D. GENE-ENVIRONMENTAL INTERACTIONS 



The fourth ohjcctive was to study gene-environmental interactions. 



Poultry breeders have long suspected that strains hrcd in one geo- 

 graphical location or under specific conditions would not perform 

 equally well in other locations or under other conditions. The use of 

 randombrcd populations as controls at the different State experiment 

 stations in the region provided an opportunity to evaluate gene-environ- 

 mental interactions. 



Nutritional Environment 



Single Comb White Leghorns at the Pennsylvania Agricultural Ex- 

 periment Station were fed two different layer-breeder diets. The rations 

 were isocaloric but the level of nutrients in one was approximately 80% 

 of that of the other (23). The data were analyzed for survivors egg 

 number for 10 months. The dominance and maternal variance as a pro- 

 portion of the additive genetic variance was more than 10 times as high 

 for the high ration as for the low ration. The sire component estimate 

 of heritability was 10% for the high ration and 28% for the low. These 

 results indicate that selection for total egg production on the low diet 

 would be superior to high ration selection regardless of the ration used 

 for the progeny. 



Raising Sexes Separately 



Two trials using chicks from a commercial broiler cross were 

 studied at the New Hampshire Agricultural Experiment Station (29) . 



Contrary to some reports, raising sexes separately did not reduce 

 the variation in 9-week body weight below that observed for each sex 

 when both sexes were brooded together. The coefficients of variation 

 were larger for males brooded separately than for those brooded mixed, 

 but for females the relationship was reversed. 



Randombred Populations 



The first 6 years of data collected from the Cornell randombred 

 control population were analyzed for nine different traits (27) . Year 

 to year variability was found but no trends were evident. Estimates of 

 heritability from this and other studies in NE-6 are given in Table 1. 

 Maternal effects accounted for 13% of the total variance in egg produc- 

 tion. Maternal effects also were important for 8, 32 and 55-week body 

 weight and for sexual maturity. 



In another analysis Cornell and USDA workers used data for eight 

 economic traits from the Cornell randombred control populations main- 

 tained at Ithaca, New York and Lafayette, Indiana (56). Most estimates 

 of the variation due to genetic-environmental interaction were small. 

 The three largest estimates were 9% for egg production, 8% for 8-week 

 body weight and 6% for 32-week egg weight. As in the previous study, 

 the variation due to maternal effects was relatively large for sexual 

 maturity and egg production. The sire effects were small. These results 

 suggest the need for a reappraisal of breeding programs for these traits 

 with more investigations on methods of capitalizing on the maternal 

 effects. 



19 



