|
Centre for Genetic
Improvement of Livestock All The Answers |
||||
|
|
Realism and
flexibility -- those are the critical characteristics a University of
Guelph research group says will propel a new technology its developed to
new heights in dairy herd simulation.
Drs. Andy Robinson, Gerald Jansen, Jane Tosh, and Ching Lin, Centre for Genetic Improvement of Livestock, are completing what they believe is one of Canada's most advanced dairy data simulation programs, the Dairy Data Simulator (DDS). What Are The Long Range Effects? The DDS is a software program that gives researchers a way to create a realistic population of computer-generated dairy cows to predict the long range effects of experimental production practices. The program is highly flexible, capable of assisting in matters that include nutrition, genetics and management. "We're building a sophisticated research tool," says Robinson. "The Dairy Data Simulator can be used to quickly and efficiently answer a countless number of questions that the dairy industry may have." The simulation process begins when researchers input information about the virtual population they want to create; such as the number of cows they'd like to use and the experimental changes they're making. After the initial input, the computer takes over, mimicking the Canadian dairy industry as closely as possible. A dairy population of thousands is artificially generated with the genetic make-up and variability of current Canadian-bred Holsteins. This simulation herd is then divided into herds averaging 45 head (the average number in a typical Canadian dairy herd). The virtual cows are kept for three lactations and poor performers are culled and replaced when needed. Young bulls are tested according to industry standards; the best are kept and used for breeding for up to 12 years in a simulated artificial insemination unit. |
Creating Virtual Cattle
Within a matter of hours, many generations of virtual cattle are produced. Output from the program is given in terms of milk 'test day' records (the simulator is also the only program of its kind that's directly compatible with the model used by the Canadian Dairy Network to generate bull and cow proofs).
The first project for the simulator will be to look at the effects of reducing the number of milk test days per cow each year. Although milk testing is critical to producers in terms of finding the best individuals to breed, it can be very expensive. The program has already simulated a reduction of test days from 10 times a year to five, and found that it doesn't affect progress of the breeding program decades into the future. The possibilities don't end there -- the DDS may soon be giving answers to researchers from a wide range of fields. "We can look at anything from breeding strategies to new feed additives to the type of floor surface used in a stall," Robinson says. "If we know the initial effects on just a few cows, we can accurately extend it to cover thousands." This research is supported by DairyGen (formerly the Cattle Breeders' Research Council) and Agriculture and Agri-Food Canada. |
||
