Skip to main content
ARS Home » Research » Publications at this Location » Publication #121512

Title: HETEROSIS AND BREED DIFFERENCES FOR YIELD AND SOMATIC CELL SCORES OF US DAIRY CATTLE IN THE 1990S

Author
item Vanraden, Paul

Submitted to: American Dairy Science Association Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 7/24/2001
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Yields and somatic cell scores (SCS) of 8806 F1 crossbreds and 81,409 purebreds were examined using test-day data from 320 US herds that each had >=10 crossbred cows born since 1990. Comparisons were made within herd, year, and 3-mo calving season, although some groups contained only purebreds or only crossbreds. The model included effects of age and lactation stage, parent breeds, regression on sire predicted transmitting ability, and general heterosis. Estimates of heterosis for first parity were 4.6% for milk yield, 5.3% for fat yield, 4.9% for protein yield, and 0.6% (unfavorable) for SCS. Corresponding estimates for all parities were 4.7, 5.0, 4.9, and 0.3%. A model without an adjustment for sire merit produced slightly lower estimates. For Brown Swiss x Holstein crossbreds, daily fat yield (1.08 kg) and protein yield (0.88 kg) equaled yields for Holsteins, but milk yield was slightly lower (27.6 versus 28.9 kg). For Jersey x Holstein crossbreds, fat yield was equal to Holsteins, but protein yield (0.85 kg) and milk yield (25.7 kg) were lower. With cheese yield pricing, crosses of Brown Swiss or Jerseys with Holsteins may be more profitable than purebred Holsteins if heterosis for health, fertility, and mortality traits are considered. Yields for all other breeds and breed crosses were lower than yields for purebred Holsteins, but feed costs are lower for breeds with smaller body size. Although genetic progress is slower within the less numerous breeds, increasing breed differences, current heterosis for yield traits is sufficient to make some dairy breed crosses worth considering. A combined national evaluation of data for all breeds and crossbreds may be desirable but would require an extensive programming effort.