Models of pig growth: problems and proposed solutions
Models of pig growth: problems and proposed solutions
复制标题
DOI:
10.1016/s0301-6226(97)00061-4
复制
发表时间:
1997-11-01
期刊:
影响因子:
--
通讯作者:
Kyriazakis, I
中科院分区:
文献类型:
--
作者:
Emmans, GC;Kyriazakis, I
A system to predict the growth and body composition of a pig from a knowledge of its genotype and diet is one of both scientific and practical interest and concern. Three of the major problems that need to be solved in order to produce such a predictive system are considered here: (i) the prediction of the potential rate of protein retention from a description of the pig and its state; (ii) the prediction of the rates of retention of the two other components of the lipid-free body which are ash and water; (iii) the prediction of the actual rates of retention of protein and lipid from a knowledge of the rate of intake of food and its composition. The existing proposed answers to these problems are described and criticised. We then present what, in our view, are the best solutions currently available. The first problem is solved by the use of the Gompertz growth equation. The second problem is dealt with by allometry but with a better theoretical underpinning. It is proposed that the third problem, i.e., that of predicting the actual rates of retention of protein and lipid, when the rate of protein retention is below the potential rate because of the limiting food supply, can be solved by using a simple rule. The rule is that the marginal material efficiency of using the ideal protein supply above maintenance for growth, is directly proportional to the energy: protein ratio of the food up to a maximum value. Thus the amount of extra protein gained per unit of extra food energy depends only on the amount of extra (ideal) protein supplied with it and is independent of liveweight and of pig genotype. The solutions to the three problems constitute a framework which can be applied to predict the response of a pig to its diet at a time and over time. Given the small number of parameters that need to be evaluated, the framework can be extended to predict the responses of different kinds of pigs to the same diet. Providing that the distributions of the values of the parameters are known, the approach can be adopted to predict the responses of the individuals which comprise a population of pigs to their diet. (C) 1997 Elsevier Science B.V.