Precision and Accuracy of Methodologies for Estimating In Vitro Digestibility of Thinopyrum ponticum (Tall Wheatgrass) Hay and Haylage Fed to Beef Cattle1

Digestibility of feeds is a commonly used nutritive parameter and can be estimated through several techniques. The objective of the present study was to compare the precision and accuracy of different IVDMD techniques used to estimate in vivo DM digestibility (IVdig) of tall wheatgrass hay and haylage (IVdig: 42.2 to 53.3%). Forages were conserved after 1 of 3 regrowth periods (66, 96, and 162 d), generating a representative range of digestibilities, and were fed ad libitum to steers. The study analyzed goodness-of-fit for predictions of in vivo digestibility using in vitro apparent DM digestibility from Tilley and Terry (T&Tdig) and a DaisyII Incubator (DAISYdig), and in vitro ruminal DM degradability using the gas-production technique (GASdeg). In addition, 2 predictive equations from the literature were tested, Van Soest (VSdig) and Rohweder (Rodig), based on fiber content of forage. The in vitro techniques showed higher correlations with IVdig (R2 = 0.97, 0.94, and 0.93, respectively) than VSdig and Rodig (R2 = 0.64 and 0.35, respectively). Biases were observed in all techniques, with correction factors of 0.34 for T&Tdig, 0.74 for GASdeg, and 0.88 for DAISYdig. As a consequence, DAISYdig showed the greatest concordance (ρc = 0.85) compared with T&Tdig and GASdeg (ρc 0.34 and 0.72, respectively), the predictive equations showed the poorest fits (ρc 0.21 and 0.33 for VSdig and Rodig, respectively). Therefore, the recommended techniques are DAISYdig or GASdeg, depending on requirements. However, all in vitro techniques showed biases, highlighting certain limitations for conserved forages.