Holstein.clickHolstein dairy, from calf to bulk tank

Home / Breed & Genetics / Genetics & Selection

Record Keeping for Genomic Evaluations: Why Parentage and Performance Data Matter

A genomic test is only as good as the identification, pedigree and performance records that sit behind it.

By the Holstein.click editors · 5 min read

Genomic testing has become routine on Holstein farms. A hair or tissue sample goes to the lab, and a few weeks later a report arrives with numbers for production, health and fertility. It is easy to treat that report as magic, but it depends on ordinary farm records: who the animal is, who her parents are, when she was born, and how she and her relatives performed. This guide explains why those records matter and what to keep. For help finding genetics advisors and testing providers, see agriculture.directory. Our guide on genomic testing and net merit covers how to use the results.

What a genomic evaluation actually uses

The Council on Dairy Cattle Breeding (CDCB) explains on its genomic impact page that marker effects are estimated from a reference population of animals that have both genotypes and phenotypes. Those marker effects are then used to predict genomic breeding values for all genotyped animals. The page notes that phenotypes need to be collected on only a small, strategically chosen group of genotyped animals, and that genomic evaluations are considerably more reliable than parent average. The U.S. system succeeds, CDCB says, because of a strong foundation of performance data, quality data collection and traditional evaluations.

In plain terms: the genotype on your heifer is compared against a huge database of animals whose records are trusted. If the records feeding that database are wrong, the predictions get worse. Your records, multiplied across thousands of herds, are what make the system work.

Scale and gain

CDCB reports that about one million genotypes are submitted each year, with more than 90 percent from females, and that the national database passed ten million animals genotyped. The SNP list used in predictions was updated to 69,200 markers, gene tests and quantitative trait loci. CDCB also reports that average annual gain in Net Merit dollars rose from $40.33 in the late 2000s to $79.20 more recently, roughly doubling after genomic evaluations began in 2009, and that genomic selection has doubled annual genetic gain for production traits in U.S. Holsteins, with gains for fitness traits such as fertility, udder health and productive life higher still.

That is a major shift, and it is based on cow-level records collected by farmers.

Parentage: the record that cannot be wrong

Pedigree has always mattered for registration. In the genomic era it matters even more, because a wrong sire or dam skews the comparison. CDCB says it runs SNP-based parentage verification and discovery on every genotype it receives, following procedures certified by the International Committee for Animal Recording (ICAR), and it offers whole-genome SNP parentage verification certificates.

Holstein Association USA, in its article on parentage verification, describes how parentage is confirmed automatically during SNP testing, and that testing done through the Association sends the genotype directly to it. For tests done elsewhere, owners should ask the testing organization to send the genotype. Some situations call for additional steps, such as every third embryo transfer calf needing confirmed parentage and importing countries often requiring microsatellite certificates.

The practical lesson is simple: record the sire and dam when the calf is born, and treat genotyping as a check, not a replacement. If the test says the recorded sire is wrong, correct the record. That correction improves the evaluation of the heifer and her relatives.

Pedigree also helps with haplotypes

The Association's genomic testing page notes that pedigree information is needed to confirm haplotype status, since carrier status often cannot be determined from genotype markers alone. It lists fertility-impacting haplotypes HH1 through HH6 and encourages planning matings to avoid carrier-to-carrier pairings. That planning depends on knowing who is related to whom.

Performance data: what to keep

Parentage tells the system who the animal is. Performance records tell it how she and her relatives did. The page from Holstein Association USA notes that CDCB fees depend on how much and what type of data a herd contributes, and herds that supply more quality data qualify for lower fees. Whether or not fees are your motivation, the following records are the foundation:

  • Identification: a permanent, unique ID that is never reused, plus matching tags and electronic ID.
  • Birth records: date of birth, dam, sire, and whether it was a single or twin birth, along with ease of calving.
  • Breeding events: dates, sire used and result, including pregnancy checks and losses.
  • Calvings: dates, calf sex, calf ID and any difficulties.
  • Milk recording: regular DHI tests for milk, components and somatic cell count. Our DHI guide explains the reports.
  • Health events: mastitis, lameness, metritis, ketosis, displaced abomasum, with dates. Health data is the least standardized, and it is increasingly valuable.
  • Exits: date and reason for every sale, cull or death.

Common record problems

  1. Reused or duplicated IDs. The same number given to two cows over the years makes history unreliable.
  2. Missing or guessed sires. Natural service or multiple AI sires can leave doubt. A genotype can sort it out.
  3. Late entries. Events written down days later from memory are often wrong.
  4. Inconsistent health terms. If one employee records sore feet and another records lameness or foot rot, the data cannot be compared.
  5. Unlinked systems. Records kept in a notebook, a software package and a DHI file may disagree.

The same issue appears internationally. In a study of Slovak Holsteins, researchers worked from raw veterinary records and had to filter them for usability before estimating genetic parameters. We summarize it in benchmarking health traits genetically.

A routine that works

Pick one person responsible for data. Record events the same day, using a short list of agreed terms. Review the data monthly with your DHI supervisor or veterinarian. When a genomic report comes back, compare parentage results with your records and fix any conflicts. When you choose heifers to keep and cows to breed to beef, make the decision from clean information, as discussed in the breeding system guide.

For more on selection and breeding concepts, see farm.fm's page on artificial selection. Costs, test options and requirements change, so confirm them with Holstein Association USA, CDCB or your genetics provider before ordering.

More from this topic: Breed & Genetics → Genetics & Selection