FISHERY BULLETIN: VOL. 70, NO. 3 



June to the end of September (Bigelow, 1927; 

 Parr, 1933). This barrier is marked by the 

 presence of a surface pool of cold upwelled water. 

 At other seasons of the year there is open con- 

 tinuity betvv^een coastal temperature conditions 

 north and south of Cape Cod. This temperature 

 barrier coincides with the poleward and equator- 

 ward boundaries of many species of marine 

 fauna, both pelagic and benthic (Bigelow and 

 Sears, 1939; Parker, 1948; Fritz, 1965; Schopf, 

 1965). 



The seasonal ranges of surface- and bottom- 

 water temperatures in most areas within the 

 Gulf of Maine ai'e less extreme than those char- 

 acteristic of the Middle Atlantic Bight. The 

 seasonal range of temperature in the shoal-water 

 area of Georges Bank, however, is as great as 

 that at similar depths in the area between Nan- 

 tucket Shoals and New Jersey (Colton and Stod- 

 dard, 1972 and In press). Parr (1933) has de- 

 scribed the ecological relationships between ho- 

 mothermal and heterothermal regions separated 

 by a seasonal temperature barrier and has dis- 

 cussed the effect that these contrasting temper- 

 ature regimes have on the period, extent, and 

 character of the migratory activities of shallow- 

 water fish populations. There is a marked sea- 

 sonal alternation in the composition of the fish 

 fauna of the Middle Atlantic Bight. In summer 

 this area contains a very rich population of 

 southern migratory fishes which follow the 

 northward dispersal of isotherms as far north 

 as Cape Cod. Winter temperature conditions 

 compel this population to vacate the area, and 

 it is replaced to a considerable extent by north- 

 ern species, such as herring and cod, which at 

 this season have free access to penetrate as far 

 south as the neighborhood of Cape Hatteras. 

 Although there is no thermal barrier during the 

 winter, the qualitative composition of the fish 

 fauna is not uniform from north to south around 

 ■Cape Cod. As Parr (1933) has pointed out, this 

 seasonal diffusion of northern populations into 

 areas where no significant temperature differ- 

 ences exist is distinguishable in both its char- 

 acter and effects from the northerly and south- 

 erly migrations of complete populations such as 

 occur under thermal compulsion within the 



Middle Atlantic Bight. In contrast to the Middle 

 Atlantic Bight which contains very few perm- 

 anent residents, the bulk of the fish fauna in the 

 Gulf of Maine are endemic to the area and there 

 is little seasonal alternation in species compo- 

 sition. 



Hutchins (1947) recognized four basic types 

 of temperature zonation which depend on max- 

 imum and minimum temperatures critical for 

 survival and repopulation. With the exception 

 of estuarine species such as the green crab which 

 may be subject to winter kill, the distribution of 

 benthic organisms in the area off New England 

 appears to be controlled by summer rather than 

 winter temperatures (Hutchins, 1947; Schopf, 

 1967) . The equatorward boundary of potentially 

 southward-migratory forms is limited by sum- 

 mer temperatures too high for survival and the 

 poleward boundary of northward-migratory 

 forms by summer temperatures too low for re- 

 production (type 3 of Hutchins, 1947). This is 

 in agreement with the assumption of Hutchins 

 (1947) that at localities having a temperature 

 barrier at only one season (summer at Cape 

 Cod), all distributions terminating there are 

 associated with that barrier. 



It is difficult to categorize fish species by strict 

 temperature zonation characteristics. Fish do 

 not have built-in integrators and so must react 

 to prevailing rather than to average conditions. 

 Unlike many benthic invertebrates, fish are able 

 to make short-term excursions and are not de- 

 pendent solely on passive migrations. It should 

 also be stressed that the distribution within the 

 broad limits set by temperature extremes for 

 survival and reproduction is controlled to some 

 extent by other factors such as food supply 

 (Blackburn, 1969) and substrate conditions 

 (Bigelow and Schroeder, 1953). Of the species 

 considered in this paper, only the American 

 plaice and the butterfish exhibited the expected 

 response to cooling, e.g., extension of the south- 

 ern range of northern endemic forms and dim- 

 inution of the northern range of southern mi- 

 grants. 



The butterfish is the only species considered 

 which shows a sensitivity to small changes in 

 tem]ierature and which migrates under the 

 thermal compulsion. Ripe butterfish are taken 



652 



