Table l. — Annual data on elephant seal numbers, shark attacks, water temperatures, and frequency of attacks relative 



to seal numbers at the Farallon Islands, central California. 



' Ttiought to be an attack on a sea lion, 



'Cows arrived witti fresh shark-bite wounds; in 1979 the remains of a cow, likely a shark victim, was seen floating in the water. 



and Nash 1978), it was worthwhile to consider 

 the relationship between temperature and sharks 

 at the Farallones. As shown in Table 1, water 

 temperature during the fall when most shark 

 attacks and sightings occurred, compared among 

 the years, fluctuated up and down but did not 

 relate clearly to yearly fluctuation in shark attack 

 frequency. The same was true for temperatures 

 during the winter elephant seal breeding season. 

 The year 1978 provides an instructive example 

 of this. In the spring-summer, temperatures were 

 unusually high. In spite of record numbers of 

 elephant seals only one shark attack was ob- 

 served, and that occurred in July, long after the 

 seal population peak (only 24 were present; 452 

 sea lions, though, were present). Later in the fall, 

 temperatures were lower than during spring but 

 much shark activity was evident (Table 1). The 

 only major relationship that was evident be- 

 tween shark attacks and water temperature was 

 that all but one observed attack occurred when 

 temperature generally exceeded 12° C (as sum- 

 marized in Figure 2). Unfortunately, there is no 

 published information of the seasonality of white 

 sharks to explain this. Off Durban, South Africa, 

 where water temperatures are higher than at the 

 Farallones, Bass (1978) found small white sharks 

 more abundant when temperatures dropped to the 

 annual low (equivalent to highest Farallon tem- 

 peratures), but the number of white sharks the 

 size of those at the Farallones did not change. 



The factor that best relates to the general 

 increase in white sharks is an increase in elephant 

 seal numbers. The fall, winter, and spring popula- 

 tions of elephant seals have been increasingly 

 rapidly at the Farallones over the past several 

 years (see Le Boeuf et al. 1974; Ainley et al. 

 footnote 3; Table 1). The seal population during 



the fall, the period of most shark attacks, has 

 increased about 3.9 fold since 1972, the first year 

 of the period when shark attacks have been seen 

 consistently year after year. In the fall data there 

 is a direct relationship between the number of 

 shark attacks and the number of seals (r = 0.895; 

 P<0.01). The ratio of attacks to the number of 

 seals, except during spring and the 1979 seal 

 breeding season, has remained at about 0.01-0.02. 

 Shark attacks during the elephant seal breed- 

 ing season (winter) have been observed less often 

 than during the fall but they may be increasing 

 during that period, too, if the 1977-79 seasons are 

 any indication. Interestingly, attacks were first 

 seen during winter in the year when the elephant 

 seal population surpassed 120 animals (1978), the 

 same population level that occurred in conjunc- 

 tion with the first fall sighting of elephant seal/ 

 shark interactions (1972). This further indicates 

 a density-dependent relationship between shark 

 predation and elephant seal populations. 



Acknowledgments 



The Farallon Island field station of the Point 

 Reyes Bird Observatory was supported financially 

 by members and donors and by contracts from 

 various agencies. In addition to our members, 

 we wish especially to thank the Lucius M. 

 Beebe Foundation, Charles E. Merrill Trust, 

 McNaughton Foundation, Dean Witter Founda- 

 tion, San Francisco Foundation, Packard Founda- 

 tion, Standard Oil of California, and the U.S. Fish 

 and Wildlife Service. Marine mammal observa- 

 tions from 1975 to 1979 were under contract from 

 the Marine Mammal Commission. We also wish to 

 thank the U.S. Coast Guard and particularly the 

 Oceanic Society for logistic support. William I. 



944 



