Two experiments were conducted to evaluate factors influencing the release rate of octacosane (C28) and dotriacontane (C32) synthetic alkanes from controlled-release devices (CRD) inserted into the rumen of sheep, and the resultant accuracy of alkane-based estimates of herbage intake. In the first experiment, estimates of the rate of release of alkanes from CRD suspended in the rumen of 7 rumen-fistulated ewes were obtained by directly measuring the rate of CRD plunger travel. In the second experiment, CRD were administered orally to 24 castrate male sheep to evaluate the effects of level of feeding (1200 v. 600 g DM/day), frequency of feeding (continuous v. once-daily), and number of CRD inserted (1 v. 2) on temporal trends in faecal alkane concentrations and ratios and on the accuracy of estimation of intake. In both experiments, animals were also dosed orally, once daily with paper pellets containing 56 mg hexatriacontane (C36).
In Expt 1, linear alkane release from the CRD commenced after a lag period of 2-3 days, at mean rates of 40.1 mg/day (C28) and 41.7 mg/day (C32), close to the desired rates of 40 mg/day. There was no evidence of curvilinearity, and differences in release rate between sheep were small (CV 4.1%). Alkanes attributable to the CRD appeared in faeces after 2-3 days and reached plateau concentrations by 6-7 days. Data are presented to indicate that the release rate was not affected by the removal of CRD from the rumen for measurement, and that the release rates in fistulated (Expt 1) and intact sheep (Expt 2) were similar.
In Expt 2, there was no effect of level of feeding on CRD release rates and faecal alkane concentrations were consistent with the differences in faecal output caused by feeding level. As a result, the accuracy with which known intakes were estimated was unaffected by level of feeding. Similarly, there were no effects of feeding frequency on the accuracy of estimated intake, implying that in this study, feeding frequency had no effect on CRD release rates. The administration of 2 CRD to half the animals in this experiment increased faecal concentrations of C28 and C32 and the synthetic : herbage alkane ratios associated with these alkanes [e.g. faecal C32 : tritriacontane (C33)] but these changes were not significantly different from those expected from the release rates estimated in Expt 1. As a result, the accuracy with which known intakes were estimated was not influenced by the number of CRD administered.
In both experiments, once-daily dosing with C36 alkane resulted in a more rapid rise in faecal C36 concentrations and faecal C36 : pentatriacontane (C35) ratios to maximal levels (3-4 days), compared with alkanes derived from the CRD (6-7 days), but thereafter, faecal C36 concentrations were much more variable, especially in the fistulated sheep. As a result, intake estimates based on the C36 : C35 alkane pair significantly over-estimated known intakes by 13% (Expt 1) and 8% (Expt 2). Data are presented to indicate that C36 alkane functioned satisfactorily as a faecal output marker and, with C32 : C33 estimates of intake, provided accurate estimates of herbage digestibility.
The results indicate that intra-ruminal controlled-release devices provide a satisfactory means of delivering an accurate, daily dose of alkanes for the estimation of herbage intake, faecal output, and digestibility in sheep.