As part of a multibreed cattle project for studying genetic variation between breeds, four adult females from each of five breeds of cattle were kept for up to 2 years in a non-pregnant, non-lactating state on each of four feeding levels. The breeds, which were Hereford, Aberdeen Angus, Dexter, British Friesian and Jersey, provided a wide range of genetic potential for body size and milk yield. The feeding levels were arranged to make the weight of lipid in the whole body about 0·05, 0·15, 0·25 and 0·35 times body weight. After an animal had been assigned to a fixed daily intake, its body weight and condition score were monitored over a period varying from 6 to 18 months until an equilibrium body weight and body composition had been established.For standard adult body weight, Akg, and for an equilibrium body weight, Wkg, and an associated metabolizable energy intake, f/MJ, the maintenance efficiency of a breed adjusted for breed size was defined as Em = W/fA0·21. The ‘lactability’ of a breed, that is, its genetic potential lactation yield, Ykg, adjusted for breed size, was defined as Y = Y/A. The between-breed regression of Em on Y had a coefficient of -0·043 (s.e. 0·007). There was, thus, a highly significant dairy-beef gradient in the equilibrium maintenance efficiency of these non-pregnant, non-lactating adult females.An expected value for the equilibrium maintenance requirement, of a breed or individual as a function of its lactability was estimated to be:which implies that the maintenance requirement of dairy breeds is about 0·2 times greater than that of beef breeds, a result which was strongly supported by a literature survey, although mean estimates from feeding trials and fasting trials were anomalous.Combining this result for milk yield with published results on growth rate, leads to the hypothesis that the greater the maximum gross efficiency of a breed for meat or milk production, the less efficiently it maintains itself, not because of any difference in fasting metabolism, but because of a reduced efficiency of food utilization for maintenance.

Males and females from Soay, Welsh Mountain, Southdown, Finnish Landrace, Jacob, Wiltshire Horn and Oxford Down sheep breeds and a breed of feral goats were slaughtered when proportionately 0-40, 0-52, 0-64 or 0-76 of mature live weight. On average, carcasses contained 583 g muscle, 243 g fat and 174 g bone per kg. Individual cuts (four commercially higher-valued and six lower-valued cuts of the shoulder, rib, loin and gigot joints) ranged from 375 to 670g muscle, 129 to 625 g fat and 0 to 294 g bone per kg.Allometric coefficients are given for 32 traits. The distribution of bone in the carcass joints changed little as animals matured but carcass muscle and fat distributions changed significantly.Breeds differed significantly in proportion of carcass muscle, fat and bone and also in their distribution. Welsh Mountain, Southdown, Wiltshire and Oxford Down all deposited carcass fat about three times more rapidly than did the Soay, goat, Finnish Landrace and Jacob. The Oxford Down had the lowest proportion of muscle (514 g/kg) and most fat (317 g/kg). The Southdown had least bone (148 g/kg) and the feral goat most muscle (662 g/kg). Southdown and Soay had the most muscle in commercially higher-valued cuts and Jacob and feral goat the least. Males had significantly more of their muscle in the neck and shoulder and significantly less in the gigot and flank.As breed size increased, proportion of carcass muscle and bone decreased and proportion of carcass fat increased. These breed regressions were attributed to the small but exceptionally lean feral Soay and goat breeds, rather than to a failure of genetic size-scaling. There were no significant trends with breed size among the domesticated breeds. t , r t " . u^n a ;" o quantified by allometry and also by the T H E proportion of muscle, fat and bone in a M J . J J u i I » • n •• v slaughtered at four degrees of maturity. c u s ' Changes in proportions of muscle, fat and MATERIAL AND METHODS

Animals from Soay, Welsh Mountain, Southdown, Finnish Landrace, Jacob, Wiltshire Horn and Oxford Down sheep breeds and a breed of feral goats were slaughtered when 0-40, 0-52, 0-64 or 0-76 of mature body weight. The effects of stage of maturity, breed and sex were examined in the following traits: gut content as a proportion of slaughter weight; blood, omental plus mesenteric fat, perirenal fat, alimentary tract, 'remainder' of non-carcass parts and hot carcass, each as a proportion of fleece-free empty body weight; also shoulder, rib, loin and gigot joints and their commercially higher-valued (prime) and lowervalued cuts, each as a proportion of carcass weight.As a proportion, gut content declined as animals matured, as also did blood, alimentary tract, 'remainder', prime shoulder, shank and prime gigot. Increasing proportions were internal fat depots, the hot carcass, rib, loin and gigot flank. Males accumulated significantly less internal fat and more shoulder than females.When equally mature in live weight, males had proportions that were significantly lower for perirenal fat, rib and gigot and higher for shoulder and pelt plus head, feet and organs.There were significant inter-breed regressions on mature body weight. The proportion of blood, alimentary tract, pelt plus head, feet and organs, and prime gigot declined with breed size, while hot carcass, brisket, rib and loin proportions increased. These regressions were attributed to a non-random sample of breeds. The small Soay breed had proportions of internal fat and hot carcass which were well below average, and higher proportions of blood, alimentary tract and 'remainder'. Soay sheep also had a higher proportion of commercially higher-valued cuts in the carcass. In contrast, the large Oxford Down breed had low proportions of non-carcass components (except alimentary tract) and high proportions of lower-priced cuts and prime rib. Feral goats had a greater, and Southdown sheep a smaller, proportion of gut content than other breeds.

Body composition was studied in 20 mature, non-pregnant, non-lactating cows from five breeds (Hereford, Aberdeen Angus, Dexter, British Friesian and Jersey) kept on four feeding levels until they attained equilibrium body weights that were proportionately 0-7,0-9,1-1 or 1-3 of their normal adult body weight. Significant breed differences were found in the proportions of body tissues and organs and these were associated with breed differences in lactability (i.e. genetic milking potential adjusted for body size). As a proportion of body weight, intraabdominal fat, liver, spleen and uterus increased significantly with lactability and hide decreased significantly. Empty gut and gut fill also increased with lactability but not significantly. Liver proportion in dairy breeds was 1 -26 times the proportion in beef breeds. Corresponding values for intra-abdominal fat and hide were 1-43 and 0-83. The most dramatic increases with feeding level were in the proportions of subcutaneous fat, both intra-abdominal fat depots, and the udder. All fat depots were completely depleted when body weight decreased to about 0-6 of its normal adult value. Strong decreases occurred in the proportion of muscle, carcass bone and offal. The proportion of empty gut decreased significantly with increased feeding level. Liver, tail, thymus and possibly gut fill were the only traits entirely unaffected by feeding level.The near-constancy of liver proportion at equilibrium implies that the rapid response of the liver to a change in feeding level is eventually matched in magnitude by the slower responses in other tissues and organs, so that the original proportionality of about 1 kg body tissue for each 10 g liver is eventually restored.