10 MASS. EXPERIMENT STATION BULLETIN 364 



6. Body Weight on March 1 and Persistency 



Poultrymen generally attach much importance to the weight of birds at the 

 close of the winter laying season. In this connection it is desirable to know whether 

 large birds are likely to be more persistent. The coefficient of correlation between 

 body weight on March 1 and annual persistency will supply some information 

 on this point. 



Number of birds 1338 



Mean weight March 1, pounds 6.62 



Weight standard deviation + . 75 



Mean annual persistency, days 334.01 



Persistency standard deviation ±41 .56 



Coefficient of correlation ' - .0905 ± .0183 



Correlation ratio . 1655 



Range in body weight was wide, extending from 4 to 9.5 pounds. The co- 

 efficient of variation in weight was somewhat above 11 percent. A mean of 

 6.62 pounds indicates that the birds were large at maturity. 



Mean persistency was 334 days and the range was from 82 to 366 days. The 

 value of the mean indicates that the majority of the birds were highly persistent 

 layers. 



The coefficient of correlation is of negative order and is statistically significant. 

 The regression of persistency on weight was found to be nonlinear so that the 

 correlation ratio, .1655, may be used to measure association. The magnitude 

 of this constant is sufficient to indicate a slight negative relation between adult 

 weight and persistency of laying. These data are suggestive that heavier birds 

 at maturity may not be as persistent as birds of somewhat lighter weight. 



Relation between Body Weight and Mortality 



7. Body Weight at Housing, and Mortality 



Since the pullets were all housed at about 170 days of age, the relation between 

 body weight at housing and subsequent mortality may be considered. In Table 4 

 the birds are grouped into classes with respect to weight at housing, and mortality 

 in each weight class is recorded for a 365-day period. The data cover the flocks 

 hatched in 1934, 1935, and 1936 in all experiments, and indicate that pullets 

 that weighed less than 4.25 pounds when placed in the laying houses had a sig- 

 nificantly higher mortality rate (32.46 percent) than birds having greater weight 

 when housed (25.87 percent). 



Mean mortality rates for the different weight classes were fitted to a straight 

 line, omitting four birds weighing less than 3.5 pounds and eight birds weighing 

 more than 6.75 pounds. These data showed an essentially normal frequency 

 distribution with respect to weight. The standard error of estimate was cal- 

 culated for mortality rate, and a range covering twice the value of the standard 

 error of estimate was found to include all mortality means. The regression of 

 mortality rate on weight at housing was strictly linear, and mortality rate in the 

 laying houses may be expected to decrease from about 30 percent for the smaller 

 birds to about 22 percent for the larger birds. This is not an intimate relation 

 between weight at housing and subsequent mortality, yet it is significant. 



