Ranking of in vitro fermentability of 20 feedstuffs with an automated gas production technique: Results of a ring test

A ring test was completed to determine within laboratory repeatability, and reproducibility among laboratories, of a gas production technique for measuring fermentation kinetics of feeds in rumen fluid. Eight laboratories in different European countries received 20 feed samples. All used the same gas production equipment and method. Gas production profiles were fitted to a multiphasic model and the most important parameters were statistically compared. Although the within laboratory repeatability of, and levels of, calculated parameters differed among laboratories, in most cases there was a high correlation between pairs of laboratories (r = 0.93-1.00), although in some cases the correlation was only moderate (r = 0.73-0.87). Within laboratory gas production profiles were influenced by variation in microbiological activity of the rumen fluid and air pressure changes at the end of the run. Among laboratory differences were observed in calibration factors, type of donor animal and diets fed. Variation sources can be minimized using within run standards.