TABLE 4-0. — Probability of no differences in weights of tagged 

 and \intagged pups, St. Paul Island, 1962 



[in kilograms] 



NEP 



REEF 



ZAP 



POL 



Combination ■'■ 



First weighing 

 (2-3 September) 



Males 



Females 



.OA 



<.01 

 .10 



.03 

 <.01 



<.01 

 <.01 



<.01 

 <.01 



Second weighing 

 (2-3 October): 



.12 



.07 



.10 

 .26 



.45 



<.01 

 .08 



.24 



■'■ Probability derived by combination method in George W. Snedecor's Statis- 

 tical methods, p. 271 (see footnote 11, p. 48). 



first weighing and for two of the eight at the 

 second weighing (table 39). During the third 

 weighing none of the weight differences within 

 rookeries were significant. There were sig- 

 nificant differences in weights between some 

 of the rookeries, therefore the data were ana- 

 lyzed further, using the combination of prob- 

 abilities described in Snedecor. " The results 

 are given in the righthand column of table 40. 

 Weight differences were significant for tagged 

 and untagged males and females at the first 

 weighing and for males at the second weighing. 

 The differences were not significant for fe- 

 males at the second weighing or for either sex 

 at the third weighing. 



Differences in weights of pups between 

 rookeries were fairly consistent during the 



'^George W. Snedecor. 1956. Statistical methods. 

 The Iowa State College Press, Ames, Iowa, 534 p. 



three weighings. Pups from Northeast Point 

 Rookery were consistently above average in 

 weight, while those from Polovina Rookery 

 were consistently below average. Pups from 

 Reef and Zapadni Rookeries varied in weight 

 between these extremes. Pups from a specific 

 rookery have not been consistently heavier or 

 lighter than those from other rookeries from 

 year to year. 



The three weighings in 1962 gave some 

 information on the growth rate and total 

 weight gain of pups in autumn. Table 41 lists 

 the changes in mean weights between weigh- 

 ings for the various groups. If the growth rate 

 is constant, 60 percent of the total weight gain 

 would have occurred between the first and 

 second weighing and 40 percent between the 

 second and third weighing. A tendency for a 

 higher than expected percentage of weight gain 

 to occur between the first and second weighing 

 was not significant. 



48 



