POPILATTON HETEROGENEITY EST PACIFIC PILCHARD 



223 



2. The direct proportionality of growth of 

 scale to growth of fish is used to obtain approxi- 

 mations of mean calculated lengths of sardines. 



3. Walford's (1946a) transformation plot is 

 applied to mean observed length data and mean 

 calculated length data sampled in different 

 regions. 



4. Comparison is made of growth character- 

 istics, k (slope of the regression, or rate of decelera- 



tion of growth) and /., 



(•"-'; 



— intercept 



or eal 



culated ultimate size) of six year classes sampled 



in the most northern region of catch, Canada, 

 am! the most southern, San Pedro. 



5. From analysis of covariance no significant 

 differences are evidenced in the mean slopes. /,-. 

 of the transformations between each of these six 

 year-classes in Canada and San Pedro. 



ti. For each year class tested, two distinct 

 ^-intercepts, or levels, of mean transformations 

 are maintained in the northern and southern 

 areas, and from the covariance tests the differ- 

 ences are significant at the 1-percent level. The 

 growth characteristic /. thus differs significantly 

 in the northern and southern catches. 



7. The use of the straight-line transformation 

 as a method of expressing growth differences not 

 readily apparent in conventional length-on-time 

 growth curves was noted by Walford (1946a). 

 The applicability of the usual statistical tests of 

 significance to such regressions of /„ . , on l„ is now 

 also apparent. Possible meanings of significant 

 difference in /• and /, as physiological characters 

 arc discussed. 



8. Complete intermixture and homogeneity in 

 populations of adult fish as sampled by the fishery 

 in different regions is not evidenced from data on 

 mean calculated lengths. 



0. The apparent cline in the growth charac- 

 teristic / appears indicative of intraspecific 

 populations in which there is limited intermingling, 

 and suggests a series of overlapping coastal 

 migrations of more than one stock. 



10. The recent discovery of an area of intense 

 spawning off central Lower California, more or 

 less discrete from the widespread spawning area 

 off southern California, points to the tentative 

 hypothesis that spawning grounds off Lower 



California give rise to the southern components 

 found in catches off San Pedro (and to a lesser 

 degree off central California), while the larger 

 fish originate off southern California and occa- 

 sionally to the north. 



11. Evidence from qualitative and quantitative 

 differences in individual scale and growth pat- 

 terns indicates some independence in the fished 

 stock of the Pacific Northwest and southern 

 ( lalifornia. 



12. The decline in catches in central California 

 and the Pacific .Northwest has been almost 

 entirely responsible for the sudden decline in 

 total catch of the entire coast. It is suggested 

 that the centering of the available population in 

 the southern part of the range may have pro- 

 duced a series of southern year classes. 



13. Ratios between year-class strength of 1- 

 or possibly 2-ring fish taken at San Pedro and 

 taken along the coast as a whole may prove a 

 useful index in determining whether a year cla^s 

 is primarily southern or has more northern 

 components. 



II. The measure of year-class strength used by 

 Walford (1946b), i. e., the number of 3-ring 

 pilchard caught at all ports, is compared with 

 the San Pedro- Pacific coast ratios of I- and 2-ring 

 fish. To date only those year classes which 

 appear predominantly southern by these criteria 

 have been much below normal strength. 



I .">. Vital Statistics of the fishery which indi- 

 cated that new sources of fish became unexpect- 

 edly available in the L949 50 season appear to be 

 explained at least in part by an influx, probably 

 from the south, of small fish onto the southern 

 and central California fishing grounds. 



16. Bimodality in length-frequency composi- 

 tion is further evidence from growth that pil- 

 chard caughl along the Pacific coast do not 



constitute a homogeneous population. 



17. Whether heterogeneity in growth charac- 

 teristics is the expression of genotypic difference 



or a phenotypic response of a species to its 

 environment is not yet determined. 



IS. In view of differences in slocks on the 

 various fishing grounds along the Pacific coast, 

 the study of population dynamics not only for the 

 coast as a whole but also by geographic areas 

 appears desirable. 



