disturbances at =£100 m more than at distances 

 101-200 m and 201-300 m, and were least reactive 

 to disturbances at 201-300 m (Table 6). 



TABLE 6. — Weights associated with 

 the distance/seal response interaction 

 term [12] of the log-linear model that 

 fits the data in Table 5. Relative mag- 

 nitude of weight indicates importance 

 of the variable. Sign of the weight in- 

 dicates direction of effect ( + is more, 

 — is less). 



'Seals were most reactive to disturbance 

 2 Seals were least reactive to disturbance. 



After actual disturbances (type I), the number of 

 seals that eventually rehauled was always lower 

 than the original number. On KI, the average time 

 it took seals to rehaul regardless of season, was 28 

 ± 20.8 min (range 5-100, n = 187). In 96 instances, 

 no seals rehauled, primarily due to tidal height. 

 During rising tides, they rehauled only 16.2% of 

 the time (n = 37 disturbances), at low slack tide, 

 55.6% of the time (n = 124), and on falling tides 

 61.5% of the time (n = 26; x 2 = 13.82, P < 0.001). 



Disturbances were of short duration and seals 

 rehauled after the disturbance source had left the 

 area, except for disturbance from commercial bait 

 harvesters, who remained in the vicinity for entire 

 low tide cycles. Bait harvesters likely prevented 

 seals from hauling out at all (zero-seal distur- 

 bance). During December, January, and February, 

 we recorded the presence of the harvesters on 13 d. 

 After being disturbed, seals did not return to the 

 haul out site on eight of those days. They were 

 disturbed briefly by the harvesters and then re- 

 hauled on 3 d, and there was no change in seal 

 numbers on 2 d. 



PWI apparently was not an important alterna- 

 tive site when KI was disturbed. A weak correla- 

 tion between seal numbers at PWI after they were 

 disturbed from KI existed during the winter (r = 

 - 0.42, n = 7) and summer (r = - 0.40, n = 152), 

 but not during the breeding season (r = - 0.14, n = 

 123). In 45 of these instances, however, (winter 3, 

 breeding 14, and summer 28), disturbances oc- 

 curred simultaneously at KI and PWI, thereby 

 precluding seal movement to PWI. During field 

 observations, the movement from KI to PWI 

 after disturbance was actually observed on 11 oc- 

 casions. 



FISHERY BULLETIN: VOL. 82, NO. 3 



DISCUSSION 



The population of harbor seals at Bolinas La- 

 goon is much higher than previously recognized, 

 and in contrast to seasonal peaks during the 

 breeding season at other seal haul out sites 

 (Fancher 1979; Johnson and Jeffries 1977 9 ; 

 Loughlin 1978; and Allen and Huber 1983 10 ), the 

 peak at Bolinas Lagoon occurred during summer 

 after the pupping season. The peak at Bolinas may 

 be caused in part by an influx of seals, possibly 

 from San Francisco Bay, only 24 km south, or from 

 Double Point, 10 km north, where numbers decline 

 after the pupping season (Risebrough et al. 1978 11 ; 

 Allen and Huber 1983 footnote 9). The summer 

 increase also coincides with a marked increase in 

 fish abundance in Bolinas Lagoon and Bolinas 

 Bay; fish abundance and species diversity are 

 greater in the lagoon from May to September than 

 from November to February (J. Gustafson 12 ). 

 Scheffer and Sperry (1931), Spalding (1964), and 

 Pitcher (1977 footnote 5), suggested that harbor 

 seals are opportunistic, preying primarily upon 

 small schooling fish. In a study by Brown and Mate 

 (1983), peak abundance of seals in Netarts Bay, 

 Oreg., also occurred in the fall and coincided with 

 the seasonal abundance of chum salmon. Move- 

 ment to Bolinas Lagoon at a time of high food 

 availability may be a consequence of the seal's 

 opportunistic feeding strategy. 



Time of day and tide were important factors that 

 influenced daily haul out patterns of seals. The 

 peak in numbers during early afternoon is consis- 

 tent with studies on the Farallon Islands (Ainley 

 et al. 1977 13 ) and in San Francisco Bay (Fancher 

 1979). Though seals were seen hauled out on KI at 

 night on 10 occasions, the sharp drop in numbers 

 during late afternoon suggests that diurnal haul- 

 ing out is preferred. The diurnal pattern may also 



9 Johnson, M. L., and S. J. Jeffries. 1977. Population evalu- 

 ation of the harbor seal (Phoca vitulina richardi ) in the waters of 

 the State of Washington. U.S. Dep. Commer., N.T.I.S. PB-270 

 376, 27 p. 



10 Allen, S. G., and H. R. Huber. 1983. Pinniped assessment 

 in the Point Reyes/Farallon Islands National Marine Sanctuary, 

 1982-83. Annual Report to U.S. Department of Commerce, 

 Sanctuary Programs Office, 64 p. 



"Risebrough, R. W, D. Alcorn, S. G. Allen, V. C. Alderlini, L. 

 Booren, R. L. DeLong, L. E. Fancher, R. E. Jones, S. M. McGinnis, 

 and T. T. Schmidt. 1978. Population biology of harbor seals in 

 San Francisco Bay, California. U.S. Dep. Commer., N.T.I.S. 

 PB81-107963, 67 p. 



12 J. Gustafson, Environmental Consultant, Resources and 

 Ecology Projects, Mill Valley, CA 94941, pers. commun. August 

 1979. 



13 Ainley, D. G„ H. R. Huber, R. P Henderson, T J. Lewis, and 

 S. H. Morrell. 1977. Studies of marine mammals at the Faral- 

 lon Islands, California, 1975-76. U.S. Dep. Commer., N.T.I.S. 

 PB-266 249, 32 p. 



498 



