Using Models to Predict Methane Reduction in Pasture- Fed Dairy Cows

Because New Zealand is a signatory to the Kyoto Protocol Framework Convention on Climate Change (1997), it is required to report national greenhouse gas emissions and maintain them at 1990 levels by 2008-2012. New Zealand has one of the highest per capita levels of the greenhouse gas methane [Woodward et al., 2001] and most of this arises from gut fermentation from ruminants. Not only does this ruminant-produced methane contribute to greenhouse gas emissions, but it is also wasteful in terms of energy loss for the animal. This paper investigates the ability of four methane models within MOLLY [Baldwin, 1995] to model the methane production from a trial conducted at Dexcel, a dairy research company based in Hamilton, New Zealand. In this trial, sixteen Friesian and Jersey dairy cows grazed on either perennial ryegrass (Lolium perenne) or sulla a condensed tannin containing forage legume (Hedysarum coronarium). MOLLY was run over 3 days, the period of methane production data collection, and each of the eight ryegrass-fed cows were modelled. (The sulla-fed cows were not modelled due to lack of information on the composition of sulla). MOLLY is one of the cow component models used in the Whole Farm Model (WFM) [Sherlock et al., 1997; Bright et al., 2000] which has been developed at Dexcel. The WFM simulates a pasture-based, rotationally grazed dairy farm and consists of a framework to which are attached four different components – climate, management, pasture and cow. Three of the methane models were empirical – based on regression equations, the remaining model being a mechanistic one. One of the empirical models and the mechanistic model were found to be the best predictors of methane production for the ryegrass-fed cows. The process of modelling methane production at the whole farm level will highlight areas where knowledge of methanogenesis is lacking, and direct research to the most promising feeding strategies to reduce methane emissions.