A model of enteric fermentation in dairy cows to estimate methane emission for the Dutch National Inventory Report using the IPCC Tier 3 approach

The protocol for the National Inventory of agricultural greenhouse gas emissions in The Netherlands includes a dynamic and mechanistic model of animal digestion and fermentation as an Intergovernmental Panel on Climate Change (IPCC) Tier 3 approach to estimate enteric CH4 emission by dairy cows. The model differs from an IPCC Tier 2 approach in that it predicts hydrogen sources (i.e., production of acetate and butyrate, microbial growth on amino acids as an N source) and sinks (i.e., production of propionate and the remainder of the volatile fatty acids (VFA), microbial growth on ammonia as an N source, saturation of unsaturated long chain fatty acids) in the rumen and large intestine, and elimination of excess hydrogen by methanogenesis. As a result, the model predicts CH4 emission by considering various dietary characteristics, including the types of carbohydrate, protein, fat, intrinsic degradation characteristics of feeds, as well as ruminal fractional passage rates, fluid volume and acidity, instead of assuming a fixed CH4 energy conversion factor in the Tier 2 approach. Annual statistics of diet and performance of the average dairy cow in The Netherlands from 1990 until 2008 indicate that dry matter intake and yield of fat and crude protein corrected milk (FPCM) per cow/year increased by 20 and 34% respectively. Based on annual data for diet and FPCM, the model predicted an increase in enteric CH4 emission from 111 (1990) to 128 (2008) kg/cow/year. As a result, CH4 emission per kg FPCM milk decreased by 13%. The predicted fraction of gross energy intake lost as CH4 energy gradually declined and was close to 0.06, which is the IPCC (1997) Tier 2 default value of 0.06 for dairy cows, but ~10% lower than the IPCC (2006) updated value of 0.065. The 15% uncertainty value for predicted CH4 emissions for a reference diet was lower than the 20% assumed under Tier 2. Our analysis indicated that uncertainty of model predictions of CH4 emission is determined mostly by errors in feed intake estimation, in the representation of the stoichiometry of production of VFA from fermented substrate, and in the acidity of rumen contents. Further uncertainty of predicted CH4 emission was due to errors in estimation of dietary composition of ingredients and in chemical compositions of dietary components. Results demonstrate that prediction of CH4 should not solely focus on representing effects of nutrition on overall digestion and apparent feed utilization by cows, but that additional attention is needed to address effects of nutrition on intra-ruminal fermentation conditions, and their effects on formation of VFA and the rumen hydrogen balance.