The differences are certainly markedly less 

 than those found in the Pacific herring (51.68 

 vertebrae in San Diego Bay as compared to 

 55.67 vertebrae in samples from the Shumagin 

 Islands as reported by Rounsefell 1930), yet no 

 similar subdivision of the herring populations 

 into subspecies has been made. 



In a similar study of the Australian pil- 

 chard, Blackburn (1951) differentiated three 

 major groups which he designated as "races' . 

 The eastern and southeastern groups of pilchards 

 were delineated like the corresponding groups 

 of anchovies; the third group of pilchards occu- 

 pied waters off southern Western Australia . 

 The three races" were separated on the basis 

 of growth rate and spawning seasons . The 

 eastern and southeastern races were subdivided 

 into smaller, more or less separate stocks. A 

 major difficulty in studying subpopulations in 

 Australian pilchard and anchovy has been the 

 virtual absence of a commercial fishery for 

 these species . 



SUMMARY 



In this paper I have reviewed research on 

 "subpopulations" of eight Pacific fishes. The 

 species dealt with were the following: 



Pacific herring 



Yellowfin tuna 



Albacore 



Pacific mackerel 



Northern anchovy 



Anchovetta 



Nehu 



Australian anchovy 



(Clupea pallasi) 

 ( Neothunnus macropterus) 

 (Germo alalunga) 

 (Pneumatophorus diego) 

 (Engraulis mordax) 

 (Cetengraulis mysticetus) 

 (Stolephorus purpureus) 

 (Engraulis australis) 



In none of the studies were the workers 

 able to distinguish between differences induced 

 by the environment and those which had a genetic - 

 al basis. Hence, the workers were dealing with 

 "groups" as defined by Marr, rather than with 

 "subpopulations" and "stocks". For all species, 

 the indirect approach has been utilized (meristic 

 and/or morphometric characters), and for five 

 of the species some tagging has been done. 

 However, for only two species has tagging been 

 extensive enough to be of value in subpopulation 

 studies, i.e., Pacific herring and Pacific mackerel . 



In studying subpopulations of clupeid and 

 engraulid fishes, most workers have relied on 

 meristic characters, especially vertebral counts. 

 In studying subpopulations of scombroid fishes 

 (tunas and mackerels), on the other hand, mer- 

 istic characters have shown too little variation 

 to be of value, hence techniques have been de- 

 veloped for utilizing morphometric characters . 



For all species studied it has been possible 

 to show "population heterogeneity" . However, it 

 has not been possible to determine the extent of 

 separateness of the various parts of a population 

 by the indirect approach . Even such advances 

 in this field as the "concept of overlap" discussed 

 in the paper by Royce do not entirely resolve 

 this difficulty. For determining the extent of 

 intermingling there appears to be no substitute 

 for effective tagging experiments . 



I have summarized the techniques used in 

 studying subpopulations of the species discussed 

 in this paper in the following table: 



69 



