| An optimization method was developed for sire selection based on net return and risk of genetic merit of sires. Expected income (net present value) of a sire realized through the offspring was proposed as the composite criterion for selection. This considered the revenue corresponding to his predicted transmitting ability for milk, fat, and protein yields, cost of dystocia corresponding to his expected progeny difference for dystocia, and other fixed and variable costs. Variance of income (risk) of a sire was a function of his reliability estimates. Expected income-Variance frontier developed for a pool of sires based on quadratic programming provided the minimum risk combination of sires for an intended expected income. The combination of sires that maximizes the 95% lower confidence boundary of the frontier (that maximizes 95% guaranteed future income) determined the optimum set of sires to be selected.; The optimization method explained above was applied to a simulated pool of young (pedigree tested) and proven (progeny tested) sires to determine the optimum proportion of young sire use. Young sires were riskier and, on the average, had higher expected income (low semen cost) than proven sires. The representation of young sires in the optimum set of sires that maximizes 95% guaranteed income determined the optimum proportion of young sire use. The optimum proportion of young sires was 34 percent for the pool that simulated the current Holstein population in the United States. The proposed method can be used to define the best set of sires where reliability estimates are included in the selection criteria in addition to predicted breeding values.; Long term inbreeding, genetic and economic gains associated with cloning were estimated. Random and rotational mating systems for full-sib clones were considered. Production of more than 50 clones could keep inbreeding coefficient below 5 percent for 10 generations. Break-even costs per clone for modern progeny testing schemes were {dollar}83 and {dollar}41 per clone with one and five clones produced per dam, respectively. Technology of cloning and infrastructure enhancements should be developed further to lower the cost of cloning below the break-even levels for commercial use of cloning to be economically viable.; Genetic and phenotypic (co)variance components were estimated from a multiple trait animal model for 305-day milk, fat, and protein yields, days open, number of services and percent cow mortality during lactation for Holsteins. Restricted maximum likelihood estimates based on expectation-maximization algorithm for the traits were similar to those from previous literature. Genetic and phenotypic relationships between yield and fertility were antagonistic. Genetic correlations between yield traits and cow mortality were unfavorable but phenotypic correlations were favorable. Evidently, modern management practices provide better management for better cows resulting in reduced mortality of high producing cows with poor genetic potential in terms of survival ability. |
