BOEHLERT ET AL.: TIME SERIES OF GROWTH IN SEBASTES 



(U 





a 



1B90 1900 1910 1920 1930 1940 1950 1960 1970 1980 



Year 



Figure 2. — Distribution of the years of birth; males and females are com- 

 bined. A. Sebastes ptiiiiiger. B. S. diplopma. 



throughout the time series. After about 1972, all 

 age gi'oups showed concordance in a relatively 

 abrupt increase in gi'owth rates. This increase 

 may have been shorter lived for age gi'oups 5 and 

 6, as suggested by the decrease in positive 

 growth rates for the last few years of the record. 

 Positive gi'owth rates continued to increase to 

 the end of the record in the 5 yr running aver- 

 ages for age gi'oups 1-4, though with some sig- 

 nificant year-to-year variability in the yearly 

 data. 



The time series for S. diploproa (Fig. 4) was 

 much longer than for S. pinniger (Fig. 3). Age 

 gi'oups 2-A generally showed a gi-adual decreas- 

 ing trend in gi'owth rates prior to about 1965 or 

 1970. This pattern was similar to that observed 

 for S. pi)iiiige)- age gi'oups 2—4 over the more 

 limited record length. This long-term trend in 

 decreasing growth rates was less evident in S. 

 diploproa age gi'oups 1, 5, and 6. Growth rates 

 for age group 5 were nearly normal throughout 



the record. Growth rates for age group 6 were 

 relatively low in the early part of the record and 

 slightly higher than normal for the period 1930- 

 55. They decreased briefly from 1955 to 1960, 

 then were steady, near-average, or slightly in- 

 creasing until about 1970. Age gi'oups 2-6 of S. 

 diploproa increased in growth rates during the 

 1970's, coinciding with the high gi'owth rates of 

 all age groups of S. pi)unger after about 1972. 

 Growth rates for age group 1 of S. diploproa 

 were strikingly different from all other age 

 gi'oups of both species of Sebastes, with a pre- 

 dominance of variabiHty over much shorter time 

 scales (ca. 5 years) and no evidence of a rapid 

 increase during the 1970's. 



The relationship between growth variability in 

 different age groups summarized above from the 

 time series in Figures 3 and 4 were compactly 

 described by the principal components of the six 

 age gi'oup time series. We included only years 

 for which yearly mean values of growth were 



795 



