Use of the gas production technique to predict the relation between the intake level and indicators of in vitro ruminal fermentation

Abstract

The potentialities of the in vitro gas production technique were evaluated as a method to predict the relations between the simulated intake level and fermentation indicators. The in vitro simulated intake levels were 1.9, 2.1, 2.2 and 2.5% live weight. The diets were made with fixed amounts of maize meal and grass, while the amount of sugarcane meal varied (equivalent to 0; 6.0, 10.0 and 20.0 g kg-1 LW). At 24 h, the degradability of dry matter and nitrogen, the gas production and short chain fatty acids (total and individual), synthesis efficiency of microbial biomass and NH3 concentration were determined. An experimental randomized block design was used, the Pearson correlation indexes were determined among the studied variables and multiple linear regressions were performed to study the functional relation between the simulated intake level (response variable) and the variables of the degradability of dry matter, gas production, total short chain fatty acids and synthesis efficiency of microbial biomass (predictor variables). The Pearson correlation between the simulated intake level and fermentation indicators showed a high positive correlation for gas production, short chain fatty acids (acetic, butyric and totals) and synthesis efficiency of microbial biomass (P <0.05). There was high negative correlation with the degradability of dry matter and nitrogen (P <0.01). The multiple linear regression equations showed that R2 improved when the predictor variables increased from 2 to
3. The dry matter degradability showed the highest values of partial R2, although when excluding the gas production at 24 h explained 65.7% of the variability. The obtained results suggest that the voluntary intake levels that were simulated in vitro from the data of dry matter degradability and gas production can be predicted, while the inclusion of total short chain fatty acids and the synthesis efficiency of microbial biomass as predictor variables increased the determination coefficient of the obtained models.
Key words: degradability, fermentation, sugar cane, prediction, models