Smith and Abramson Leopard shark tag-recovery data from San Francisco Bay 



377 



the fewest were returned in February (2%), January 

 (4%), and July (5%). The return rate for 'healthy' fish 

 was 9.9% and that of 'injured' fish, 15.8%. Clearly, 

 there is no indication that handling during the tagging 

 process induced excess tagging mortality. The sex ratio 

 of recaptured fish was exactly the same as at tagging 

 (53% males) indicating little if any selective mortality 

 with regard to sex. 



Movements 



Bimonthly distribution patterns were plotted using 

 recapture information from 101 recoveries made dur- 

 ing 1979-88, for which recapture location and date 

 were known (Fig. 3). From about March through Aug- 

 ust recoveries were restricted to the bay, while from 

 September through February a marked pattern of 

 dispersal occurred within and outside of the bay. One 

 shark traveled as far as Elkliorn Slough near Monterey, 

 about 140 km south of the entrance to San Francisco 

 Bay. 



To discern year-to-year differences in the temporal 

 and general geographic pattern. Table 4 shows recov- 

 eries by year and recapture area (south San Franciso 

 Bay, central and north bays, ocean/outside bay). The 

 year 1983 was unique in that the greatest number of 

 recoveries in any one year was made that year (22 fish, 

 plus 2 fish recovered dead). It was also the only year 

 that ocean recoveries were made as early as Septem- 

 ber, and the only year when an unusual number of 

 recoveries (6 fish) was made in the central bay in June. 

 (During the 9-year study period, there was only one 

 other instance, in 1981, of a fish being taken in the cen- 

 tral bay in June.) 



No distribution pattern appeared to be correlated 

 with fish size or sex, but due to the small sample size, 

 the results do not preclude the possibility that the pat- 

 tern may vary with sexual maturity or age. The sex 

 ratio of fish taken in the ocean was normal (1:1); how- 

 ever, one commercial fisherman, who caught 3 of the 

 12 tagged fish taken in the ocean, reported that all 3 

 were gillnetted from the same depression on the sea 

 floor in Half Moon Bay (one taken 13 Nov. 82, the other 

 two taken in 1986 on 20 Oct. and 17 Nov.), and all 3 

 were large females with developing embryos. 



Mortality rates and virtual population analysis 



The oldest ages reported for this species are 24 years 

 for a male of approximately 135 cm in length and 20 

 years for a 130-cm female (Kusher 1987, fig. 8). How- 

 ever, we arbitrarily decided to use a maximum age of 

 30 years in Hoenig's (1983) regression equation since 

 there are records of fish larger (180 cm; Kato et al. 

 1967) than those that were aged; we also noted that 



the largest female aged by Kusher (140 cm) was, at 18 

 years, not the oldest. Because the largest fish cited 

 above was taken prior to the time when there was more 

 than nominal fishing directed at this species, we feel 

 that M may be assumed approximately equal to the Z 

 from Hoenig's regression equation. The result of tak- 

 ing t,„„j. = 30 in the equation was an estimated M of 

 0.14. 



While we have no explicit estimate of a tag-shedding 

 rate from this experiment, the results of Kato and Car- 

 vallo (1967) from multiple tagging of sharks at the 

 Revillagigedo Islands, Mexico, indicated a minimal an- 

 nual shedding rate of 0.095. We have taken this value, 

 converted it to an annual instantaneous loss rate of 

 0.10, and added it to the previously estimated 0.14 to 

 obtain X=Af -I- 0.1 = 0.24 for use in lieu of natural mor- 

 tality in the VPA. We assume, and believe it reasonable 

 given the apparent hardiness of the animals and tough- 

 ness of their skin, that there was no instantaneous (type 

 I) tag shedding. 



With regard to the percent of tag recoveries reported 

 to us, we again have no "built-in" estimate, but the 

 California Department of Fish and Game has conducted 

 extensive studies on the return rates for various re- 

 ward levels associated with tagged striped bass in the 

 same general area that the leopard shark tagging oc- 

 curred (Stevens et al. 1985). Based on their experience 

 (D. Kohlhorst, Calif. Dep. Fish Game, Stockton, CA 

 95205, pers. commun., Nov. 1988), it seemed that a 

 50% reporting rate would be the best value to use as 

 an adjustment for tag recoveries in this case. To utilize 

 this adjustment in the VPA, we use half the actual 

 number of tag releases as the target for convergence. 



