|dc.description.abstract||An experiment was conducted to investigate the applicability of near infrared reflectance spectroscopy (NIR) to predict DMI and in-vivo NDF digestibility in horses. Nine mature, sedentary stock-type geldings, ranging from 10 to 20 yr of age were randomly assigned to one of three treatments within three simultaneous 3 x 3 Latin Squares to compare intakes and digestibilities of alfalfa (A), coastal Bermudagrass (C), and mixed grass hay (MGH). The 51-d study consisted of three 17-d periods. Each period consisted of a 4-d dietary adjustment period, 10-d feeding period, and 3-d total fecal collection period. Total fecal collections were conducted for 72 h, with feces collected and weighed every h. Representative hay and fecal samples were remitted to Dairy One Forage Lab (Ithaca, NY) for analysis.
There was a main effect of treatment (P = 0.02) on overall mean DMI. Horses consuming A had a greater mean DMI (P = 0.02) as compared to horses consuming MGH (7.31 vs. 5.41 kg/d). Overall mean DMI for horses consuming A and C (7.31 vs. 6.79 kg/d) were similar (P = 0.71). Dry matter intake was negatively correlated (r = -0.44) with NIR analysis of NDF with sulfite and ash correction (aNDFom). Dry matter intake was significantly (P = 0.02) related to hay aNDFom; however, the R2 value was 0.17, which indicated that 17% of the variation in DMI could be explained by the aNDFom content of hay. Dry matter intake was negatively correlated (r = -0.45) with NIR analysis of undigested NDF at 30 h (uNDF30). Dry matter intake was significantly (P = 0.02) related to hay uNDF30; however, the R2 value was 0.17, which indicated that 17% of the variation in DMI was explained by the uNDF30 content of hay.
There was a main effect of treatment on in-vivo NDF digestibility (P = 0.01). Correlations between in-vivo NDF digestibility and NIR hay analysis were not significant (P > 0.61) for aNDFom, uNDF30, or NDF digestibility at 30 h (NDFD30). Significant correlation coefficients between aNDFom content and in-vitro estimates were observed; 0.99 for uNDF30 and -0.68 for NDF digestibility at 30 h. Results from this study indicate that the use of NIR to estimate intake and in-vivo NDF digestibility in horses warrants further research. Perhaps the further collection of data will lead to development of NIR calibration equations that will be robust enough to predict NDF digestibility in horses.||