528 



Fishery Bulletin 89(3). 1991 



80- 



E 40 



^E 



£ 20 



80 



40 



20 40 60 80 100 120 140 160 180 

 Age (d) 



Figure 6 



Scattergrams of widow rockfish Sebastes entomelas standard 

 length on age for each year, 1985-88 (upper), with fitted linear 

 regressions to each year of data (lower). 



occurred in 1985 (0.974 mm/day), whereas the lowest 

 growth rate (0.563 mm/day) was seen in 1986. Results 

 of ANCOVA (Table 3) show that significant differences 

 (P 0.001) in growth rate occurred among years. Results 

 of all possible pairwise LSD tests comparing annual 

 slope estimates (Fig. 3) show that growth in 1985 was 

 distinctly fast. Growth rates were relatively low in 1986 

 and were moderate and similar in 1984, 1987, and 1988. 



Chilipepper Sufficient numbers of fish were collected 

 during 1985-88. Standard length and age data from 

 these years (Fig. 5) indicate an apparent linear rela- 

 tionship. Regression statistics resulting from fitting 

 each year's data to the linear growth model are pro- 

 vided in Table 2. Annual growth rates ranged from a 

 low of 0.399 mm/day in 1986 to a high of 0.555 mm/day 

 in 1987. Results from ANCOVA (Table 3) show that 

 annual differences exist in growth rate (P 0.002); pair- 

 wise LSD comparisons of slopes (Fig. 3) indicate a very 

 clear separation between the slow growth that occurred 

 in 1985-86 and the relatively fast growth in 1987-88. 



Widow rockfish Like chilipepper, widow rockfish 

 were collected in sufficient numbers only in the years 

 1985-88. The scatter of data (Fig. 6) shows an ap- 



parently linear increase in standard length, at least 

 over the range of ages sampled. A problem with the 

 data is that in 1986 only 9 fish were collected, with ages 

 ranging 80-110 days. This resulted in a marked reduc- 

 tion in age contrast (30 days). 



The data from each year were fitted separately with 

 linear regression (Table 2), with estimates of growth 

 rate ranging from 0.299 mm/day in 1986 (N9) to 0.607 

 mm/day in 1987. When ranked, the descending se- 

 quence of estimated annual growth rates (i.e., slopes) 

 was: 1987>1988>1985>1986. Results of ANCOVA 

 (Table 3), however, indicated that slopes did not vary 

 significantly among years (P 0.184), although all pos- 

 sible pairwise LSD comparisons (Fig. 3) showed that 

 growth rate in 1986 was slower than in other years. 

 The reason the ANCOVA failed to reveal differences 

 in slope among years was likely due to a small sample 

 size and low age contrast in 1986. An F-test for an- 

 nual differences in adjusted means (i.e., elevation) did 

 reveal significant differences (Table 3, P 0.001). Pair- 

 wise LSD comparisons of annual adjusted means in- 

 dicated that pelagic juveniles were larger in 1987 than 

 in other years. When ranked, the descending sequence 

 of annual adjusted means was: 1987 > 1986 > 1988 

 >1985. 



