Cowgame: animal breeding simulation software

Abstract

A primary aspect of beef production systems is the ability to capitalize on the theories of animal breeding and selection. However, a great deal of time and effort can be spent experimenting with poor breeding programs or selection criteria. The Cowgame is a computer simulation software application designed as a teaching tool to allow students, animal breeders, or interested parties to experiment with simple or complex aspects of animal breeding theory. This software is designed to be run on an IBM-PC 486 computer or equivalent with approximately 16 megabytes of random access memory and a math coprocessor chip. Run times associated with this software range from 10 minutes to 48 hours depending on the number of herds and the type of computer the software is run on. The Cowgame was developed to model real-life beef cattle herds. The quantitative, correlated characters of birth weight, weanin(ir weight, yearling hip height, and yearling weight with the qualitative character for hydrocephalus are simulated for each anii-nal. Five sires and 50 dams are simulated for each herd. A maximum of 125 herds can be developed at once. The five sires are mated to 10 dams each to produce offspring in the first generation. The herd owner then enters selection and mating decisions based on sires, dams, and offspring to allow the software to simulate the next generation. For each (veneration ,-al all available phenotypic data are analyzed using Multiple Trait t Derivative Free Maximum Likelihood (MTDFREML) to calculate Expected Progeny Differences (EPDS) and accuracies so that the analysis is completely independent of simulation. Output of phenotypes, EPDS, accuracies, and progeny averages are available for each animal in the population and are reported to each owner for each generation. Genetic progress can be measured by the owner because true simulated breeding values and averages are available upon request from the software. This option is restricted to the administrator of the project, so a herd owner may go through several generations before the administrator makes the information available. The software was tested for means, variances, and covariances to indicate that the simulation i-models were properly implemented.