Skip to main content
ARS Home » Plains Area » Mandan, North Dakota » Northern Great Plains Research Laboratory » Research » Publications at this Location » Publication #190996
Title: FAECAL NEAR INFRARED REFLECTANCE SPECTROSCOPY TO PREDICT DIET QUALITY FOR SHEEP

Author
item LI, H. - CHINA AG UNIV BEIJING, CH
item TOLLESON, D. - TEXAS A&M UNIVERSITY
item STUTH, J. - TEXAS A&M UNIVERS9TY
item BAI, K. - CHINA ACADEMY AG SCI,BEIJ
item MO, F. - CHINA AG UNIV BEIJING,CHI
item Kronberg, Scott

Submitted to: Small Ruminant Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/15/2005
Publication Date: 3/1/2007
Citation: Li, H., Tolleson, D., Stuth, J., Bai, K., Mo, F., Kronberg, S.L. 2007. Faecal near infrared reflectance spectroscopy to predict diet quality for sheep. Small Ruminant Res. 68:263-268.

Interpretive Summary: Faecal near infrared reflectance spectroscopy (NIRS) is a non-invasive method to determine the diet quality of herbivores, but the viability of this technique for sheep (Ovis aires) has not been reported in the US. A prediction equation using tradition laboratory-derived forage quality data and faecal near infrared spectra was developed to determine if faecal NIRS can predict diet quality of forage-fed sheep. In 2002 (n = 15) and 2003 (n = 20) mature ewes (55 ± 2.4 kg) were fed individual diets for 7 days. Diets ranged from 4.3 to 23.5 % crude protein (CP) and 52.4 to 75.8 % digestible organic matter (DOM) and were composed of various grass, forb and browse components. Daily intake was recorded. Faecal samples were collected on days 6 and 7. CP was determined by micro-Kjeldahl and DOM by an in vivo corrected in sacco technique. Statistical techniques were used to develop the predictive equations. Equation validation was accomplished by cross validation, and validation results indicate acceptable predictive ability. To determine the effect of individual animal variation on faecal NIRS predictions, five ewes were fed different forages in two, 7-day trials. Predicted percent CP and DOM from both trials indicate minimal effect on NIR predicted diet quality due to individual animal variation. Predicting diet quality of forage-fed sheep can be done using faecal NIRS.

Technical Abstract: Faecal near infrared reflectance spectroscopy (NIRS) is a non-invasive method to determine diet quality in herbivores, but has not been reported for sheep (Ovis aries) in the US. A diet reference chemistry:faecal near infrared spectra calibration (n = 78) was developed to determine if faecal NIRS can predict diet quality of forage-fed sheep. In 2002 (n = 15) and 2003 (n = 20) mature ewes (55 ± 2.4 kg) were fed individual diets for 7 days. Diets ranged from 4.3 to 23.5 % crude protein (CP) and 52.4 to 75.8 % digestible organic matter (DOM) and were composed of various grass, forb and browse components. Daily intake was recorded. Faecal samples were collected on days 6 and 7. CP was determined by micro-Kjeldahl and DOM by an in vivo corrected in sacco technique. Partial least squares (PLS) and stepwise regression (SWR) techniques were used to develop predictive equations. Calibration results for percent dietary CP were: SWR, R2 = 0.93, SE calibration (SEC) = 1.27 and PLS, R2 = 0.95, SEC = 1.08. Calibration results for DOM were: SWR, R2 = 0.78, SEC = 1.58 and PLS, R2 = 0.80, SEC = 1.51. Equation validation was accomplished by cross validation, predicting an independent validation set, and by predicting day 7 samples with a day 6 derived equation within this study. Validation results indicate acceptable predictive ability. To determine the effect of individual animal variation on faecal NIRS predictions, five ewes were fed different forages in two, 7-day trials. Predicted percent CP and DOM from both trials indicate minimal effect on NIR predicted diet quality due to individual animal variation. Diet quality of forage-fed sheep can be accomplished by faecal NIRS.