Measuring and modelling in-vitro gas production kinetics to evaluate ruminal fermentation of feedstuffs.

In this thesis, the possibilities of kinetic gas production measurements for the evaluation of ruminant feedstuffs have been examined. Present in-vitro methods were mostly end- point methods. There was a need for a kinetic in-vitro method that described ruminal fermentation, due to new techniques in fabricating additives and new developments in plant breeding.When a feedstuff is incubated with CO _{<font size="-2">2</font>} /HCO<font size="-2">₃ -</font>buffered rumen fluid, the rumen microorganisms ferment the carbohydrates from the feed to volatile fatty acids (VFA) and gas. The VFA also release CO _{<font size="-2">2</font>} from the buffer. When followed over time, this direct and indirect gas production result in sigmoidal curves. The ratios between the VFA produced (acetic, propionic and butyric acid) depend on the substrate. When rapidly fermentable carbohydrates are available, the relative amounts of propionic acid increased, whereas with slow fermentable carbohydrates, the relative amount of propionic acid decreased at the favour of the acetic and butyric acid. This affects the gas production, because at the production of propionic acid, no gas is produced as a fermentation end product, in contrast to acetic and butyric acid. In other words, the amount of gas produced depends on the fermentation pattern (Chapter 2).The gas production methods available all required frequent periodical readings. For a frequent use, it was desirable to have an automated method, which was realised through a liquid displacement system with automated registration (Chapter 3). Later this method was replaced by a more sophisticated one, using pressure transducers (Chapter 7).For comparison of the gas production curves, it was necessary to fit the data to a mathematical model. Several known models were tested, but they did not fit the curves satisfactorily. The fit was especially poor in those cases when rapid early gas production was observed together with a low final gas production rate. Therefore a new model was developed: the "Modified Gompertz Model' (Chapter 4). This model divided gas production in gas produced upon fermentation of rapidly and slowly fermentable material. From this model the following parameters were derived to describe the curves: length of lag phase, total gas production, maximum gas production rate, time at which gas production rate is maximal, and time at which 95% of the gas is produced.The effects of addition of cell wall degrading enzymes to grass silage was evaluated using the automated method and the model. Addition of those enzymes that resulted in a significant degradation of cell walls in the silage, resulted in a shortened lag phase and a lower maximum gas production rate, indicating that the enzymes degraded cell wall material to more rapidly fermentable components, leaving a more slowly fermentable fraction. The total amount of gas remained unchanged (Chapter 6).Another application was the evaluation of the effects of phenolic acids (p-coumaric and ferulic acid), added externally or present in different concentrations in the feed (stems of different maize inbred lines). When added externally, p-coumaric acid inhibited gas production and dry matter disappearance. Ferulic acid was not inhibitory. Maize stems with higher phenolic acid contents showed lower gas production rates and lower total gas production. A causal relationship could not be derived due to interaction with cell wall content (Chapter 7). The present method to determine ruminal degradation kinetics is the nylon bag method, where feed is incubated in the rumen in nylon bags and disappearance of feed components is followed gravimetrically over time. For different maize products a significant relationship was found between the total gasproduction and the unfermentable fraction. The gas production rate and rate of disappearance of organic matter did not agree. This might be caused by the fact that the nylon bag method distinguishes a soluble, an undegradable and a potentially degradable fraction, where degradation kinetics are only determined of the latter two fractions. The gas production method takes the whole sample into account. Future research should reveal the fysiological meaning of the gas production parameters. So far this method can supply a ranking of different feedstufs and/or additives