51 Estimation of in vitro digestibility measures within equine fecal inoculum using feedstuff fiber composition

Abstract

Previous studies support the use of equine fecal microbes to simulate hindgut fermentation to calculate in vitro true digestibility (IVTD) of feedstuffs. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) composition have been shown to predict IVTD, in vitro NDF digestibility (IVNDFD), and in vitro ADF digestibility (IVADFD). However, little work has been done to predict these digestibility measures within horses. Thus, the study objective was to determine correlation and estimation of in vitro digestibility measures using equine fecal inoculum from feedstuff NDF and ADF composition. Four mature stock-type geldings (8.5 ± 3.9yr; 558.4 ± 31.0kg; 6.0 ± 0.4 BCS) were used in a duplicated 4 × 6 Factorial Design. Horses were acclimated to a basal diet of 1.8% BW as-fed of Coastal Bermudagrass hay (Cynodon dactylon)and 0.2% BW as-fed of Bluebonnet® Equilene® Pellets for 21d. Feces were collected on d22 via rectal-grab for use as incubation inoculum. Dried, ground feed samples (0.5g: alfalfa hay, Coastal Bermudagrass hay, soybean meal, rice bran, hempseed meal, Bluebonnet® Equilene® Pellets), sealed in ANKOM filter bags, were placed in duplicate within 2 DaisyII incubators for 48hr. After incubation, fiber bags were removed and dried, then analyzed for NDF composition to determine IVTD. Split samples were analyzed to determine pre-incubation NDF composition of feedstuffs. Correlations and linearregressions were performed within R Statistical Program©. Both NDF (r = −0.936) and ADF (r = −0.897) were highly correlated to IVTD. However, ADF had a greater correlation to both IVNDFD (r = −0.807) and IVADFD (r = −0.848) than NDF (r = −0.684; r = −0.720). A linear regression between NDF and IVTD produced the model: IVTD (g/kg DM) = 1052.30 – 0.932 * NDF (g/kg DM); R2 = 0.930; P < 0.001. A lower goodness-of-fit was found between ADF and IVTD: IVTD (g/kg DM) = 1033.25 – 1.528 * ADF (g/kg DM); R2 = 0.866; P < 0.001. An additive model of IVTD versus NDF+ADF mildly increased goodness-of-fit (R2 = 0.953). Sample ADF strongly predicted IVADFD (IVADFD (g/kg ADF) = 881.91 – 1.595 * ADF (g/kg DM); R2 = 0.835; P < 0.001) but only moderately predicted IVNDFD (IVNDFD (g/kg NDF) = 855.15 – 1.518 * ADF (g/kg DM); R2 = 0.749; P < 0.001). These results indicate the effectiveness of using feedstuff NDF to predict IVTD when using equine fecal inoculum. Feedstuff IVNDFD and IVADFD are predicted with less precision, likely due to the lack of foregut digestion of feedstuffs that would occur in vivo. The use of in vitro total-tract digestion modeling in horses may estimate equine hindgut fiber digestion with greater precision.