species (see Jensen, 1939 for example) . 



Despite this widely Known and accepted 

 fact, the implications with respect to the inter- 

 pretation of meristic data, for example, seem 

 to be sometimes overlooked or ignored. Only 

 under certain ideal conditions would such en- 

 vironment modified characters be of value . 

 These conditions are: (1) the environmental 

 conditions, and consequently the characters, 

 are not duplicated at different geographical 

 locations. (2) The differences produced are 

 really distinctive. (3) The behavior of the fish 

 is such that the differences are not obscured by 

 mixing (or that the sampling problems are not 

 unduly complicated). Under these ideal condi- 

 tions such characters would serve as useful 



natural tags" with respect to geographic origin 

 of fish, extent of dispersion, migrations and 

 related problems . They would not, however, 

 provide any information on the more fundament- 

 al problem of genetic difference . 



One has only to consider what is Known of 

 sardine spawning to realize how meaningless 

 (from the standpoint of subpopulation characters) 

 meristic characters may be. Aside from the 

 fish which spawn in the bays of Lower Californ- 

 ia in the fall (and which probablv represent a 

 special situation as is discussed elsewhere), 

 sardine spawning is largely confined (for what- 

 ever reason) to waters between 13°C and 16.5°C, 

 as measured at a depth of 10 meters. Thus, 

 there is a 3 .5° range over which eggs may be 

 deposited and which should lead to meristic 

 differences, at least at the extremes. If the 

 eggs are spawned during a period of no winds, 

 the temperature profile of the water should be 

 fairly constant and the immediate environment 

 of any particular egg should be comparatively 

 unchanging. However, if the temperature pro- 

 file is stratified (as it would be after an extended 

 period of calm), eggs from a single spawning 

 could be exposed to different temperatures ac- 

 cording to their depth distribution. 



There is the further situation wherein 

 wind stirring after spawning results in the 

 relatively rapid mixing of stratified water with 

 the subsequent reduction of temperature in the 

 upper part of the stirred layer and an increase 

 in the lower part of the stirred layer. Experi- 

 ments on other fishes have shown that sudden 



temperature changes during certain sensitive 

 periods during development can produce large 

 changes in meristic characters. 



Thus, it is apparent that, on the one hand, 

 identical environmental conditions may exist at 

 points widely separated in space and time (see 

 above and Ahlstrom, 1954) and that, on the other, 

 eggs from a single spawning may be exposed to 

 a variety of environmental conditions including 

 rapidly changing conditions . 



Tagging experiments: Pilchard tagging 

 experiments (using internal tags) were started 

 in British Columbia in 1935, in California in 

 1936, and in Oregon in 1937. These experiments 

 have been discussed by Brock (1940), Clanc and 

 Janssen (1945) Hart (1937, 1938, 1938, 1938, 

 1938, 1939, 1940, 1941, 1942, 1943, 1943, 1944, 

 1945, 1945>, Hoy (1938). Janssen (1937, 1938, 

 1938, 1939, 1939), Janssen and Aplin (1945). 

 andShuman (1939). 



The results of most of these experiments 

 have been summarized by Clark and Janssen ( 1945) . 

 They conclude (p . 41): 



"Fish tagged by the California Division 

 of Fish and Game in Mexican waters have been 

 retaken in the California fishery off San Diego, 

 San Pedro, Monterey and San Francisco. Sar- 

 dines tagged in central and southern California 

 have been recovered in the California fisheries 

 and in the Pacific northwest. Fish marked in 

 the Pacific northwest by the Canadian and Oregon 

 governments have been recovered in the Calif- 

 ornia fisheries . 



"The rapidity of movement of tagged fish 

 varies with size. The largest sardines will 

 migrate from southern California to British 

 Columbia in five to six months and fish tagged 

 off British Columbia have made the southern 

 migration in the same time interval . Smaller 

 fish do not move as rapidly and several years 

 will elapse before an average -size lot of tagged 

 sardines has spread to distant fishing grounds." 



Clark and Marr (1955:32-33) have recently 

 re-examined the tagging data and have given the 

 following summary: 



"(a) Fish tagged at any locality from 



112 



