Yoklavich et al.: Larval rockfishes and their physical environment off central California 



769 



Ralston and Howard (1995): 0.1165, 0.10, 0.08, 0.06, 

 0.04, and 0.03/d for fish of age <15, 15-27, 28-42, 

 43-70, 71-100, and >100 d, respectively. 



Larval identification 



Preflexion and flexion rockfish larvae were identi- 

 fied primarily from melanistic pigment patterns, in- 

 cluding: length of ventral midline series, occurrence 

 of dorsal midline pigmentation, and occurrence of 

 pectoral fin pigmentation. Distinctions among lar- 

 vae with shared primary pigment patterns were 

 based on secondary pigment characters (e.g. exter- 

 nal dorsal midline pigment configuration, lower lip 

 pigment, and pectoral fin size and pigment pattern), 

 and on reproductive seasonality of adults (Wyllie 

 Echeverria, 1987; Love et al., 1990). 



Preextrusion, yolk-sac, and preflexion stages have 

 been illustrated or described in some detail for 49 of 

 the 52 species of rockfishes off central California (see 

 Kendall [1991] for a general review). More advanced 

 larval stages have been described for relatively few 

 species of rockfish. We used serial descriptions of lar- 

 vae from preflexion through flexion to identify 

 shortbelly (Moser et al., 1977), blue (Moreno, 1993), 

 cowcod (S. Levis; Moser et al., 1977), bocaccio {S. 

 paucispinis; Moser, 1967), and stripetail rockfish (S. 

 saxicola; Laidig et al., 1996). All postflexion larvae 

 with dorsal, anal, and pectoral fin rays were identi- 

 fied to species. 



Squarespot rockfish, S. hopkinsi, were abundant 

 as preflexion larvae and were identified cautiously 

 from a description of their preextrusion stage (Moser 

 et al., 1977). Two groups of species also were identi- 

 fied on the basis of shared pigment patterns. The 

 "copper complex" informally is synonymous with the 

 subgenus Pteropodus (Jordan et al., 1930; Kendall, 

 1991), and comprises copper (S. caurinus), gopher 

 (S. carnatus), black-and-yellow (S. chrysomelas), 

 grass (S. rastrelliger), china (S. nebulosus), and quill- 

 back (S. maliger) rockfishes. We include brown (S. 

 auriculatus) and kelp (S. atrovirens) rockfishes in a 

 "copper complex +" because their pigmentation is 

 indistinguishable from members of Pteropodus. A 

 "Sebastosomus +" group includes black, olive (S. 

 serranoides), yellowtail (S. flavidus), widow (S. 

 entomelas), and bank (S. rufus) rockfishes. Identifica- 

 tion of this group is limited because yolk-sac and early 

 preflexion larvae are indistinguishable from many other 

 species, including most in the subgenus Sebastomus. 



Data analysis 



Larval abundance was calculated for each replicate 

 at each station by multiplying the number of larvae 



by depth of tow and by dividing by the volume of 

 water filtered per sample. Abundance was expressed 

 as number of larvae per 10 m 2 . Abundance data were 

 examined for normality and were log-transformed 

 where appropriate. Among-station and among-time 

 differences in total abundance of larval rockfishes were 

 tested for significance by using a 2-factor analysis of 

 variance (ANOVA), including only those samples col- 

 lected during the first eight periods in 1991-92 (i.e. 

 those with three replicates per station and five stations 

 per time period). Between-year comparisons of abun- 

 dance were made with 2-factor ANOVA (i.e. factors 

 were year and time) by using only those time peri- 

 ods and stations equally represented in both years. 

 Specific time periods and stations that contributed 

 to significant factors in the models were identified 

 by using post-hoc pairwise multiple comparisons of 

 cell means with Bonferroni-adjusted probabilities. 

 Differences in mean size and growth rate of the two 

 dominant species of larval rockfish among stations, 

 collection times, and years were similarly analyzed. 



Results 



The physical environment 



Our 1991-93 study period off central California was 

 characterized by a prolonged El Nino event. Anoma- 

 lously warm, low-salinity water in nearshore areas 

 during much of our study indicated an onshore dis- 

 placement of the California Current, similar to that 

 reported by Lynn et al. (1995) for this time period 

 and by Simpson ( 1984) for the 1982-83 El Nino event. 

 Upwelling was reduced and delayed relative to other 

 years, and distinct persistent upwelling plumes were 

 not evident within the survey area during the sam- 

 pling periods. The two years differed, however, in the 

 intensity, duration, frequency, and direction of wind 

 events that affected transport processes. 



Coastal winds off Monterey prior to the 1991-93 

 El Nino event were primarily from the northwest 

 during December to April (Fig. 2A), as estimated from 

 daily wind components measured at buoy 46042 

 (Fig. 1) and averaged from 1988 to 1991. Associated 

 with this prevailing wind pattern was a steady shoal- 

 ing of cold, high salinity water, indicative of up- 

 welling, in the upper 200 m of the water column in 

 the outer part of Monterey Bay, as measured by a 

 total of 37 CTD casts from stations H3, C7, and Ql 

 (Fig. 1) during April 1989, December 1989-April 

 1990, and December 1990-March 1991 (Fig. 2, B and 

 C; raw data from Rosenfeld et al., 1994b). 



During winter 1991-92, daily wind speeds were 

 largely greater than the average calculated during 



