100 



u. 



w 



J 60 



350 - 

 O 



>■ BOTH ISLANDS 



/ 



^2 UNTAGGED AND UNMARKED 



IB MARKED ONLY 



E3 TAGGED AND CHECKMARKED 



62 63 64 65 



Figure 12. — Counts of dead pups, Pribllof Islands, Alaska, 

 1950-51 and 1953-65. 



and marking program on St. Paul Island in 

 1965. Untagged pups v^ith this mark weighed 

 more than tagged and marked pups but less 

 than untagged and unmarked pups on two of 

 the four rookeries where pups were weighed 

 (fig. 13 and app. table 22). On Zapadni Reef 

 and Polovina rookeries, untagged pups bearing 

 the first digit mark were lighter than tagged 

 and marked or untagged and unmarked pups. 

 An analysis of variance test adjusted for dis- 

 proportionate subclass sizes showed that the 

 interaction for rookeries versus type of mark 

 was significant (P<0.01). The main effects 

 were not tested because of the significant 

 interaction. On all rookeries, untagged and 

 unmarked seals were heavier than tagged and 

 marked seals or seals with marks only, and 

 males were heavier than females. 



The use of scales that were too sensitive 

 to the movement of pups during weighing has 

 been a constant problem since the weighing 

 began. In 1957 and 1958, a metal cone for 

 holding the pups was attached to a dial spring 

 scale suspended from a wooden tripod. The 

 spring scale, however, was too sensitive to 

 movement of the pups, and the cone caused 

 the pups to move considerably in their attempts 

 to escape. Both problems were partially solved 

 in 1959 by putting the pups in burlap bags 

 sewed to construction-steel hoops and placing 

 the bag containing the pup on a platform scale. 

 Though the platform scale had no dampeners, 

 it was less sensitive than the spring scale to 

 movement of the pups, and the pups were less 

 inclined to struggle when confined in the bag. 

 The bags, however, changed in weight as they 

 dried or absorbed rainwater. Frequent weigh- 

 ing of the bags and corresponding adjustments 

 of the scale eliminated much of the error from 

 this source. In 1963 and 1964, the same plat- 

 form scale was used, but the pups were placed 

 in 20-gallon plastic garbage cans. The pups 

 tended to move more when in this container 



VOSTOCHNI AND 

 MORJOVI 



REEF POLOVINA 



ROOKERIES 



Figure 13. — Mean weights of seal pups about 1 Septennber, 

 St. Paul Island, 1965. 



than when confined in the burlap bags. The 

 method of holding the pups, therefore, was 

 changed in 1965 and a new platform scale, 

 with dampeners, was used. One man stood 

 on the scale and held each pup during weigh- 

 ing, a method that effectively eliminated move- 

 ment of the pups. The weight of the man was 

 recorded after each series of 25 weighings, 

 for later subtraction. 



Barricades for holding small groups of pups 

 for weighing have been used each year, and 

 weighed pups have been released 50-75 feet 

 from the site of weighing so that they would 

 not be weighed twice. Weights were recorded 

 to the nearest 0.2 kg. from 1957-64 and to 

 the nearest 0.5 kg. in 1965. 



Population Estimates 



This section presents estimates of the popu- 

 lation based on the most recent data. Methods 

 used and sources of data have been described 

 by: Roppel, Johnson, Bauer, Chapman, and 

 Wilke (1963); Roppel, Johnson, and Chapman 

 (1965); and Roppel, Johnson, Anas, and 

 Chapman (1965). 



Estimates based on tag recoveries .-- The 

 number of pups born, as estimated from tag 

 recoveries (includes tagged and lost-tag seals) 

 from both sexes, are presented in tables 10 

 through 14. To show within- season variability, 

 the data were divided into 5-day periods and 

 estimates were based on recoveries from 3- 

 and 4-year-old males for each period (table 

 12), No apparent reason exists for the much 

 greater variability in estimates based on re- 

 coveries from 4-year-old males. Estimates 

 based on recoveries from females show con- 

 siderable year-to-year variation and generally 

 are much lower than estinnates based on re- 

 coveries from males. For these reasons the 

 data for the sexes have not been combined. 



15 



