A method of using a genetic algorithm to examine the optimum structure of the Australian sheep breeding industry: open-nucleus breeding systems, MOET and AI

A genetic algorithm was evaluated as a means of using a computer model to determine the optimum structure of the Australian sheep breeding industry. The model simulates an open-nucleus 3-tiered sheep breeding system to investigate the benefit of strategies such as multiple ovulation and embryo transfer (MOET) and artificial insemination (AI) in terms of both genetic improvement and dollar values. The model indicated that both MOET and non- MOET systems benefit from an open nucleus, and both could also benefit from the promotion of some ewes from the commercial flocks. However, MOET systems require a relatively large number of rams, whereas non- MOET systems should have a smaller number of rams despite the much larger number of ewes needed. Under the conditions studied here, there was only limited advantage of MOET over non-MOET schemes. The optimum structure of the MOET, non-MOET and closed systems differ substantially. A genetic algorithm is a simple method for considering a single change (such as reproductive rate, or open v. closed) that may require major changes in the industry structure to achieve the full benefits. The method could also be of value in many other situations requiring optimisation of complex models.

[1]  J. James,et al.  On the derivation of economic weights in livestock improvement , 1986 .

[2]  S. K. Ercanbrack,et al.  Effects of inbreeding on reproduction and wool production of Rambouillet, Targhee, and Columbia ewes. , 1991, Journal of animal science.

[3]  N. Jackson,et al.  Optimal structure for a cooperative nucleus breeding system. , 1972 .

[4]  Samir W. Mahfoud.  Niching methods for genetic algorithms , 1996 .

[5]  R. Ponzoni.  The derivation of economic values combining income and expense in different ways: An example with Australian Merino Sheep , 1988 .

[6]  D. S. Keller,et al.  A comparison of factors reducing selection response in closed nucleus breeding schemes. , 1990 .

[7]  W. G. Hill.  Prediction and evaluation of response to selection with overlapping generations , 1974 .

[8]  Martin Wattenberg,et al.  Stochastic Hillclimbing as a Baseline Mathod for Evaluating Genetic Algorithms , 1995, NIPS.

[9]  H. N. Turner,et al.  Quantitative Genetics in Sheep Breeding , 1969 .

[10]  Brian J. Horton.  Factors limiting the value of AI and MOET in a three-tiered sheep breeding system , 1995 .

[11]  P. Bird,et al.  The choice of discount rate in animal breeding investment appraisal. , 1980 .

[12]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[13]  J. James.  Effective Population Size in Open Nucleus Breeding Schemes , 1978 .

[14]  Brian J. Horton.  Optimal structure for a three-tiered co-operative sheep breeding system using AI and MOET , 1992 .

[15]  G. Evans.  Application of reproductive technology to the Australian livestock industries. , 1991, Reproduction, fertility, and development.