The nutrition of grazing ewes during pregnancy and lactation:: a comparison of alkane-based and chromium/in vitro-based estimates of herbage intake

The n-alkane and chromium/in vitro procedures for estimating herbage intake were compared in grazing ewes during late pregnancy, early lactation, and mid-lactation. To ensure differences in herbage intake, the ewes were grazed in 4 plots of phalaris-dominant pasture at 2 levels of stocking: 17.1 ewes/ha and 30.8 ewes/ha. To investigate whether either procedure for estimating herbage intake was influenced by supplement consumption, half of the ewes at each stocking level received 500 g/day air-dry of a pelletted supplement (1 : 1 milled oat grain : sunflower meal). Supplement intakes were estimated using tritiated gypsum as a marker. During intake measurement periods, ewes were dosed twice daily with both alkane capsules and capsules containing chromium sesquioxide. For the last 6 days of the 12-day dosing period, rectal faecal samples were taken twice daily, immediately before the dosing. Over these same periods, wether sheep fitted with faecal collection harnesses were similarly dosed and sampled, and their total faecal output collected to establish the faecal recovery of chromium and the alkanes. Herbage intakes were estimated using the C27/C28, C29/C28, C31/C32, and C33/C32 alkane pairs. Estimates of intake based on the shorter alkane pairs were lower than those estimated with the C33/C32 alkane pair, by amounts which differed between the periods. Evidence is presented that estimates based on the last pair of alkanes (C33/C32) are the most accurate and are also more accurate than those based on the chromium/in vitro procedure. The relationship between these 2 methods for estimating intake was different in mid-pregnancy compared with either stage of lactation. The consumption of supplement did not interfere with any of the methods for estimating herbage intake. Estimates of faecal output based on the use of chromium, C28 alkane, or C32 as an external marker were statistically identical, indicating that the difference between the 2 methods for estimating herbage intake was not related to a failure to accommodate the incomplete recovery of any of the markers used or to the failure of rectal grab samples to be representative of total faeces. Our results indicate that herbage collected by oesophageally fistulated (OF) sheep was representative of that grazed by the ewes and could thus be used to provide the herbage alkane data needed to estimate herbage intake by the alkane method. However, the in vitro digestibility values obtained from the OF samples did not represent the digestibilities actually occurring in vivo. This was the main cause of the observed difference between the 2 methods for estimating intake. Possible reasons for the differences between the in vitro and in vivo estimates of digestibility are discussed.