Potential selection response in composite and parental sheep breeds

Theoretical expectations and experimental evidence were examined to determine the relative magnitude of additive genetic variance in composites compared with parental breeds. A single-locus two-allele, additive and dominance effects model was used to clarify expectations from varying (i) degree of dominance (d), (ii) variance among mean gene frequencies per locus of parental lines (Vq) and (iii) mean gene frequency of the dominant allele in the composite (q$\sb{\rm c}$). Results showed that expected heterosis retained in a composite ($\Delta$H%) is proportional to d and to Vq and is maximum when q$\sb{\rm c}$ is intermediate. Ratio of additive genetic variance between composite and parental averages ($\Delta$Va$\sb{\rm c}$%) increases the most with Vq when d = 0, but changes with higher degrees of d from $\u3e$100% when q$\sb{\rm c}$ $\u3c$.5 to $\u3c$100% when q$\sb{\rm c} \u3e$.7. Thus, at any level of $\Delta$H%, $\Delta$Va$\sb{\rm c}$% is always less for higher d. Heritability (h$\sp2$) for ewe performance in two composites were compared with their constituent purebreds: Comp 1 = 1/2 Finn 1/4 Rambouillet 1/4 Dorset under accelerated and Comp 2 = 1/2 Finn 1/4 Suffolk 1/4 Targhee under annual lambing. Data included 17868 records of 6177 ewes and 828 sires. Estimates of h$\sp2$ from sire components and repeatability for individual records were higher in composites than parent breeds for size and weight of litters at birth and were lower in composites for fertility and weaned lamb output per ewe lambing or exposed especially under accelerated lambing. However, estimated h$\sp2$ from regression of ewe-mean records of daughter on dams were higher for composites than parent breeds under both accelerated and annual environments in output per ewe lambing and per ewe exposed. Weighted mean differences in h$\sp2$ for composites vs parent breeds were slightly positive for size and weight of litters at birth but mostly negative for number and weight weaned per ewe exposed. Thus expected greater response from selection in composites may depend primarily on greater selection intensity from heterosis in viability of lambs