44 



CORRELATION STUDIES ON WINTER FECUNDITY 



By F. a. Hays, Ruby Sanborn and L. L. James 



High winter egg production is very desirable from the poultry keeper's point 

 of view for two reasons: first, prices for eggs are much more remunerq,tive 

 during the winter months than at any other season; second, winter egg yield 

 is intimately correlated with annual egg yield (Hervey 1923). 



The number of eggs that a pullet will lay from first egg to March first de- 

 pends upon seven pairs of Mendelian factors as has been shown by Hays 

 (1924). This being the case, winter egg production cannot be correctly con- 

 sidered as a simple physiological character but rather as the manifestation of 

 the interaction of the characters of a complex. Such traits as sexual maturity, 

 winter pause, intensity, and broodiness have a definite and measurable relation 

 to the number of winter eggs when large numbers are considered. Each of 

 these four major traits affecting winter fecundity is unquestionably subjected 

 to and modified by varying conditions that we may call "environmental" for 

 lack of a more specific term. 



In the studies reported below an attempt has been made to measure by 

 means of the coefficient of correlation the degree of association of some measure- 

 able variables with winter egg yield. The chief value of such a study lies in the 

 fact that, knowing the relative importance of the variables considered in re- 

 lation to winter production, the breeder should be able by controlling major 

 variables such as age at first egg, hatching date, rate of growth, etc., to attain 

 higher winter averages and to secure greater uniformity in the winter fecundity 

 of his flock. For example, if the flock is mated in such a way as to secure only 

 genetically early maturing pullets, the age at first egg will range from about 

 150 to 210 days, while in a flock such as ours that is not genetically pure for 

 early maturity the age at first egg ranges from 150 to 300 days. 



Data Available. 



A total of 959 Rhode Island Red Pullets hatched in eleven weekly broods 

 from March 25 to June 3, 1923 are studied. This flock includes rather hetero- 

 genious breeding when winter fecundity is considered. Included in this flock 

 are birds bred for the following characteristics: high winter and annual fe- 

 cundity, non-broodiness and broodiness, good color, inbred and outbred, and 

 hatching power of eggs. Taken as a whole, this flock may be considered good, 

 but not equal to the standards set by those birds bred for fecundity alone. 

 The mean of 803 birds that have complete winter records is 45 eggs. 



Coefficient of Correlation 



Coefficient of correlation is here used as a measure of association or depend- 

 ence of one trait upon another. For example, in a particular flock of pullets 

 if the degree of correlation between age at first egg and winter egg record is 

 — .64, such a constant indicates that large winter records depend on early ma- 

 turity in 64 birds out of every hundred, and that sexual maturity is one very im- 

 portant factor in determining the winter record of such pullets. A -f sign be- 

 fore the coefficient .64 would mean that late sexual maturity is associated with 

 large winter egg records. Other factors, such as hatching date, 150-day 

 weight, weight at first egg, daily gain in weight, etc., may be measured in a 

 similar manner. A comparison of the coefficients of correlation for these 

 different factors furnishes a measure of their relative importance. Selection 

 based on the coefficient of correlation should be applied to flocks rather than to 

 individual birds as is evident in tables 1 and 2. 



The word mean has the same meaning as average. Standard deviation shows 

 the range of variation of a group of individuals above or below the mean or 

 average. If the mean hatching date is 6.68 and its standard deviation is 

 ■±.9,.95, the interpretation is that the average range in hatching date of the flock 



