FISHERY BULLETIN: VOL. 87. NO, 2, 1989 



a refuge in size from this predator at about 125 

 mm. 



Hines and Pearse (1982) present size-frequency 

 and density data for several red abalone populations 

 in central California under different predation 

 regimes which contrast markedly with the results 

 from Johnsons Lee. The Hopkins Marine Life 

 Refuge has been within the sea otter range for many 

 years and the abalone population is dependent on 

 an extensive crevice refuge. The density of emer- 

 gent abalones was higher than at Johnsons Lee (0.18 

 m"^ vs. 0.06 m"^) but the effects of intense otter 

 predation are apparent in the population structure; 

 the average size of emergent red abalones was 8 cm 

 (vs. 15 cm at Johnsons Lee) and of shells was 10 cm 

 (vs. 11 cm). At a site north of the otter range, the 

 average size of red abalones was 18 cm, more than 

 twice the size at Hopkins, and an average shell size 

 of 21 cm was found at an intermediate site recently 

 invaded by the mammals (Hines and Pearse 1982). 

 These rather dramatic differences underscore the 

 importance of studying abalone population dynamics 

 in a biogeographic context. 



In his monograph on California abalones. Cox 

 (1962) reported that the San Miguel Island fishery 

 produced predominantly red abalones while all the 

 other Channel Islands produced mostly pink aba- 

 lones. (These were the only two haliotids fished at 

 this time in California.) In this 5 yr study, only three 

 live pink abalones were found on the transects at 

 Johnsons Lee, all in 1978. Unpubhshed CDFG block 

 landing records for 1983 indicate that the red aba- 

 lone harvest exceeded the take of pink abalones on 

 Santa Rosa, San Nicolas, and Santa Cruz Islands 

 as well as on San Miguel Island. The reason for this 

 apparent shift in species composition is not clear. 

 Growth and larval survival of H. rufescens are op- 

 timal in cooler temperatures than for H. corrugata 

 (Leighton 1974), but there is no evidence of a cool- 

 ing trend since the early 1960s sufficient to produce 

 this shift. As both species have been fished actively 

 since the mainland south of Point Conception and 

 the Channel Islands were reopened to commercial 

 harvest in 1943 (Cox 1962), it is unlikely that com- 

 petition was important. Pink abalones are more 

 susceptible to bar cuts (Burge et al. 1975), a factor 

 which would affect both the yield to the fishery and 

 the reproductive potential of the stocks. The shift 

 may reflect relative differences in egg production 

 or larval dispersal potential. The average red aba- 

 lone landings for 1981-86 represented 17% of the 

 average landings for this species for 1950-70. For 

 pink abalones, which were not affected by sea otter 

 predation during this time period, this figure was 



3%, suggesting that H. corrugata is less resilient 

 to fishing pressure. 



Mortality Rate 



Our mortality rate estimate for H. rufescens is 

 higher than Smith's (1972) estimate (an annual turn- 

 over rate of 0.05), that was based on a length- 

 frequency sample of large animals in northern 

 California. Our estimate is consistent with the direct 

 observations of survival at Hopkins (Hines and 

 Pearse 1982), but lower than estimates of turnover 

 rate from empty shell production obtained in the 

 same study. These authors favor the estimate based 

 on shell production, considering that their choice of 

 sites for making direct observations may have been 

 biased toward good refuges from predation. With- 

 out predation by sea otters, mortality rates in the 

 two studies would appear to be similar. 



The Fournier and Breen (1983) method can only 

 estimate total mortality rate, which in this study is 

 based on both exploited and preexploited segments 

 of the population. We estimate natural mortality as 

 being less than this estimate. A second problem is 

 the assumption that each cohort resulted from an 

 initial cohort of the same size. The data (Fig. 2) do 

 not support this assumption but the effects of 

 variable recruitment are minimized by combining 

 data for several years in a standardized form. 



Natural mortality rate may vary with age in some 

 molluscs, and such variation is important in mak- 

 ing estimates from age-structured models (Apple- 

 doorn 1988). Mortality may be higher for the first 

 two age classes of red abalones, because we obtained 

 higher mortality estimates when these were in- 

 cluded (Table 6). After the first two age classes mor- 

 tality estimates were stable, supporting the assump- 

 tion that natural mortality is independent of age. 



Implications for Management 



Several authors consider abalones to be particular- 

 ly susceptible to "recruitment overfishing" sensu 

 Gulland 1973 (e.g., Harrison 1969; Sainsbury 1977; 

 Mottet 1978; Breen 1980). Some reasons for this 

 belief are as follows (see also Harrison 1986). Low 

 adult natural mortality rates lead to dense "top- 

 heavy" populations with a high proportion of large 

 adults. Because fecundity increases exponentially 

 with length, much of the breeding capacity of the 

 population can be concentrated in these large indi- 

 viduals. Diving is an extremely efficient harvesting 

 method on open substrate types (e.g., Beinssen 

 1979b). Abalones have a high imit value, so economic 



334 



