Table 1 gives the sample size (N) and the mini- 

 mum and maximum total lengths used in the anal- 

 ysis for each species and each sex. Table 2 shows 

 estimates of ^/-intercept (a), slope (6), standard er- 

 ror of estimate (Sy^), correlation coefficient (r), 

 and F for each species and sex. Analysis of covari- 

 ance was used to determine if separate lines for 

 males and females significantly reduced the variance 

 from a common line (Kleinbaum and Kupper 1978). 

 Analysis of covariance was also used to test for 

 significant differences in the relationship of otolith 

 length to total length between the sexes at the P 

 = 0.05 level and the P = 0.01 level (Table 2). The 

 highest values of r and examination of scattergrams 

 (Fig. 2) indicate that the length relationships are 

 linear over the observed range of values. Limiting 

 the application of these regressions to the ranges 

 of observed values is advised. 



Results and Discussion 



Linear regressions were run on each sex in order 

 to investigate possible sexual differences. In 17 of 

 the 30 species investigated, the relationship between 

 otolith length and fish length is significantly differ- 

 ent between males and females (Table 2). Sexual size 

 dimorphism has been observed in 11 of the 17 

 species in Table 2. These species (plus S..alutus) in- 

 clude most commercially and sport-caught rock- 

 fishes in the northeastern Pacific Ocean. The six 

 species for which growth curves have yet to be con- 



Table 1 —Sample sizes and size ranges used in the linear regres- 

 sions of total lengtli versus otolith length for Sebastes. Measure- 

 ments are in millimeters. 



500 



450 



E 

 E 



X 



O 



z 



UJ 



400 



350 



< 

 O 



300 - 



250 - 



Female= ° Range a — a 

 Male= • Range 



12 13 14 15 16 

 OTOLITH LENGTH (mm) 



17 18 



Figure 2.— Linear regression of total length on otolith 

 length of widow rockfish, Sebastes e7itomelas. The range 

 of values for males (•-•) and females (o-o) at each whole 

 millimeter of otolith length. 



384 



