In vitro methods as predictors of voluntary intake and digestibility of hays fed to sheep

Eleven sun-cured hays were used to study the suitability of different in vitro methods to predict their voluntary dry matter (DM) intake (VDMI; g DM/kg liveweight) and in vivo DM digestibility (DMD; g/kg). The methods used were: (1) gas production at different incubation times when hays were incubated in vitro with buffered rumen fluid, (2) DM disappearance at different incubation times with cellulase (CEL), (3) release of sugars to the supernatant after incubation with cellulase, and (4) optical density of the supernatant at lambda = 280 nm (as an indicator of phenolic compounds release) after incubation with cellulase. All kinetic data were fitted to first-order kinetics models to estimate the rate of degradation and the potential degradability, and the average degradation rate and effective degradability (ED) were calculated. The most accurate prediction of VDMI was by using the average degradation rate of the CEL method in a single regression equation, which accounted for 0.78 of the variation in intake (residual s.d. = 1.40). When only data from grass-rich hays were considered, the inclusion of potential degradability and rate of degradation of the CEL method in a multiple regression equation accounted for 0.98 of the variation in intake (residual s.d. = 0.32). The ED of the CEL method explained 0.93 and 0.92 of the variation in the in vivo DMD for all hays and grass-rich hays, respectively (residual s.d. = 18.1 and 21.7, respectively). The predictive ability of the different methods is discussed and compared with that of the in situ technique that has been previously reported.