156 



Natural mortality of blue rockfish, 

 Sebastes mystinus, during their first 

 year in nearshore benthic habitats 



Peter B. Adams 

 Daniel F. Howard 



Tiburon Laboratory, Southwest Fisheries Science Center 

 National Marine Fisheries Service. NOAA 

 3150 Paradise Drive, Tiburon, California, 94920 



Much of the variation in year-class 

 strength among marine fishes has 

 been attributed to high mortality 

 during the egg and larval stages 

 (Ricker, 1954; Cushing, 1973; 

 Beverton, 1984). However, for 

 groundfish species that migrate 

 from a pelagic to a benthic juvenile 

 stage, there has been increasing 

 suggestion that natural mortality 

 during the early benthic juvenile 

 stage may control year-class strength 

 (Beverton, 1984; Sissenwine, 1984). 

 These arguments stem from the fol- 

 lowing: 1) demonstration of den- 

 sity-dependent natural mortality of 

 juvenile groundfish (Lockwood, 

 1980; Meyers and Cadigan, 1993); 

 2) cases where larval abundance 

 and later year-class strength are 

 uncorrelated (Bakun and Parrish, 

 1980; Saville and Schnack, 1981; 

 and Fogarty et al., 1987); and 3) the 

 reasoning that total natural mor- 

 tality during the juvenile stage may 

 be greater than during the egg or 

 larval stage, because the juvenile 

 stage lasts much longer (Sissen- 

 wine, 1984). These arguments were 

 reviewed specifically for rockfish 

 (Sebastes spp. ) by Love et al. ( 1991 ), 

 but the results were inconclusive. 



Materials and methods 



Off northern California, young of 

 blue rockfish, Sebastes mystinus, 

 released in the winter spend three 



to five months offshore as pelagic 

 larvae and early juveniles and then 

 settle to the benthos. During the 

 course of monitoring annual re- 

 cruitment of these benthic near- 

 shore juveniles, data were collected 

 from which mortality rates could be 

 estimated for juvenile blue rockfish 

 during their first year in the 

 benthic stage. Annual population 

 density estimates were the aver- 

 aged counts made between July 

 (when recruitment to nearshore 

 habitats was complete) and mid- 

 September of each year. The mor- 

 tality estimates were based on 

 counts from July through the fol- 

 lowing April. Exceptionally high or 

 low counts were excluded when the 

 distribution of juveniles was influ- 

 enced by unusual oceanic condi- 

 tions. Because the sampling began 

 after settlement was complete, 

 these estimates do not include mor- 

 tality that occurs during or imme- 

 diately following settlement. 



First-year juvenile blue rockfish 

 were counted along strip transects 

 at three northern California sites 

 during different years. Dark Gulch 

 was sampled from 1985 to 1989, 

 Salmon Point in 1987 and 1988, 

 and Horseshoe Point in 1988 (Fig. 

 1). Each site covers approximately 

 10,000 m 2 and consists of high-re- 

 lief rocky reefs surrounded by lower 

 reefs and boulders with occasional 

 sand patches interspersed. Rock 

 surfaces shallower than 15 m were 



covered with benthic algae and bull 

 kelp (Nereocystis luetkeana) from 

 July through November, after 

 which time the bull kelp was ripped 

 up by winter storms. 



One-minute strip-transect counts 

 were made by two observers using 

 SCUBA over bottom depths be- 

 tween 5 and 22 m (Fig. 2). Transects 

 always began in the offshore por- 

 tion of the study site and moved 

 inshore. Direction changes were 

 made perpendicular to the previous 

 course, usually after one to three 

 counts on the same heading. Ob- 

 servers swam 2 m off the bottom 

 and counted first-year juvenile blue 

 rockfish within 3 m in any direc- 

 tion during one-minute transects. 

 Counts were made only when hori- 

 zontal visibility was greater than 

 4 m. Twenty counts were made at 

 each site (average total counts=19, 

 range 10-35). All counts were made 

 between 1000 and 1400 hours. 



Results and Discussion 



Natural mortality was estimated 

 by using a catch-curve analysis 

 based on the exponential decrease 

 in fish abundance with age (Ricker, 

 1975; Vetter, 1988). Juvenile blue 

 rockfish density during each one- 

 minute count was transformed to 

 ln(.v+l) and plotted against age 

 (days). The slope of the least- 

 squares regression through the de- 

 creasing portion of the log-trans- 

 formed density was the instanta- 

 neous rate of decrease, or natural 

 mortality, provided there was no 

 significant movement either into or 

 out of the area. The mortality coef- 

 ficients were calculated from the 

 transformed one-minute counts 

 rather than from daily averages in 

 order to demonstrate the patchy 

 nature of the data where a large 

 portion of zero counts occurred ran- 



Manuscript accepted 25 July 1995. 

 Fishery Bulletin 94:156-162 ( 1996). 



