Authors: Lyons, Gary, Sharma, Shekhar, Aubry, Aurelie, Carmichael, Eugene, Annett, Ronald
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-11-2016
• Subjects: Digestibility; Faecal analysis; Feed; FTIR; Intake; Sheep; Lignin (sa); F.NIRS; NDFom; DM; N; SD; ADFom; FTIR; PLS; Faecal analysis; DMD; Intake; R2c/R2cv; GE; Feed; Digestibility; DMI; SECV; GED; GEI; AshI; PCA; RPD value; SEC; Sheep; NI
• ISSN: 0377-8401; 1873-2216
• DOI: 10.1016/j.anifeedsci.2016.08.014

•Sheep intake and digestibility predicted by Fourier Transform Infrared spectroscopy.•Good prediction models developed for a number of parameters.•Prediction models for some parameters performed poorly.•Further development of all models required to improve performance.•The proposed technique would be beneficial to animal scientists and industry.
The aim of this study was to investigate the application of Fourier Transform Infrared (FTIR) spectroscopy to provide calibration equations for assessing faecal composition, intake and digestibility in sheep. Four grass-based feed types and corresponding faecal samples were collected from two separate sheep digestibility experiments. The feed (n=41) and faecal samples (n=132) were analysed for chemical composition, digestibility, were scanned using FTIR spectroscopy and intakes were recorded. Data were analysed using principle component analysis and partial least squares (PLS) regression techniques for calibration equation development. Calibration accuracy and performance were assessed by calculating regression coefficient of cross validation (R2cv) and the ratio of performance to deviation (RPD value) respectively. Results indicated that the composition of feed and faecal samples along with associated digestibilities and intakes were significantly different for a number of parameters studied. Some of the PLS regression equations generated were comparable to those reported in the published literature, and based on calibration statistics and performance those developed for faecal ash and neutral detergent fibre were good and could be used for quantification (R2cv=0.9–0.8, RPD=3.0–2.5); those for acid detergent fibre, lignin, feed ash intake and dry matter digestibility (DMD) were adequate for qualitative analysis and screening (R2cv=0.8–0.7, RPD=2.3–2.0); and the calibration equations for dry matter, nitrogen, dry matter intake, gross energy, gross energy intake and gross energy digestibility were poor and unsuitable for prediction (R2cv≤0.7, RPD≤1.8). Further development of calibration equations and in particular the prediction of DMD, could be of benefit to animal scientists and the livestock industry.