ARC (1980), in its Chapter 2, gives a method for predicting the rate of feed intake of a sheep of a given weight, W, on a given feed (feed class and energy metabolisability, q). In its Chapter 3 a method is given for predicting the growth rate, GR, of a sheep of a given sex and W from its ME intake. Where GR and W are known the latter procedure can be used to predict the rate at which feed, needs to be eaten, from the ME content and q values of the feed, to support the observed GR. A third method is that of Emmans (1985), which predicts the rate of intake of a non-constraining feed with energy as the first limiting resource by a given animal in a given initial state in a thermally neutral environment; the animal is described by its mature protein weight, Pm kg, its mature lipid:protein ratio, LPRm, its Gompertz rate parameter for protein growth, B/day, and its fattening rate parameter, b, where (b + 1) is the allometric exponent for body lipid on body protein in potential growth.

The three methods were used to predict the intakes of Suffolk rams and the predictions compared with the intakes observed in an experiment.

The data were from 66 Suffolk rams born in mid-January, weaned at 8 weeks into single pens and given free and continuous access to the feed shown in Table 1. Their intakes and weights were recorded for successive periods of 2, 2, 2, 2, and 3 weeks from 20th March. The sheep were split into three groups on the basis of their weights on 20th March with mean values of 22, 27 and 33 kg. On the basis of their weights the second and third groups were judged to be 9 and 19 days respectively ahead of the first group.