Animal Science Department


Date of this Version



North Central Regional Publication No. 262


(c) 1980 University of Nebraska-Lincoln, The Agricultural Experiment Station Institute of Agriculture and Natursl Resources


An objective of the NC-103 regional swine breeding project is to evaluate inter-population and intra-population performance of domestic and exotic strains of swine. Several cooperating stations have conducted experiments relative to this objective. Projects were not exact replicates, but sufficient overlap allowed combining the information to yield more precise estimates of heterosis and breed effects than was possible from the analyses of data from any single experiment. This publication summarizes data available from NC-103 cooperating stations on breed and heterosis effects in swine. An extensive crossbreeding experiment has been conducted and results published by Canadian researchers Fahmy and Bernard, 1971 (8); Fahmyetal., 1971 (9); Fahmyetal., 1975 (l0); Fahmyetal., 1976(11); Holtmann et al., 1975 (13). Several of the breeds were U.S. breeds and many crosses were similar to those in projects contributing to NC-103. To provide additional information on breeds and crosses, the Canadian data were also utilized. These data provided comparisons with breeds found primarily in Canada. For a summary of the Canadian study, see the report of Fahmy and Holtmann, 1977 (12).

Data used were individual experiment breed group means obtained from scientific publications, N C-1 03 annual reports, or personal communication with project leaders. Breed group means were subjected to a weighted least squares analysis where each mean was weighted by the number of observations in the mean. For each trait, the data were grouped and several analyses were conducted so that each analysis contained the largest possible subset of the data. For example, some experiments included purebreds and crossbreds whereas others compared several crossbred combinations. Breed and heterosis effects were estimated only from experiments that included both purebreds and crossbreds. Purebreds were then deleted and breed effects were estimated from analyses utilizing all two-breed cross means. A third analysis of each trait was made utilizing only those experiments that included purebred and crossbred females mated in such a way that heterosis for maternal effects could be estimated. For sow productivity traits, a fourth analysis was conducted on all three-breed cross means (crossbred female and purebred boars) to compare the performance of various combinations of crossbred females. The same breeds were not used in each experiment. However, most studies included Yorkshire and Duroc-Yorkshire crosses. Therefore, to insure a connected design for each analysis, only experiments that included Yorkshire or Duroc-Yorkshire crosses were included. A description of the traits analyzed, the experiment stations from which data were obtained and the reference for each source of the data are presented in Table l.