GENETIC EVALUATION OF LAMB PERFORMANCE IN OUTAOUAIS-ARCOTT AND SUFFOLK SHEEP, THEIR CROSSES AND CANADIAN-ARCOTT-SIRED OR HAMPSHIRE-SIRED 3-BREED CROSSES

A crossbreeding study was used to assess the growth performance of the Outaouais Arcott and Suffolk breeds, their two breed crosses and spec ific three-breed crosses sired by either Canadian Arcott or Hampshire rams. In a sheep flock maintained under a semi-confinement management system, Suffolk lambs, weighing 3.8 kg at birth, were not significantl y heavier than Canadian Arcott- or Hampshire-sired crossbred lambs. La mbs of the Outaouais breed were 26% lighter (P<0.05) than those of the Suffolk breed, and their two breed crosses had birth weights between the two parental purebreds. The relative ranks among the breeds and th eir crosses had not changed by 42 d of age with lamb weights ranging f rom 10.6 to 12.6 kg. By 120 d of age, the Canadian Arcott- and Hampshi re-shed lambs weighed approximately 28-29 kg and were not significantl y different in weight from the Outaouais and Suffolk breeds and their two breed crosses. Additive and maternal genetic effects in the Suffol k breed were significantly greater than in the Outaouais breed for bir th weight by 12.9 and 12.6%, respectively. By 42 d, no difference was detected between breeds for direct genetic effect. However, the matern al effect favoured the Suffolk breed by 15.8% (P<0.05). By 120 d, the direct genetic effect was Significantly greater in the Outaouais breed by 9.8%, but maternal genetic effects continued to favour the Suffolk breed by 12.5% (P<0.05). Heterosis effects were generally small (<3.1 %) and not significant for lamb weights. The terminal sire breeds show ed no significant difference between Canadian- and Hampshire-sired lam b weights. Generally, results obtained for two analyses (least squares model and multi-trait animal model) of the same data set were consist ent. However, the animal model would be preferable when estimating par ameters and breeding values from an unbalanced data set with unequal s ubclass frequencies.