192 



NORTH AMERICAN FAUNA 74 



Table 31. Frequency of occurrence of corpora lutea of pregnancy in ovaries of 

 ''potentially est rous"^^ females taken in the Bering and Chukchi seas from No- 

 vember to August.^ 



Percent with 





No. of 



No. with corpus 



corpus luteum 



Month 



specimens 



luteum of pregnancy 



of pregnane)' 



NoN'ember 



1 



0 



0 



December 



13 



0 



0 



January 



3 



0 



0 



February 



2 



1 



50 



March 



12 



9 



75 



April 



15 



11 



73 



May 



98 



65 



66 



June 



56 



40 



71 



July 



9 



3 



33 



August 



14 



11 



79 



^^Not carrying a near- or full-term fetus (No\ember to June) and not ha\'ing born a new 

 calf (April to August). 



^Data from Gol'tsex- (1978 and unpublished data) and J. J. Burns and F. H. Fay (unpub- 

 lished data). 



alread\' had a large corpus luteum of pregnancy; by April and May, that 

 proportion had not risen, and it apparently did not change thereafter (Table 31). 



The condition of the earliest recorded corpus luteum in a female taken on 14 

 February, relative to that of corpora lutea in other pinnipeds and other 

 carnivores (Wimsatt 1963; Harrison 1963; Craig 1964; Harrison and Weir 1977), 

 indicated that o\'ulation had occurred 2 to 4 weeks earlier, in mid-January to late 

 January. The comparati\'e condition of those in females taken in March 

 suggested that ovulation had occurred at least 1 month earlier. Conversely, the 

 corpora lutea in females taken in April to June were mostiy much larger than 

 those in Februar\- to March, and their histological condition was consistent with 

 that of corpora in other pinnipeds at the time of implantation (cf. Craig 1964). 

 W^hereas some females may o\-ulate later than February. I suspect that they 

 mostly fail to be fertilized. 



Breeding Behavior 



Because the mating season of walruses previoush- was thought to be in April to 

 June, at the same time as calving (Allen 1880: Chapskii 1936; Belopol'skii 1939; 

 NikuUn 1941; Freiman 1941; Brooks 1954; Krylov 1969), most of the 

 observations thought to be rele\ ant to breeding behavior were made at that 

 time. In retrospect, it is not surprising that the observ'ers saw no remarkable 

 social organization, for most of the herds in spring are made up either of 

 subadult and adult males or of adult females with their dependent young. Mixed 

 herds of adult males and females, such as might be expected to occur commonly 

 during the mating season, are conspicuously scarce to absent at that time 

 (Nikulin 1941). 



Hanna (1923) apparently considered that the large herds found on certain 

 islands in the Bering Sea in summer were "breeding rookeries." comparable to 



