Comparison of fitting of some mathematical models to describe the ruminal fermentation kinetics according to gas production technique for alfalfa hay

In this study, the mathematical models were used for evaluation of ruminal fermentation kinetic of alfalfa hay. These models included exponential (EXP), Michaelis-Menten (MIC), Mitscherling (MIT), Weibull (WEB), Korkmaz-Uckardes (KOR) and France (FRC). The in vitro gas production was carried out in 4 separate periods. Three syringes containing feed samples (3 replicates) were considered for each period and the volume of gas produced in each period at different incubation times (144 hours) was fitted for these models. Mean Square Error (MSE), coefficient determination (R2) and Mean Percentage Error (MPE) were used for models goodness of fit. Durbin-Watson and Shapiro-Wilk tests, Bayesian Information Criterion (BIC), Akaike’s Information Criterion (AIC) and Accuracy Factor (AF) were used for selection of the best model. The results showed that MSE in FRC (0.852) and MIC (0.917) models was lower than that of EXP (7.437) model (P