DIFFERENTIATION OF MITOCHONDRIAL DNA IN 

 ATLANTIC HERRING, CLUPEA HARENGUS 



I. KORNFIELD AND S. M. BOGDANOWICZ' 



ABSTRACT 



To investigate genetic relationships among spawning stocks of Atlantic herring, Clupea harengus, in 

 the Gulf of Maine and Gulf of St. Lawrence, mitochondrial DNAs from ripe females at three localities 

 were examined by restriction endonuclease analysis. Using seven variable restriction enzymes, 

 mtDNAs from 69 completely characterized individuals produced 26 composite digestion patterns. The 

 majority of individuals (65%) possessed composites which were common to two or more spawning 

 localities; the other individuals displayed locality specific "unique" composites. Analysis of relation- 

 ships among these unique composites suggested that some may have been derived from other areas. 

 These results are not consistent with the idea that separate genetic stocks of Atlantic herring exist 

 in the Gulf of Maine. 



The relationships among discrete spawning 

 stocks of Atlantic herring, Clupea harengus , are 

 problematical. A large number of stocks and stock 

 complexes are recognized throughout the eastern 

 and western North Atlantic; these delineations 

 are based largely on meristic characters, spawn- 

 ing time, and spawning location. Tagging studies 

 in the western North Atlantic have shown exten- 

 sive migration and mixing of stocks during nonre- 

 productive periods (Creaser et al. 1984). More 

 limited studies of spawning fish have demon- 

 strated that some tagged individuals returned to 

 their spawning locations (Wheeler and Winters 

 1984). Recent work has advanced the hypothesis 

 that specific environmental attributes essential 

 for growth and survival of larval herring largely 

 determine where Atlantic herring will spawn 

 (lies and Sinclair 1982). The notion that spawn- 

 ing occurs near areas suitable for larval retention 

 could explain the discontinuous or patchy distri- 

 bution of spawning areas. Similarly, the occur- 

 rence of fall spawning and spring spawning At- 

 lantic herring stocks may be a function of 

 completion of larvae growth and metamorphosis 

 constrained by resources within the larval reten- 

 tion area (Sinclair and Temblay 1984). There is 

 thus a reasonable model to explain the existence 

 of geographically or temporally discrete spawn- 

 ing stocks. However, the genetic structure among 

 these different spawning groups is unresolved. 

 Implicit in the Atlantic herring stock concept is 



'Department of Zoology and Center for Marine Studies, Uni- 

 versity of Maine, Orono, ME 04469. 



Manuscript accepted March 1987. 



FISHERY BULLETIN: VOL. 85, NO. 3, 1987. 



the idea that individual fish belong to defined 

 groups by virtue of returning to specific spawning 

 sites. If this is the case, there should exist high 

 genetic continuity among individual Atlantic 

 herring within stocks and relatively lower conti- 

 nuity among stocks. That is, genetic differences 

 should be observable among stocks. Unfortu- 

 nately, meristic characters useful for stock defini- 

 tion are under environmental influence and have 

 a complex genetic basis. Electrophoretic charac- 

 terization of allozyme variation should poten- 

 tially permit identification of genetic discontinu- 

 ities among stocks. However, despite the 

 availability of a large number of polymorphic 

 markers and adequate sample sizes, significant 

 genetic heterogeneity among Atlantic herring 

 stocks has not been demonstrated (Anderson et 

 al. 1981; Kornfield et al. 1982; Grant 1984; Riv- 

 iere et al. 1985). The inability of allozyme analy- 

 sis to differentiate among herring stocks could 

 occur for two alternative reasons. Herring stocks 

 could have originated so recently that there has 

 been insufficient time for stock specific allozyme 

 variation to accumulate. Further, natural selec- 

 tion may be acting to homogenize allele frequen- 

 cies that may characterize stocks. Thus, standard 

 allozyme analyses may not be sufficiently sensi- 

 tive to detect genetic variation which distin- 

 guishes stocks. Alternatively, herring stocks 

 could be largely composed of individuals that do 

 not return to natal spawning sites. Under this 

 explanation, herring stocks would not represent 

 discrete genetic groups but rather random assem- 

 blages of spawning individuals. Management of 



561 



