556 



Fishery Bulletin 103(3) 



ID 



CQ 



> 



o 



1991-27 



1991-86 



1991-168 



1991-178 



1988-100 



1.0 

 0.5 

 0.0 

 -0.5 

 -1.0 



1.0 

 0.5 

 0.0 

 -0.5 

 -1.0 



1.0 

 0.5 

 0.0 

 -0.5 

 -1.0 



1.0 



0.5 

 0.0 

 -0.5 

 -1.0 



1978 1980 1982 1984 1986 1988 1990 



0.5 



00 

 -0.5 

 -1 



1978 1980 1982 1984 1986 1988 1990 



Year 



Figure 2 



The time series of 6 18 (•, and left axis) and the linear model residuals 

 (□, and right axis) taken from each black rockfish [Sebastes melanops) 

 otolith used in the present study. The solid line is the linear model fit to 

 the 6 18 data. A residual value of zero indicates perfect agreement between 

 observed and predicted year-specific average residuals. Sample numbers 

 corresponding to Table 1 are given inside each graph. 



northerly. Furthermore, the isotope variability between 

 fish may be due to fish inhabiting different areas in the 

 early periods of life or to temporal differences in growth. 

 Finally, changes in calibration of the spectrometer be- 

 tween assays may be a source of uncertainty. 



A critical assumption behind the present study was 

 that the lowest 6 18 corresponds to the warmest water 

 temperature, and consequently the 1983 El Nino that 

 serves as the time marker. The 6 18 values may be 

 impacted by salinity in addition to water temperature 



(Dorval, 2004), and we assumed that salinity was con- 

 stant and that the changes in <5 18 values were largely 

 influenced by changes in temperature. A further con- 

 founding element to this kind of study is the ability of 

 fish to move and potentially select microhabitats with 

 different temperatures than that of the average local 

 environment. Natural date-specific markers also must 

 be monitored over a number of years to ensure that 

 they remain identifiable within the otolith (Campana, 

 2001). We addressed this concern by selecting fish of 



