Validation of the sulphur hexafluoride (SF6) tracer gas technique for measurement of methane and carbon dioxide production by cattle

Methane (CH4) and carbon dioxide (CO2) production from six crossbred yearling beef heifers (400 +/- 13.0 kg) were measured, using the sulphur hexafluoride (SF6) tracer gas technique (Tracer) and open-circuit hood calorimetry (Cal) to validate the former in estimating rumen CH4 and CO2 production in the field. Animals were individually fed a diet consisting of 50% barley concentrate and 50% alfalfa cubes at 1.3 x maintenance requirements daily. Heifers were divided into two groups for individual animal 24-h gas measurements by each method. Each group of heifers was rotated between the Cal and Tracer techniques for 6 consecutive days in an incomplete block design. Methane production ranged from 108 to 145 L d(-1) (mean 130 +/- 4.0 L d(-1)) using the Cal technique, and 90 to 167 L d(-1) (mean 137 4.0 L d(-1)) using the Tracer technique. The mean CH4 production (L d(-1)) was not different (P = 0.24) between the two methods. Carbon dioxide production with the Tracer technique was 20% higher than CO2 production with the Cal technique (P < 0.01). The range of CO2 production was 1574 to 2049 L d(-1) (mean 1892 +/- 74.0 L d(-1)) by Cal, and 1541 to 3330 L d(-1) (mean 2353 +/- 74.0 L d(-1)) by Tracer. Day-to-day variation in CH4 production was not different within each method (P > 0.05); however, animal-to-animal variation (11.7%) was significant for the Tracer technique (P = 0.04), but not for the Cal technique (P = 0.53). Comparison of the equality of variance between the two methods showed that there were no differences in variations (P > 0.05) between Cal and Tracer for CH4 production. On the other hand, variations in CO2 production were not equal (P > 0.05) between methods. Day-to-day variation in CO2 production was significant using Cal, but not Tracer (P > 0.05). Animal-to-animal variation in CO2 production was 1.6 and 11.8% by Cal and Tracer techniques, respectively. It can be concluded that the SF6 tracer technique accurately estimated rumen CH4 production, but CO2 production was 20% higher. The study suggests that for CH4 measurements using the SF6 tracer technique, more animal numbers are needed than for Cal to reduce animal-to-animal variation.