Feed intake of sheep as affected by body weight, breed, sex, and feed composition

• Published in: Journal of animal science Vol. 88; no. 2; pp. 467 - 480
• Main Authors: Lewis, R.M, Emmans, G.C
• Format: Journal Article
• Language: English
• Published: Champaign, IL American Society of Animal Science 01.02.2010; Am Soc Animal Sci
• Subjects: ad libitum feeding; Animal Feed; animal genetics; Animal Husbandry; Animal productions; Animals; Biological and medical sciences; Body Size - genetics; Body Size - physiology; body weight; Body Weight - genetics; Body Weight - physiology; breed differences; Breeding; Dietary Proteins - metabolism; Eating - genetics; Eating - physiology; equations; estimation; Feed and pet food industries; feed composition; feed intake; feeds; Female; females; Food industries; Fundamental and applied biological sciences. Psychology; gender differences; genetic size-scaling; Genotype; growth traits; Hybridization, Genetic - genetics; Hybridization, Genetic - physiology; liveweight gain; Male; males; methodology; prediction; quadratic polynomial equations; ruminant nutrition; Sex Factors; sheep; Sheep - genetics; Sheep - physiology; sheep breeds; sheep feeding; statistical models; Terrestrial animal productions; traits; validity; Vertebrates; Farming animal; Animal feeding; Body weight; Selection; Size; Sex; feed intake; Vertebrata; Meat animals; Mammalia; Genetics; Sheep; genetic size-scaling; Chemical composition; feed composition; Artiodactyla; Ungulata; Feed
• ISSN: 0021-8812; 1525-3163
• DOI: 10.2527/jas.2008-1735

The hypotheses tested were that genetic size-scaling for mature BW (A, kg) would reduce variation in intake between kinds of sheep and that quadratic polynomials on u = BW/A with zero intercept would provide good descriptions of the relationship between scaled intake (SI, g/A⁰.⁷³ d) and degree of maturity in BW (u) across feeds of differing quality. Both sexes of Suffolk sheep from 2 experimental lines (n = 225) and from 3 breed types (Suffolk, Scottish Blackface, and their cross; n = 149) were recorded weekly for ad libitum feed intake and BW; recording of intake was from weaning through, in some cases, near maturity. Six diets of different quality were fed ad libitum. The relationship between intake and BW on a given feed varied considerably between kinds of sheep. Much, but not all, of that variation was removed by genetic size-scaling. In males, the maximum value of SI was greater than in females (P = 0.07) and was greater in Suffolk than in Scottish Blackface, with the cross intermediate (P = 0.025); there was no difference between the 2 Suffolk lines used (P = 0.106). The quadratic polynomial model, through the origin, was compared with a split-line (spline) regression for describing how SI varied with u. For the spline model, the intercept was not different from zero in any case (P > 0.05). The values of u at which SI achieved its maximum value (u* and SI*) were calculated. Both models fit the data well; the quadratic was preferred because it predicted that SI* would be achieved within the range of the long-run data, as was observed. On a high quality feed, for the spline regression, u* varied little around 0.434 (SD = 0.020) for the 10 different kinds of sheep used. For the quadratic, the mean value of 0.643 (SD = 0.066) was more variable, but there were no consistent effects of kind of sheep. The values of u* and SI* estimated using the quadratic model varied among the 6 feeds: 0.643 and 78.5 on high quality; 0.760 and 79.6 on medium protein content; 0.859 and 73.3 on low protein content; 0.756 and 112 on a low energy content feed; 0.937 and 107 on ryegrass; and 1 (forced, as the fitted value of 1.11 was infeasible) and 135 on Lucerne. The value of u* tended to increase as feed digestibility decreased. We conclude that genetic size-scaling of intake is useful and that a quadratic polynomial with zero intercept provides a good description of the relationship between SI and u for different kinds of sheep on feeds of different quality. Up to u [congruent with] 0.45, intake was directly proportional to BW.