Ralston and Howard: Year-class strength and cohort variability in Sebastes mystmus and 5. flavidus 



717 



specific synchrony in the recruitment patterns of com- 

 mercially harvested stocks on the west coast of North 

 America, particularly on spatial scales similar to 

 ours. Similar findings had been previously reported 

 from commercial stocks in the northwest Atlantic 

 (Koslow, 1984). 



Although there is striking broad-scale synchrony 

 in our data (Fig. 2), there is also reason to believe 

 that in some years mesoscale differences in environ- 

 mental conditions may strongly influence abundance 

 patterns. For example, results from 1991 seemed to 

 show an uncoupling of the abundance trends of blue 

 and yellowtail rockfish at Mendocino County with 

 those at Sonoma County, as well as with the trawl 

 survey data (Fig. 2). That year was distinctive be- 

 cause northerly distributed species (e.g. Sebastes 

 emphaeus, Sebastes melanops, and Sebastes pinniger) 

 were relatively common in the trawl catches. 



Ontogeny and year-class strength 



Year-class strength in blue and yellowtail rockfish is 

 apparently established prior to the late pelagic juve- 

 nile stage. The strong correspondence between esti- 

 mates of year-class strength obtained from trawl 

 surveys of pelagic juveniles collected during May- 

 June and from direct observations of settled juve- 

 niles in nearshore habitats three months later indi- 

 cates that reproductive success is governed prima- 

 rily by events that occur earlier in the life history, 

 presumably during the larval stage (Houde, 1987; 

 Myers and Cadigan, 1993). 



This is not to suggest that substantial mortality 

 does not occur after the pelagic juvenile stage. In- 

 deed, over the same time period (1983-92), the 

 among-year variation in nearshore settled juveniles 

 was much greater than that for offshore pelagic ju- 

 veniles (CVs are presented in Table 1). However, 

 because cohorts maintained their rank in year-class 

 strength through the settlement transition to 

 nearshore habitats (Figs. 2 and 3), the increased vari- 

 ability evident in the Mendocino and Sonoma data is 

 not associated with a reordered strength of year classes. 



A possible reason for the greater CVs of recently 

 settled rockfish is that the data collected nearshore 

 are more strongly affected by spatial patchiness in 

 the distribution of young-of-the-year juveniles. Ow- 

 ing to the inherently broader spatial scale of the trawl 

 survey (i.e. each tow covered -1 km and the survey 

 area extended 200 km in a north-south direction), 

 this latter sampling method would have integrated 

 spatial patchiness on the scale of 10 3 -10 5 m. This 

 hypothesis predicts a weakening of the correlation 

 between pelagic and settled juvenile abundances with 

 an increase in the difference in CVs. However, blue 



rockfish from Sonoma showed the greatest increase 

 in CV through the settlement transition (1.98 to 3.68) 

 and yet had the highest correlation with pelagic ju- 

 venile numbers (0.86). 



A competing hypothesis to explain the apparent 

 increase in CVs is that of depensatory mortality 

 during settlement. This type of mechanism could 

 further deplete already weak year classes, while hav- 

 ing virtually no effect on strong ones. The result would 

 be an increase in interannual variability in abun- 

 dance without altering relative year-class strength. A 

 logical source of depensatory mortality at settlement 

 is predation by a predator assemblage, which takes 

 a fixed number of settling fish each year. However, 

 Hallacher and Roberts (1985) showed that many 

 kelp-dwelling fishes prey heavily on blue rockfish at 

 the time of settlement but not during the rest of the 

 year (see below). Similarly, recent work by Hobson 

 et al. 5 has shown that three kelp-forest inhabitants 

 that are not highly piscivorous, i.e. Hexagrammus deca- 

 grammus, Sebastes melanops, and Sebastes mystinus, 

 feed heavily on recently settled juvenile blue rock- 

 fish in years when settlement is strong. These stud- 

 ies indicate that compensation during settlement is 

 more likely than is depensation. 



Whatever the cause of the greater CVs of settled 

 juveniles, the strong correspondence between trawl 

 and diver estimates of year-class strength in blue 

 and yellowtail rockfish indicates that reproductive 

 success has been established by the end of the pe- 

 lagic juvenile stage (see also Myers and Cadigan, 

 1993). By estimating levels of variability in fecun- 

 dity, one can infer at a much earlier stage that 

 interannual differences in spawning output are too 

 small to account for fluctuations in recruitment to 

 the fishery (Table 1; see also Shepherd and Cushing, 

 1980). A seven-year study of weight-specific fecun- 

 dity in yellowtail rockfish (Eldridge and Jarvis, 1995) 

 yielded an among-year CV of 0.10, after within-year 

 measurement error was removed (see Methods sec- 

 tion). This amount of variation is insufficient to ac- 

 count for fluctuations observed at the time rockfish 

 cohorts recruit to the fishery. 



Levels of variability at the young-of-the-year ju- 

 venile stage, however, are more than adequate. In 

 fact, it would seem that rockfish go through a phase 

 of compensatory mortality, from the settled juvenile 

 stage to the time a cohort enters the fishery (i.e. ages 

 3-4 in rockfishes). Note that for a log-normal distri- 

 bution, the CV of year classes recruiting to rockfish 

 fisheries, based on recruitments estimated from 



5 Hobson, E. S., J. R., Chess, and D. F. Howard. 1995. Interannual 

 variations in predation on juvenile Sebastes spp. by three north- 

 ern California predators. Unpubl. manuscr. 



