Difficulties encountered in modeling CH4 output by ruminants - Archive ouverte HAL Accéder directement au contenu
Poster De Conférence Année : 2012

Difficulties encountered in modeling CH4 output by ruminants

Résumé

Modeling accurately CH4 production by ruminants is a challenge for the global assessment of this greenhouse effect gas production. For this purpose, many experimental data have been used; they mainly came from calorimetry measurements performed on ruminants for decades. Most of the equations predicting CH4 production by ruminants proposed in the literature were based on animal characteristics (weight, level of performance...) or/and feed items (diet composition ...) and level of intake. They differed mainly in terms of accuracy and genericity because of differences in the databases used for the calculations. Some equations seemed to be accurate on the basis of their low residual variation; however they actually have a limited generic value because they were built on data sets limited to one or a very small number of sources and integrate very few contexts. In contrast, some less accurate equations have the advantage of being calculated on a wide variety of data contexts. It is a priori possible to improve the precision of predictive models of CH4 production by integrating parameters related to the valorization of the diet by the animal (i.e. digestibility of the diet organic matter or OMD). However, the uncertainty of this measure of digestibility needs to be integrated in order to assess the global accuracy of such models. Moreover, the relationship between OMD and CH4 production is not simple because methanogenesis is much more sensitive to digestive interactions than OMD. Other approaches attempted to go further in details by integrating criteria even closer in the process of methanogenesis, such as, in particular, models incorporating data on the state of fermentations in the rumen. Thus, it is possible to work with models that are more mechanistic, providing a better understanding for the shortcomings of more aggregated models. For example, models based on relationships between methanogenesis and volatile fatty acids profile in the rumen have the interest to be more mechanistic because based on well-known principles of stoichiometry. However, these models are even limited for predicting CH4 production for extreme diets causing significant deviations of fermentations in the rumen, resulting in a very low CH4 production compared to the predicted one. Similarly, this approach overestimates intense production of CH4 observed with fermentation rich in acetic and butyric acids. This bias was observed in vivo as in vitro. In conclusion, the debate on the choice of a model to foresee the CH4 output by ruminants remains quite open.
Fichier non déposé

Dates et versions

hal-01369356 , version 1 (20-09-2016)

Identifiants

  • HAL Id : hal-01369356 , version 1
  • PRODINRA : 369222

Citer

Daniel Sauvant, Sylvie Giger-Reverdin, Amélie Serment. Difficulties encountered in modeling CH4 output by ruminants. International Symposium on Emission of Gas and Dust from Livestock, Jun 2012, Saint-Malo, France. 130 p., 2012, International Symposium on Emission of Gas and Dust from Livestock. ⟨hal-01369356⟩
67 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More