The feeding of whole barley grain to breeding ewes is widely recommended yet it is recognised that a proportion of the grain fed in this form is usually passed intact in the faeces. It is claimed that whole grain improves the rumen fermentation process because it is broken down more slowly than milled grain and this, together with the saving in the cost of processing, means that a loss of up to 0.10 of grain intake in the faeces is considered acceptable. However there have been claims that on silage-based diets the proportion of grain passing intact in the faeces may be higher than 0.10 (Vipond, Hunter and King, 1985). In a preliminary study at this Institute with ewes on silage based diets in late pregnancy 0.18 of grain consumed was recovered intact in the faeces.

In 1988 45 ewes in late pregnancy were fitted with faecal collection bags as described by Vipond, Hunter and King (1985) during two 4-day periods 5 and 3 weeks before lambing and during a further 4-day period 3 weeks after lambing. In the two pre-lambing collection periods ewes were allocated at random from uniform groups to one of 5 forages (4 silages and 1 hay). Three of the silages were made from primary growths of a ryegrass pasture, one cut on 25 May with a single-chop harvester (SI) and two cut on 8 June, one with a single-chop (S2) and one with a precision-chop (S3) harvester. The fourth silage was a second cut, harvested on 29 June with a precision-chop harvester (S4). Hay was of medium quality and was bought in. All forages were offered. ad libitum so that intake was 0.9 of forage offered. Ewes were offered 250 and 600 g/d whole barley 5 and 3 weeks before lambing respectively and 800 g/d 3 weeks after lambing. In the faecal collection after lambing all ewes received the same roughage (silage S4) and values for the proportion of intact grain in the faeces at this stage were used to adjust values 5 and 3 weeks pre-lambing by covariance.