Two experiments at Belclare, Co. Galway, in late autumn 1988, evaluated the use of herbage and dosed n-alkanes for estimating herbage intake by sheep. The first experiment examined faecal recoveries of dosed and herbage n-alkanes. The second experiment assessed the accuracy and precision of herbage intake estimates obtained using the n-alkane technique, and tested the effect of supplying n-alkanes to animals either in gelatine. capsules (containing different ratios of n-alkane: cellulose fibre) or in pellets prepared from shredded paper onto which the n-alkanes had been adsorbed. Individually penned wether lambs were offered freshly cut herbage ad libitum (+ 10%) and actual dry matter intake was recorded daily. Intake was estimated using the C31:C32 and C33:C32 (natural:dosed) n-alkane ratios.

There was no significant effect of n-alkane chain length on faecal recovery rate for either the dosed n-alkanes (C32 and C36), the herbage odd-chained n-alkanes (C29, C31, C33 and C36) or those used for the estimation of herbage intake (C31, C32 and C33). The accuracy and precision of the n-alkane technique for estimating herbage intake were unaffected by whether the dosed n-alkane was supplied in capsules or pellets or by the n-alkane:cellulose fibre ratio in the capsules. The bias in the estimated intake was – 8% (± 1·1%) and + 3% (± 1·2%) for estimates based on C31:C32 and C33:C32 ratios, respectively. The estimates based on C31:C32 and C33:C32 exhibited similar precision in the estimation of herbage intake, with a R.S.D. of 6% in actual intake when adjusted for variation in estimated intake and a correlation of +0·92 between actual and estimated herbage intake. The C.V. for actual herbage intake was 17%. The repeatability of actual dry matter intake over three consecutive 6-day periods was 0·54 while those of estimated intake were 0·57 and 0·60 for estimates based on C31:C32 and C33:C32, respectively. The results show that the n-alkane technique can provide an accurate and precise estimate of herbage intake.