Prediction of methane production in beef cattle within a mechanistic digestion model
| Autor: | J. L. Ellis, J. Dijkstra, E. Kebreab, S. Archibeque, J. France, A. Bannink |
|---|---|
| Rok vydání: | 2011 |
| Předmět: | |
| Zdroj: | Modelling nutrient digestion and utilisation in farm animals ISBN: 9789086867127 Modelling Nutrient Digestion and Utilisation in Farm Animals Modelling Nutrient Digestion and Utilisation in Farm Animals. Wageningen: Wageningen Academic Publishers |
| DOI: | 10.3920/978-90-8686-712-7_20 |
| Popis: | Methane is produced by ruminants as the result of microbial digestion, it represents an energy loss to the animal, and it is also a potent greenhouse gas. Mechanistic modelling can lend insight into dietary strategies aimed at reducing methane emissions from cattle, but require proper representation of aspects of underlying rumen fermentation and digestion. Proper prediction of the production of volatile fatty acids (VFA) is central to accurate methane prediction. This study evaluated the newly updated and expanded model of VFA dynamics developed by Bannink et al. (2008), in comparison to a previous model version (Bannink et al., 2006), within a rumen model (Dijkstra et al., 1992; modified by Mills et al., 2001) for its methane prediction ability in beef cattle fed high-grain diets. In an evaluation of the rumen model performed by Kebreab et al. (2008) using the Bannink et al. (2006) VFA stoichiometry, the model performed well on dairy cow data, but poorly on beef cattle data in predicting methane emission. Several modifications were therefore made to the model to adapt it for typical high-grain beef cattle diets and then evaluated for its accuracy to predict methane emissions from feedlot cattle. Passage rate of protozoa was increased proportionally with the grain percent of the diet, and the protozoal proportion of the amylolytic microbial pool was reduced accordingly. This allowed a small cellulolytic microbial pool to remain in the rumen, where it previously went extinct. Also, stoichiometry of rumen VFA production was adjusted for the use of monensin in the observed data. Preliminary results showed that while the representation of some central aspects of rumen fermentation probably improved, the model over-predicted methane production with a root mean square prediction error value of 2.86 MJ/d, with 56% of that error coming from bias and 44% from random sources. Concordance correlation coefficient value was 0.194. Results indicate that other areas of the model require improvement for predicting methane production accurately in high grain feedlot diets, such as hind gut fibre digestibility. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|