The effect of alternative economic and genetic covariation structures on the relative economic gain from selection using stayability traits.

Three alternative cost ratios and the ratio of annual fixed cost for cows to net replacement cost for three populations that differed in genetic (co)variances were investigated. We studied their effect on the value of longevity records to supplement production records to achieve genetic gains for reducing the nonfeed cost of producing a unit of milk. Data files were constructed from stochastically generated production records and culling variant records; the later records represented the relative value for cow survival of all nonproduction traits. The use of one stayability trait instead of none increased gain by 1 to 9% across the populations and cost ratios. Only 2% additional gain occurred with two stayability traits instead of one. Populations that differed in genetic correlation between production and the survival value of nonproduction traits and in survival heritabilities contributed 2 to 5% toward the greater gains. The different cost ratios contributed 1 to 3%. The effect of increasing the length of herd life independent of production is inherently secondary to the effect of increasing the production during lactation. If genetic antagonism between production and the survival of nonproduction traits is found and if the net replacement costs are high, the use of production and longevity records for selection can enhance gain by up to 10%.

Genetic value of herd life adjusted for milk production.

Cow herd life adjusted for lactational milk production was investigated as a genetic trait in the breeding objective. Under a simple model, the relative economic weight of milk to adjusted herd life on a per genetic standard deviation basis was equal to CVY/dCVL where CVY and CVL are the genetic coefficients of variation of milk production and adjusted herd life, respectively, and d is the depreciation per year per cow divided by the total fixed costs per year per cow. The relative economic value of milk to adjusted herd life at the prices and parameters for North America was about 3.2. An increase of 100-kg milk was equivalent to 2.2 mo of adjusted herd life. Three to 7% lower economic gain is expected when only improved milk production is sought compared with a breeding objective that included both production and adjusted herd life for relative value changed +/- 20%. A favorable economic gain to cost ratio probably exists for herd life used as a genetic trait to supplement milk in the breeding objective. Cow survival records are inexpensive, and herd life evaluations from such records may not extend the generation interval when such an evaluation is used in bull sire selection.

Relationships of first lactation milk and type traits to cow survival and a sire index for discounted total milk.

The probability of cow survival during each of six consecutive 13-mo periods after first calving was regressed on lactation milk yield and 15 type traits, each recorded during the first lactation. Adjusted coefficients of determination were small (.028 to .081). The ratio of variance in predicted cow survival probabilities due to milk only with that due to the 15 type traits was 12.6:1, 4.7:1, 2.5:1, .5:1, 0:1, and .02:1 for respective herdlife periods. Maximum effect of type traits on herdlife ranged from 11 to 109 d. Prediction equations for cow survival were incorporated into a sire index for discounted total milk yield of progeny. Type traits affected the sire index by their influence on the six predicted survival rates. Milk yield affected the discounted total milk index directly as well as through the survival rates. Simulated variation from milk estimated transmitting abilities for sires accounted for 99.9% of the total variation in the discounted total milk index. Since the effect of the type traits on cow survival rates was greatest during the later herdlife periods, too few animals would have survived to these later periods to reveal any significant effect from type on the discounted total milk index.

Association of bovine lymphocyte antigens with milk fat percentage differences.

A group of 246 Holstein artificial insemination bulls was examined for associations of bovine lymphocyte antigens with differences in transmitting ability for milk and fat yield and fat percent. The mean fat percentage associated with the w10 allele was .15% greater than the mean for four other alleles. No differences in phenotypic or allelic effects were detected for milk or fat yield.

Genetic parameters and environmental factors associated with type traits scored on an ordered scale during first lactation.

Sixty subjective and linearized type data were recorded on 7375 first lactation Holstein cows. Data were analyzed by paternal half-sister analysis for genetic parameters. One hundred and six sires were represented. Two kinds of heritability estimates were obtained; the first was a method 3-derived estimate for the raw scale and the second was a "nonlinear method 3"-derived estimate for an underlying scale. Whenever estimates of heritability on the raw scale were positive, larger estimates were generally obtained on the underlying scale. Some udder traits (e.g., depth, rear length, rear balance, teat placement, rear teat length, rear width) had underlying heritabilities between 10 to 30%. Suspensory strength and teat shape had no genetic variation. Of feet and leg traits only set (.14) and cleanness of hocks (.09) had heritabilities above 6%. Some functional traits (milking speed, edema) had heritabilities near 11%. Disposition, mastitis score, and milk letdown had no genetic variation. Most body traits (e.g., pin and hip width, length and levelness of rump, rib spring) had heritabilities between 15 and 30%. Method 3 phenotypic and genetic correlations relations were computed for the raw scale. Most phenotypic correlations were near zero. Traits associated with size or skeletal dimension had positive phenotypic typic and genetic correlations. Fore udder length, rear udder width, and rear udder height had positive genetic correlations (range .21 to .43). Udder depth had moderate genetic correlations with most udder traits, teat traits and body traits. The statistical models included describer, stage of lactation, and calendar month of description effects. In general, describer effects were the largest and most significant kind of environmental effect. Stage of lactation effects occasionally showed consistent trends. Month effects appeared to be unpredictable.

Mate selection by selection index theory.

Mate selection by selection index prediction of total merit in expected progeny is proposed as a rational basis for making recommendations in the choice of a bull to which a cow may be mated. Growth in USA of service programs recommending bulls to particular cows has motivated the need to rationalize mate selection processes. This paper illustrates that mate selection on the basis of highest index value for expected progeny among potential mates can justify special mate selection programs, when a nonlinear relationship exists between at least one trait in the index and merit.

Heritability of a linear combination of traits.

The heritability, h I (2) , of a linear combination of phenotypes, I, when defined as the ratio of the variance of the genetic index, σ g⋆ (2) *, to the variance of the index, σ I (2) , is shown to be different from the square of the correlation, r HI (2) , between the index and an arbitrary linear combination of genetic effects, H. The gain in H fror selection on I is shown to be proportional to h I (2) . bHg⋆ for any index.