实际的猪生长模型是否足以预测生长和营养需求

Actual pig growth models, are they adequate to predict growth and nutrient requirements?

实际的猪生长模型是否足以预测生长和营养需求？

Candido POMAR 

Agriculture and Agri-Food Canada, PO 90, Lennoxville, QC, J1M 1Z3 Canada

加拿大农业和农业食品部，PO 90，伦诺克斯维尔，QC，J1M 1Z3加拿大

Pig production efficiency is the result of the responses of individual animals. The variation of individual animal performance results from between-animal genetic variation and the interaction between this genetic variation and the environment, including health and some management practices. To illustrate the impact of between-animal variation on population responses to dietary treatments, a mathematical pig growth model was used. Population responses to increasing levels of ideal protein intake indicated that the linear-plateau model used to represent protein responses for an individual pig is compatible with the curvilinear response observed in populations of pigs. However, the form of this response is affected by the between-animal variation. Variation between animals decreases population protein deposition rate, daily gain and feed conversion ratio. The differences between individual and population responses to increasing levels of available nutrients are important and can lead to large variation in the recommend amount of nutrients required to optimise population responses. It is concluded that mathematical models designed to simulate population responses to treatments or to predict nutrient requirements need to integrate the effect of population variation on growth and performance.

猪生产效率是个体动物反应的结果。动物间遗传变异和遗传变异与环境之间的相互作用，包括健康和一些管理实践之间的相互作用导致个体动物表现的变化。为了说明动物间变异对人群对膳食处理反应的影响，使用数学猪生长模型。群体对理想蛋白质摄入量增加水平的反应表明，用于代表个体猪蛋白质反应的线性平台模型与在猪群中观察到的曲线反应相一致。但是，这种反应的形式受到动物间变异的影响。动物之间的差异降低了群体蛋白质沉积速率，日增重和饲料转化率。个体和群体对可用养分水平增加的反应之间的差异很重要，可能会导致优化群体反应所需的营养素推荐量的大幅变化。得出的结论是，设计用于模拟群体对处理的反应或预测养分需求的数学模型需要综合考虑种群变异对生长和性能的影响。

INTRODUCTION

Animal feed is currently undergoing a major conceptual shift (SAUVANT, 1992) in which diets previously formulated to meet or exceed nutritional requirements are now optimized based on animal responses. These responses are evaluated in terms of zootechnical and economic efficiency, product quality, and impact on the environment, behavior, well-being and animal health (LOVATTO and SAUVANT, 1999). This development is leading to significant changes in feed unit systems, research programs in the field of applied nutrition, but also in the methods of estimating the nutritional contributions necessary to obtain the desired animal responses (SAUVANT et al, 1995, BLACK et al, 2002).

动物饲料目前正在经历一个重大的概念转变（SAUVANT，1992），其中饮食先前已经配制成满足动物的需要。 这些反应是根据畜牧技术和经济效率，产品质量以及对环境，行为，健康和动物健康的影响来评估的（LOVATTO和SAUVANT，1999）。 这一发展导致饲料系统，应用营养领域的研究计划发生重大变化（SAUVANT等，1995年，BLACK等，2002年）。

Pork production differs from other animal products in the use of mathematical modeling to represent the animal response in different nutritional, genetic and environmental contexts. Since the early works of WHITTEMORE and FAWCETT in 1974, many research groups across the The world has proposed mathematical modeling as a tool of choice to represent the interactions between the animal and its environment or to estimate the animal response to dietary intakes in different research or production contexts. Early models relied on average empirical laws to represent the different metabolic functions and the use of nutrients consumed. These models have since evolved into complexity becoming more mechanistic (less empirical), considering other influencing factors and having better characterized parameters. But in the majority of these models, the biological phenomena represented are those of a single animal and consequently, the response obtained is that of this animal, perhaps a medium, in the context of simulated production.

猪肉的生产与其他动物产品的不同之处在于使用数学模型来表示在不同的营养，遗传和环境背景下的反应。自1974年WHITTEMORE和FAWCETT的早期工作以来，世界各地的许多研究小组都已经提出了动物与环境之间相互作用的数学模型，或者估计动物对不同研究中的饮食摄入量的反应或生产环境。早期模型依赖于平均经验法则来表示不同的代谢功能和消耗的营养素的使用。这些模型已经演变成更机械的（不太经验的）关系。目标：在大多数这些模型中，生物现象是单一动物的现象，因此，在模拟生产的情况下，获得的反应是该动物的反应，可能是一种介质。

The zootechnical productivity of any animal production system, however, results from the productivity of each individual in the system. However, when we have to analyze or compare production systems, it is common practice to use the mean as an evaluation criterion. At the same time, there is little interest in the observed variances around these averages, although they may also be important components of overall system productivity (KNAP, 1995). On the other hand, the response of an individual can hardly represent that of a population since these two responses, that is to say that of an individual (representative of a population) and that of the population itself. Similarly, they are different in both shape and extent (POMAR, 1995, POMAR et al, 2003, WELLOCK et al, 2004, BERHE, 2004). Moreover, the gap between these responses increases with the heterogeneity of individuals (POMAR 1995, POMAR et al., 2003), which leads us to believe that the heterogeneity of populations must be considered in the interpretation of zootechnical responses of animals. populations and biological phenomena involved in these responses, when establishing the general laws governing the response of animals (POMAR et al, 2003, WELLOCK et al, 2004) or when determining the optimal level of nutrients necessary for growth animals (POMAR 1995, LECLERCQ and BEAUMONT 2000). In this context, we must therefore question the type of mathematical model to develop and the level of aggregation of these models when we try to simulate the response of a population of animals to optimize nutrient intake or control. 'breeding. The aim of this work is therefore to analyze the impact that the variation between animals of the same population can have on their zootechnical response and their nutritional needs and to give some reflections on the way in which the population's response must be represented in the simulation models of pork butchers.