A primary goal of the research programme of the International Livestock Centre for Africa (ILCA) is to identify opportunities for improvements in tropical African livestock systems, and studies on the productivity of cattle herds are an integral part of this research. However, the consequences of changes in cattle productivity at the herd level are manifested over time in a complex way which can only be predicted probabilistically. At the same time, field research on cattle herd productivity requires substantial resources and an extended period before results are available. For this reason, ILCA's field research programme is complemented by mathematical modelling. The objectives of this approach, integrating field research and modelling, are to provide better predictions of the consequences of change than are possible from informal calculations, to maximize the transferability of research results and to engender and support the close integration of research in all the appropriate scientific disciplines.
Due to earlier limitations in manpower and computer facilities, ILCA's initial modelling efforts were based on the use of existing models. The Animal Science Systems Group at Texas A & M University (TAMU) had developed a model of cattle production which represented an appropriate introduction to modelling techniques for livestock researchers in Africa. In 1978, ILCA, TAMU and the Botswana Government's Animal Production Research Unit cooperated in the application of the TAMU model to commercial ranching and traditional cattle systems in Botswana. ILCA Systems Study No. 1 reports on that research.
This exploratory work highlighted the need for a simulation model with stochastic features - particularly for the forage component, as year-to-year variability in the quality and quantity of forage available is a key determinant of herd productivity in nearly all African livestock production systems. It was also considered that an integer-based model, i. e. one which treats animals in the simulated herd as individual entities, was most appropriate for simulating the generally small herd production units in Africa. Additionally, the model had to be structurally flexible, so that it could be readily applied to a range of production situations without major modifications.
This Systems Study presents the rationale and formal specifications of an operational model with these features. The authors wish to thank J C M Trail for his constructive criticism and substantive comments from the initial conception to the final preparation of this report. They are also grateful to C de Haan, S Westley, C R W Spedding, P Chudleigh, J King, J Durkin, H Kahn, K Milligan and R von Kaufmann for many useful suggestions on an earlier draft, and to G Maloba for typing the final text. However, any errors or deficiencies that may remain are the authors' sole responsibility.
ABSTRACT
A general cattle herd simulation model is presented in which a herd is simultaneously represented as both a biological and an economic unit. The model is time dynamic, stochastic, non-optimizing and integer, that is, it treats animals in the simulated herd as individual entities. The model provides the user with an array of policy options for weaning, breeding, milking, buying and selling stock, and supplementation to increase production or for strategic reasons during periods of drought - so that herd performance can be evaluated under alternative production regimes.
Five general components in the model account for changes in the biological status of animals during each month of simulation: forage intake, energy requirements, production and growth, mortality and reproduction. The quantity and quality of forage on offer are specified stochastically. The parameters of mathematical relationships of the various biological processes are drawn from the literature and specified for particular systems under study based on observations from these systems. Thus, the model is data based where possible. The FORTRAN computer code is adequately modularized so alterations and refinements can be made easily.
KEY WORDS
Cattle herd, simulation model, stochastic, production alternatives, Africa, feed intake, energy requirements, growth, mortality, reproduction.
RESUME
Le présent document expose les caractéristiques d'un modèle général de simulation dans lequel un troupeau de bovins est considéré à la fois comme unité biologique et unité économique. Le modèle est dynamique dans le temps, stochastique, non-optimisant et numériquement entier dans la mesure où il traite les animaux du troupeau simulé en tant qu'entités individuelles. Il fournit à l'utilisateur toute une gamme de possibilités en ce qui concerne le sevrage, la reproduction, la traite, l'achat et la vente du cheptel, ainsi que la complémentation destinée à accroître la production ou utilisée pour des raisons stratégiques pendant les périodes de sécheresse afin que l'évaluation des performances du troupeau puisse s'effectuer par le système de facteurs alternatifs dans le cadre du système de production
Cinq composantes générales du modèle, à savoir consommation fourragère, besoins énergétiques, production et croissance, mortalité et reproduction, expliquent les changements de l'état biologique des animaux pour chaque mois de simulation. Dans le modèle, la quantité et la qualité du fourrage offert sont spécifiés stochastiquement. Sur la base des données fournies par la documentation disponible, les paramètres des relations mathématiques existant entre les divers processus biologiques sont spécifiés pour un système donné, à la lumière des renseignements tirés de l'observation dudit système. Ce modèle se fonde donc, autant que possible, sur l'utilisation de données. Le code FORTRAN automatisé a été adéquatement modulé pour faciliter les modifications et les améliorations éventuelles.
MOTS-CLES
Troupeau de bovins, modèle de simulation, stochastique, Afrique, alternative de production, besoins énergétiques, croissance, reproduction, mortalité.
