Technical note: Can the sulfur hexafluoride tracer gas technique be used to accurately measure enteric methane production from ruminally cannulated cattle?

An experiment was conducted to determine whether using ruminally cannulated cattle affects the estimate of enteric methane (CH(4)) emissions when using the sulfur hexafluoride (SF(6)) tracer technique with samples taken from a head canister. Eleven beef cattle were surgically fitted with several types of ruminal cannula (2C, 3C, 3C+washer, 9C; Bar Diamond, Parma, ID). The 2C and 3C models (outer and inner flanges with opposite curvature) had medium to high leakage, whereas the 9C models (outer and inner flanges with the same curvature) provided minimum to moderate leakage of gas. A total of 48 cow-day measurements were conducted. For each animal, a permeation tube containing sulfur hexafluoride (SF(6)) was placed in the rumen, and a sample of air from around the nose and mouth was drawn through tubing into an evacuated canister (head canister). A second sample of air was collected from outside the rumen near the cannula into another canister (cannula canister). Background concentrations were also monitored. The methane (CH(4)) emission was estimated from the daily CH(4) and SF(6) concentrations in the head canister (uncorrected). The permeation SF(6) release rate was then partitioned based on the proportion of the SF(6) concentration measured in the head vs. the cannula canister. The CH(4) emissions at each site were calculated using the two release rates and the two CH(4):SF(6) concentration ratios. The head and cannula emissions were summed to obtain the total emission (corrected). The difference (corrected - uncorrected) in CH4 emission was attributed to the differences in CH(4):SF(6) ratio at the 2 exit locations. The proportions of CH(4) and SF(6) recovered at the head were greater (P < 0.001) for the 9C cannulas (64% and 66%) compared with the other cannulas, which were similar (P > 0.05; 2C, 6% and 4%; 3C, 17% and 15%; 3C+washer, 19% and 14%). Uncorrected CH(4) emissions were ± 10% of corrected emissions for 53% of the cow-day measurements. Only when more than 80% of the SF(6) escaped via the rumen did the difference between the uncorrected and corrected CH(4) emissions exceed 20%. We concluded that using cannulated cattle introduces more variability into the SF(6) technique used with a head canister, a technique that is already highly variable. Thus, use of cannulated animals is not recommended when using the SF(6) technique with head canister. However, if cannulated cattle are used, the cannulas need to be tight-fitting to minimize leakage, and large animal numbers are needed to overcome the additional variability.