POPULATION HETEROGENEITY IN PACIFIC PILCHARD 



215 



Early in the examination of scales for age 

 determination, qualitative differences were noticed 

 in the scale patterns of pilchard from different 

 regions of catch. An attempt was made to estab- 

 lish criteria for separating ring types somewhat as 

 had been done in herring by Norwegian investiga- 

 tors (Runnstr0m 1936). Although scales collected 

 in the Pacific Northwest showed predominantly 

 heavily marked winter rings of the northern type 

 associated with sharp cessation of growth (cf. fig. 

 5a), and those from the San Pedro fishery showed. 

 as a rule, the faintly marked annuli suggesting 

 milder winters (cf. fig. 5b), the presence of inter- 

 mediate ring types in central California made the 

 separation too subjective for quantitative treat- 

 ment. 



Further attempts to separate the vai\ ing growth 

 pal terns in individual fish were made using trans- 

 formation values (calculated by the method of 

 semiaverages) asymptotic length (/,) and rate 

 of deceleration of growth (k) for individuals 

 caught in four geographical regions. For the 

 L939 class, the regression of the values /„ on I: 

 gave scatter diagrams (fig. 6) for ( 'anada (N = 286, 

 dots) and San Pedro (N=433, circles). Even 

 though there is overlapping in the scatter of 

 individuals caught in the two areas, a huge pro- 

 portion of 4-ring sardines of the growth type 

 caught in San Pedro are outside the scatter For 

 the Pacific Northwest. This type was character- 

 ized by relatively steep slope, or slow rate of 

 deceleration, and small calculated ultimate size. 

 We interpreted this to mean that there was in 

 turn a large proportion of San Pedro 4-ring fish 

 that would never have been caught in the Pacific 

 Northwest. Likewise, a smaller proportion of 

 Canadian fish were outside the San Pedro scatter 

 and might never he caught in southern California. 



Scatter diagrams dig. 7) of San Francisco 

 (N = 514, symbolized by x) and Monterey 

 (N=783, symbol q) show many fish of the north- 

 ern type. Numerous individuals from Monterej 

 catches also fall within the scatter of character- 

 istically Sa n Pedro type (cf. fig. 6). San Fran- 

 cisco shows few of the latter hut a large propor- 

 tion of types common to Canada; namely, with 

 relatively rapid deceleration rates and greater 

 asymptotic lengths. 



These regressions of /., on /' for 1939-ycar-class 

 pilchard fcaught in the 1943-44 season were 

 curvilinear as in the aquarium fish Plaiypoecilus 



macvlMus, and some individuals with improbably 

 high calculated asymptotes 2 were found as in 

 platies (Felin 1951). 



Similar regressions for individual pilchard of the 

 same year class were plotted at the 6-ring stage 

 (fig. 8). By this time, individuals of an unlikely 

 predicted ultimate size were not represented ia the 

 catches anywhere along the Pacific coast and cal- 

 culated asymptotes were well within the probable 

 range of size for the species. As a result of the 

 disappearance of these types, the regressions ap- 

 peal- linear. 



The scalier for 6-ring pilchard at all four ports 

 included 194 individuals. Of these, 10 percent 

 were entirely outside the main scatter <>f all fish. 

 This group showed southern growth characteris- 

 tics. Out of 70 from Canada only I percent were 

 of this type with slow deceleration rale and low 

 predicted size. Of •"> 1 fish from San Francisco, 1 I 

 percent were of this type; of 57 in Monterey, 7 

 percent; and out of the small number, 13, of this 

 older age caught at San Pedro, 62 percent had the 

 typically southern growth characteristics. 



The high percentage of these fish in San Pedro 

 is in agreement with the appearance of predomi- 

 nantly southern-type annuli in San Pedro scales. 

 Man) scale samples taken from sardines oil* south- 

 ern California and a few collected oil' Lower Cali- 

 fornia show faint winter marks and wide zones of 

 summer growth; the yearlj increments decrease 

 very slowly in si/.,', illustrative of their slow de- 

 celeration in growth. (See figure 5b.) These fish 

 are characteristically of small size for their age 

 compared with northern types (see fig. 5a), and 

 so far have not been oldened to attain as great 

 age as the large northern pilchard. Their position 

 in this respect appeal's similar lo that of the warm- 

 water Mediterranean pilchard when compared 

 with .North Sea types (cf. fig. 2'. 



From the few available samples of Mexican fish 

 taken off northern Lower California there is also 

 indication of another southern growth type charac- 

 terized by a low k value (the slope is even less than 

 that for northern types), and l>\ a small calculated 

 ultimate size. (See figure 5b.) 



The pilchard scales were photographed by means 

 of infrared plates. This method was used l).\ 

 Roper (1936), to bring out year rings as distincf 



-'.I. I.. Hurt (personal communication) recorded ;i standard length ol 351 

 mm. for a pilchard caught off thi of Vancouver Island and landed 



it Mootka. 



