Survival, lifetime production, and profitability of Normande × Holstein, Montbéliarde × Holstein, and Scandinavian Red × Holstein crossbreds versus pure Holsteins

The objective was to describe the dynamics of culling risk with disposal codes for Holstein dairy cows reported by herds enrolled in the Dairy Herd Improvement program. Dairy producers could report 1 of 9 possible disposal codes or forego reporting a code. After edits, 3,629,002 lactation records were available for cows calving between 2001 and 2006 in 2,054 herds located in 38 states primarily east of the Mississippi river. The distribution of culled cows by disposal code was estimated by parity, days after calving, pregnancy status, cow-relative 305-d mature equivalent milk yield, herd-relative 305-d mature equivalent milk yield, and season. Of all herds, 57% reported all 8 different disposal codes excluding the codes dairy purposes and reason not reported. Hazard (risk) functions were calculated by parity, from 1 to 520 d since calving for open cows and from 1 to 280 d since conception for pregnant cows. Annualized live culling rate and death rate (reported code was death) were 25.1 and 6.6%, respectively. The primary disposal code was died (20.6% of all culling), followed by reproduction (17.7%), injury/other (14.3%), and low production and mastitis (both 12.1%). The risk of culling with various disposal codes varied with stage of lactation. Died and reproduction were the most frequently reported codes for cows leaving the herd during early and late lactation, respectively. Early lactation was also a critical period for culling with the disposal codes injury/other and disease, and the risk increased with days after calving for the codes low production and reproduction. The risk of culling with the disposal code died showed the greatest seasonal pattern with increased risk of death in spring and summer. A negative association was found between annualized live culling and death rates within herds. Compared with open cows, pregnant cows had a lower risk of culling with all reported disposal codes. In addition, the risk of culling was lower in high-producing cows with all disposal codes. In conclusion, the risk for culling by disposal code varied by parity, stage of lactation, season, pregnancy status, and milk yield.

Heterosis and breed differences were estimated for milk yield traits, somatic cell score (SCS), and productive life (PL), a measure of longevity. Yield trait data were from 10,442 crossbreds and 140,421 purebreds born since 1990 in 572 herds. Productive life data were from 41,131 crossbred cows and 726,344 purebreds born from 1960 through 1991. The model for test-day yields and SCS included effects of herd-year-season, age, lactation stage, regression on sire's predicted transmitting ability, additive breed effects, heterosis, and recombination. The model for PL included herd-year-season, breed effects, and general heterosis. All effects were assumed to be additive, but estimates of heterosis were converted to a percentage of the parent breed average for reporting. Estimates of general heterosis were 3.4% for milk yield, 4.4% for fat yield, and 4.1% for protein yield. A coefficient of general recombination was derived for multiple-breed crosses, but recombination effects were not well estimated and small gains, not losses, were observed for yield traits in later generations. Heterosis for SCS was not significant. Estimated heterosis for PL was 1.2% of mean productive life and remained constant across the range of birth years. Protein yield of Brown Swiss x Holstein crossbreds (0.94 kg/d) equaled protein yield of purebred Holsteins. Fat yields of Jersey x Holstein and Brown Swiss x Holstein crossbreds (1.14 and 1.13 kg/d, respectively) slightly exceeded that of Holsteins (1.12 kg/d). With cheese yield pricing and with all traits considered, profit from these crosses exceeded that of Holsteins for matings at breed bases. For elite matings, Holsteins were favored because the range of evaluations is smaller and genetic progress is slower in breeds other than Holstein, in part because fewer bulls are sampled. A combined national evaluation of data for all breeds and crossbreds may be desirable but would require an extensive programming effort. Animals should receive credit for heterosis when considered as mates for another breed.