Quantifying variation in estimated methane emission from ruminants using the SF6 tracer technique : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Animal Science at Massey University, Palmerston North, New Zealand

Abstract

With the signing of the Kyoto Protocol, New Zealand must reduce its national greenhouse gas emissions. As New Zealand has a large proportion of its national emissions as methane (~31%), and methane (CH4) has a short atmospheric lifetime, it provides a good target for mitigation strategies. The initial aim of this research was to identify high and low CH4-emitting cattle to assess factors that contribute to low CH4 production. Initial studies using the SF6 tracer technique to estimate CH4 production could not identify consistently high and low CH4 emitters. Research was therefore undertaken to confirm whether this was due to high variation in estimated CH4 yields, and to quantify the within- and between-animal variation in CH4 production when using the SF6 technique. This research showed considerable within- (coefficient of variation, CV = 7-10%) and between-animal (CV = 7-18%) variation in CH4 yield (g CH4/kg DMI) over time when using the SF6 technique. This is larger than the within- (CV = 3%) and between-animal (CV = 10%) variation reported for calorimetry. This led to the recommendation that the SF6 technique not be used in identifying animals for high or low CH4 yield. A power analysis was developed based on the measured variances for the SF6 technique. Results from this analysis provide researchers with important information on the number of animals and measurements per animal required when undertaking CH4 experiments. One of the sources of variation with the SF6 technique is the SF6 release from permeation tubes. Estimated CH4 yield increases by approximately 8.5% when going from a release rate of 3 mg SF6/day to a rate of 5 mg SF6/day. Further, an in vitro study indicated that SF6 release from permeation tubes is approximately 8% lower in rumen fluid than in air. While further research is required to confirm these results, they emphasise the need to allow time for the release rate to stabilise in the rumen for 4-5 days prior to undertaking measurements. It also led to the recommendation that release rates used in experiments should be within a narrow range, and balanced across experimental treatments.