POPULATION HETEROGENEITY IX PACIFIC PILCHARD 



207 



GROWTH CURVES OF 1939 YEAR CLASS 



Utilizing mean calculated lengths (table 2) for a 

 given year class of Pacific pilchard taken in the 

 most, northern region of catch, Canada, and in 

 the most southern, San Pedro, we may compare 

 growth curves derived from mean calculated 

 lengths with those derived from mean observed 

 lengths for each year class throughout its life in 

 the fishery. An advantage in the transformation 

 plot is graphically illustrated by figures 3 and 4. 

 It is not readily apparent from the conventional 

 growth curves of the 1939 year class (lig. :>) that 



the crosses delineating the growth curves derived 

 from mean observed lengths of the year class in 

 successive seasons express the same rate of de- 

 celeration of growth as do the mean back calcula- 

 tions of length for 4-ring, 5-ring and 6-ring fish 

 (circles, squares, and triangles) of this 1939 year 

 class. Solid symbols represent northern cal- 

 culated-length data, open symbols the southern. 

 These same curves transformed (fig. 4) show the 

 relative constancy of slope (k) as well as level 

 (y-intercept) of the regressions for northern and 

 southern populations, whether plotted from 

 observed or calculated length data. 



t 1 i r 



"I - 1 



130 



X 



A 



a t 



# t 



D 

 T 



T 



•-A 



^so 90 ioo no 120 130 140 150 J60 170 ieo 190 200 210 220 230 240 250 



LENGTH AT AGE n 



Figure 4. — Transformations of the growth curves of the L939 year class shown in figure '■'>. 



258571—54- 



