Mathematical modelling in agricultural systems: A case study of modelling fat deposition in beef cattle for research and industry

Agricultural production and marketing systems are rich in complexity and diversity. Understanding how such systems are structured and function requires the use of mathematical modelling frameworks that represent their key features. Disciplines that are required to develop and solve such models include: biology, computer programming, statistics, mathematics, economics, and social science. More recently models have been developed to investigate specific issues such as the effects of climate change, climate variability, and to assist third world countries improve their production at the farm gate. This paper presents: • An overview of mathematical modelling: - Past, present, and future - Modelling animal growth - Modelling body fat • Process of model development: - Level of aggregation - Classification of models • Case study of fat deposition in beef steers: - Biology of fat - Measurements of fat - Model development of the Davis Growth Model - Model development of BeefSpecs The case study of fat deposition in beef cattle provides an illustration of: 1. The modelling process from the cell to the animal level for the research model (Davis growth model) and; 2. The modelling at the herd level for the industry model (BeefSpecs). BeefSpecs has been developed for producers and livestock officers and is available for use from the Meat & Livestock Australia (MLA) web site (http://www.mla.com.au). Both the process of model development and the biology of fat were the building blocks to develop the research and industry models. A clear understanding of the 'how', 'why' and 'what' of any model is fundamental and it is imperative that models start out at a simple level and add complexity as required. At the very outset clear and concise objectives need to be stated. Future modelling opportunities lie in the hands of large scale data sets generated from -omic (genomics, transcriptomics, proteomics, and metabolomics) technologies. Agricultural systems that make long-term and short-term predictions will continue to make significant contributions to the agricultural industries, but the number of data inputs that are required to drive the models are contrastingly different.