Environmental effects may be behind the wide difference in the performance of your first-calf heifers, but don’t discount genetics.
If you’ve ever taken the time to chart the performance of your first-calf heifers, you’ll see a wide difference in performance. Even in the best managed herds with 30,000 lb. herd averages, one heifer will produce 36,000 lb. of mature equivalent milk while a heifer lying in the next freestall is producing 20,000.
Roughly 70% of the variance is due to environmental effects: A health event early in the calf’s life—pneumonia, scours, whatever—can affect how she performs in the parlor later in life.
But don’t discount genetics, says Pat Hoffman. Hoffman was a heifer specialist with the University of Wisconsin for years, and is now a Dairy Technical Specialist with Vita Plus Corporation.
Hoffman explains the other sources of performance variation are both genetic: parent average and Mendelian sampling effect. (Each contributes about half of the genetic variance.) Parent average is the average genetic value of the sire and cow. The better the parents, the better the average.
The Mendelian sampling effect is where that German priest comes in. Father Gregor Mendel, through experiments with pea plants in the mid-1800s, discovered how genes sort themselves at fertilization. When a sperm fertilizes an egg, the genes of each are recombined. On average, that recombination will be the average of the parents’ genetics. But any one individual can get any combination. Over a population, the recombination occurs on a bell curve with most getting the average and just a few on each tail of the curve getting the best or poorest combinations of genes. (Note: A.I. bull studs rely on Mendelian sampling to find the best next generation of sires. And that’s how and why they only bring back one in 10 bulls to active, proven line-ups.)
In practical terms, roughly 20% of calves get the best combination of genes and another 20% get the poorer combination of genes. It might seem counterintuitive, but as herd averages rise, the Mendelian effect takes on greater value, says Hoffman.
When herd averages were just 10,000 lb./cow, a 20% variance was +/- 2,000 of milk. At a 20,000 lb./cow average, the variance is 4,000 lb. And at a 30,000 lb./cow average, the variance is 6,000 lb. So the 24,000 lb./cow versus 36,000 lb./cow difference is actually to be expected, says Hoffman.
"There are numerous genetic research papers on the Mendelian effects on milk production and estimates are different within each research paper. But random genetic effects account for more than 5,000 lb. of milk in first lactation," he says. "At $20 milk, this variance is of high economic value."
"In fact, this Mendelian sampling effect has more value than ever before. And it may be growing [as we improve a herd’s genetic level] and not shrinking."
What you do about this is another matter:
Scenario 1. If you need every heifer as a replacement, your only decision is what you breed your heifers to. Low genetic value heifers merit cheap semen or beef semen (to produce dairy beef).
Scenario 2. If you don’t need all the heifers for replacement, the decision gets trickier because it now costs $2,000 (give or take) to raise a heifer to freshening.
The key is knowing which are the genetically poor heifers. If you have good records, parent average is probably good enough for Scenario 1. If you’re actually going to cull animals, parent average may or may not be.
Hoffman points out that parent average is dependent on the accuracy of records—knowing who the sire and dam are. Plus the reliability of Parent Average for production traits in a commercial herd is roughly 20 to 25%, says geneticists. Hoffman prefers genomic testing because it has higher reliability (65 to 70%) and will correctly identify parents.
But geneticists point out that a tripling of reliability does not equate to a tripling in accuracy. To get to the issue of accuracy, you have to take the square root of the reliability. So the accuracy of Parent Average for production is about .5. The accuracy of genomics is .84. Consequently, genomics is more accurate, but not 3X more accurate.
And then there’s the issue of cost. Genomic 6K or 9K chips cost in the range of $50. But to find the bottom 20% without using parent average, you have to test everybody. So the $50 test becomes $250. Perhaps that makes sense in Scenario 2 if you’re trying to save $2,000/head in rearing costs. It probably doesn’t in Scenario 1—unless you’re a registered breeder also trying to find the very top-end heifers for flushing and sexed semen treatment.
You can read Hoffman’s entire presentation on heifer performance variation here.
For more on genomic testing: