192 



Fishery Bulletin 94(1), 1996 



sis program (SAS Institute Inc., 1987). The log-lin- 

 ear option was used because edge types would not be 

 normally distributed within the month. The analy- 

 ses were used to test for interactions between edge 

 type and each factor (sex, year, and location). In addi- 

 tion, two-way interactions were tested for the relation- 

 ship of edge type to sex by year, sex by state, and year 

 by state. Since there were no samples from Washing- 

 ton in February 1985, May and June of 1985, and June 

 of 1987 (Table 1), the analyses of state, year by state, 

 and state by sex, were not done for these months. Each 

 test was conducted separately by month because edge 

 type was known to vary between months (Lenarz, 1987 1 

 and because the purpose of this study was to examine 

 differences among sexes, years, and areas. 



The times of formation of the hyaline zones are 

 discussed in light of sea-surface temperature data 

 from northern California and southern Washington. 

 Mean monthly sea-surface temperatures were ob- 

 tained from National Weather Service marine buoys 

 located off Point Arena (northern California) and 

 the Columbia Basin (where many of the Washington 

 fish were caught). 



Results 



There was a significant difference in edge type be- 

 tween the sexes for the month of March (Table 2). 

 Examination of the data indicates this difference was 

 due to a difference in the percentage of opaque zone 

 types present. Males had a slightly smaller fraction 

 of narrow opaque zones than did females (5.7% vs. 

 6.3%), and a slightly larger percentage of wide opaque 

 zones (8.4% vs. 6.9"^ ). There were no significant dif- 

 ferences between the sexes for any other month. 



There were significant differences in the timing of 

 hyaline-zone formation between the states for all 

 months that could be tested (Table 2). Both hyaline 

 and opaque zones form earlier in Washington than 

 in California (Fig. 1 ). By July, about 50% of all otoliths 

 from Washington have a wide opaque zone whereas 

 only about 20% of the otoliths from California have 

 a wide opaque zone. 



There were significant differences in edge types 

 among years for all months (Table 2). Hyaline zones 

 formed earlier in 1985 and 1987, at an intermediate 

 time in 1986, and later in 1983 and 1984 (Fig. 2, A 

 and B). 



There were significant differences in edge type for 

 all months for which the year-state interaction could 

 be tested (Table 2). This indicates that the states do 

 not vary to the same extent. Yearly differences are 

 much more pronounced in California than they are 

 in Washington (Fig. 2, A and B). In all cases, the an- 



nual variations are in the same direction; however, 

 the changes are smaller in Washington (Table 2). 



There were significant differences in edge types in 

 February, March, and April for the year-sex interac- 

 tion. There were no differences for this interaction 

 for January, May, or June. These differences may be 

 attributed in part to the strong effect of yearly vari- 

 ability. Examination of the frequency distributions 

 also suggests that a somewhat higher fraction of fe- 

 males have hyaline zones in February, March, and 

 April in 1983, whereas the fraction tends to be equal 

 in other years. 



There was a significant difference in edge types in 

 March for the state-sex interaction (Table 2). There 



