In two experiments with female Scottish Blackface or Border Leicester x Scottish Blackface sheep nine or ten animals were slaughtered in mid lactation and the remainder either in November at the time of the subsequent mating (Expt 1) or at weaning (Expt 2).
The sheep were infused periodically, including just before slaughter, with 100 μCi tritiated water (TOH) in order to measure total body water by dilution and to estimate body fat using the inverse relationship between the proportions of fat and water in the body. The accuracy of the methods was assessed when the sheep were slaughtered. In the ton lactating Scottish Blackface sheep of Expt 1 fat made up 11.5% of the total body weight with an S.D. of 8·38% whereas the 11 sheep slaughtered at mating were twice as fat (23·2, S.D. 4·01 %). The 25 Border Leicester × Blackface sheep were all thin, whether they were slaughtered in mid lactation (4·4, S.D. 2·56%) or at weaning (3·4, S.D. 2·81%).
The standard error of estimate of body water from TOH space in Expt 2 was 1·2 kg c.v. 2·8%) and lower in Expt 1.
The precision with which an animal could be weighed was very important in determining the accuracy with which body fat could be predicted from live weight and TOH space. In both experiments the standard error of estimate for body fat in lactating sheep was between 600 and 700 g, compared with 1·3–2·7 kg when body fat was predicted from live weight alone. These estimates were sufficiently accurate to be of value in following changes in body composition in live animals as their nutritional and physiological state alters and for comparing animals in groups where the average fatness is greater and the range wider than in the sheep used in Expt 2.