Multiple in vitro methods have been developed to assess forage digestibility but little is known regarding agreement in disappearance among them. This study compared 3 different rumen in vitro apparent digestibility (AD) methods (Daisy^II [D], Batch Culture [BC] and the Ankom Gas Production System [G]) at 4 incubation times ([IT];12, 24, 36 and 48 h). Additionally, results obtained at 24 h were compared with those obtained from dual-flow continuous fermenters [CF]. Annual ryegrass (Lolium multiflorum; 33.8% NDF, 16.6% CP, 91.1% OM, 27.3% WSC) was clipped from an ungrazed pasture, dried (<60°C) and ground in a Wiley mill (1 mm). Three runs of each method were conducted using rumen fluid from a cannulated Holstein cow in mid-lactation fed a 34% corn silage, 6% grass hay and 60% grain mix diet. Ankom F57 acetone pre-rinsed bags containing 0.5 ± 0.01 g of sample were used for D, BC and G. Digestibility coefficients in CF were estimated in 3 periods (7-d adaptation, 3-d collection) ran simultaneously with the other methods. The same buffer was used for all 4 methods. Data were analyzed using Mixed procedure of SAS in a model including method, IT and period as a random factor, with IT as repeated measure. Means within each IT were compared by PDIFF function. Results are presented in Table 1. Results indicate that D predicts higher AD than G and BC at IT greater than 12 h. Digestibility estimated using CF was similar to the obtained with BC and G at 24 h. We conclude that different in vitro digestibility methods could yield different results. Table 1. Dry matter digestibility (%) data of annual ryegrass samples assessed by 4 in vitro methods at 4 incubation times

In vitro method	Incubation time (h)
12	24	36	48
DaisyII	58.35	70.92a	75.93a	79.85a
Batch culture	52.88	61.29b	68.14b	71.82b
Gas Production System	51.23	58.89b	64.01b	66.44b
Continuous fermenter		58.97b

^a,bMeans within each column with a different letter indicate significant differences (P < 0.05).