Analyzing ASU's Buck Test Data
2007 Angelo State University Meat Goat Performance Test concluded on
August 31, 2007, and I thought this month we should go through the data
and see what we can learn from this year's results.
Table 1 lists the results for each individual animal that was tested. The results are sorted with the buck that had the highest average daily weight gain (ADG) first, and the lowest ADG last.
The first column lists what place the buck came in based on ADG. The next column lists the buck's ear tag #. Next is the name of the breeder, followed by the buck's birth date and the breed of the animal. In the interests of saving space I left Boer off the description of Purebred and Fullblood.
The next two columns list the buck's weight when the test commenced on 06/08/2007 and when it concluded on my birthday, 08/31/2007. As you can see I had a very happy birthday (Double M Boer Goats is my ranch).
The column labeled ADG lists the buck's average daily weight gain. The top testing buck gained over 0.9 lbs per day while on test. The average for all animals on test was only 0.561 lbs/day. The last place buck gained 0.226 lbs/day.
However, frequently those slowest growing bucks suffered from illness while at the test site. For instance, the next to last place buck was one of mine, and not only did he do poorly at the test, but shortly after we brought him back he died from pneumonia. His lungs were full of scar tissue from previous bouts of respiratory infection.
The column labeled REA lists the buck's ribeye area. This is a measure of the amount of muscle (meat) the animal has. You have to be careful when using this column because the raw data is pretty meaningless. If you rank the bucks strictly off their REA, as the university does in the next column, you will usually select the largest animals not necessarily the best ones.
Obviously, a 130 lb buck is going to have a much larger ribeye area than a 70 lb buck. When I look at these results I am much more interested in knowing whether the buck has a REA that is larger or smaller than the average for an animal of his size. This curiosity led me to develop a table of REA's by size of the buck using nine years of test data.
I then use that table to calculate the REA as a percentage of the average REA for a buck of the same weight. That information is listed in the column labeled REA % Avg. I also rank the bucks based on that statistic. If the number listed is above 100%, the buck has a large REA for his size. If it is below 100%, than his REA is below average for a buck of his size.
The difference in the REA rankings when you adjust for differences in size can be quite dramatic. For instance, the #1 buck when REA is adjusted for size was only 39th if you looked at the raw data. A number of bucks that ranked very highly using the raw REA numbers are actually below average if you adjust the numbers for size differences.
The next two columns list the buck's scrotal circumferences at the beginning and end of the test.
Finally, the last column lists the buck's sire.
The individual animal results are fun to look at, but it is the results listed in Table 2 that are of the most interest to me. Table 2 lists the results by sire.
When I did these rankings I threw out the results for sires that didn't have at least 4 sons on test. Statistically the more sons representing a buck, the more accurate the results will be.
For instance, let's say I had only tested one son of my buck T1 (aka Buckaroo) instead of the 10 sons I actually tested. Had I selected his son W110R to send, then Buckaroo would have looked like the greatest sire I've ever owned, since that buck gained 0.905 lbs/day at the test and won.
However, if I had selected his son W333Y that got sick at the test and gained a minuscule 0.286 lbs/day, I might have come to a very different conclusion about T1.
Instead, I tested 10 of his sons and they gained on average a very respectable 0.679 lbs/day. With that large of a sample I feel pretty confident that T1 is an excellent sire, but I have two other sires that are even better.
To me that is the entire point of the performance tests, to identify superior genetic lines, not necessarily outstanding individuals.
I do use the individual animal results to help determine which sires to breed to which does, since my data does show that the results obtained from breeding Sire A to the daughters of Sire B can yield superior results to breeding Sire A to the daughters of Sire C.
That isn't to say that the daughters of Sire C are necessarily bad, however, because breeding them to Sire B often produces superior results.
The only way to determine which combinations work best is to try them, and then test the resulting offspring. You aren't going to learn anything sending a small percentage of your buck crop to the test. Most years I test 100% of our bucks.
Table 3 lists the results by herd. Again, I only ranked herds that had a significant number of entries to avoid including results that are statistically not very accurate or meaningful.
Finally, Table 4 lists the results by breed. I am going to do an entire column on this next month utilizing all the test data I have (which dates back to 1999), so until then...
|This page updated 11/04/07|