FISHERY BULLETIN: VOL. 82, NO. 2 



in nets makes a similar analysis of the spring 

 cruises tenuous at best, we will not consider the 

 topic further other than to note that average 

 abundance and biomass were higher in the north- 

 ern and inshore portion of the study area during 

 both seasons (Tables 1, 2). 



CONCLUSIONS 



Despite large variation in the abundances of 

 individual species, cluster analyses of 9 yr of sur- 

 vey data have shown clear and consistent patterns 

 of community composition and distribution among 

 demersal fishes of the Middle Atlantic continental 

 shelf. Allowing for thermal variation and misclas- 

 sification of small catches, persistent site and 

 species clusters have indicated the presence of four 

 relatively constant and well-defined areas of 

 faunal homogeneity in the spring and five more 

 general areas in the fall, and five strongly recur- 

 ring species associations during both seasons. 



The spring site groups can be described approx- 

 imately as northern inner- and mid-shelf (I), ex- 

 tending from shore out to about 60-80 m from Cape 

 Cod to south of Delaware Bay; northern mid-shelf 

 (ID, occupying from around 60-80 m out to about 

 150 m from Cape Cod to Hudson Canyon; southern 

 outer shelf (III), 60-150 m, from Delaware Bay to 

 Cape Hatteras; and outer shelf-shelf break (IV), 

 >150 m. The southern inner and mid-shelf is a 

 thermally related transition zone between groups 

 I and III. The outer shelf between Delaware Bay 

 and Hudson Canyon was also a transition zone 

 (between groups II and III), but this discontinuity 

 does not appear to be related directly to tempera- 

 ture, but rather to the extent to which the north- 

 ward migration of the warm-temperate species 

 group has progressed by the time of the survey 



The five spring species groups contained one 

 group specific to this season and four which con- 

 tained common elements and properties with 

 analogous fall groups. The first group (A) can be 

 characterized as highly cryophilic, being virtually 

 restricted to site group I and containing two mem- 

 bers [Gadus morhua and Hemitripterus 

 americanus) which were relatively absent from 

 the study area during the fall. None of these 

 species were major dominants, even within group 

 I. The second group (B) is also composed of primar- 

 ily boreal, cold-water species, but in this case is not 

 completely restricted to site group I (although 

 primarily distributed there) and contains the 

 major dominant for that site, Limanda ferruginea . 

 The third group (C) may be described as ubiqui- 



tous throughout the study area with moderate or 

 better constancy to all site groups (Fig. 9). All 

 members of this group are boreal or resident, and 

 the major dominants, Merluccius bilinearis and 

 Squalus acanthias , are the nuclear members. The 

 fourth group (D) is composed entirely of warm- 

 temperate members and is restricted to the 

 warmer southern and outer shelf waters (site 

 groups II-IV). Peprilus triacanthus and 

 Stenotomus chrysops are the major dominants 

 from this group. The last group (E) is composed 

 strictly of weakly dominant slope species mostly 

 confined to the shelf break site group (IV). 



The spring warming trend noted during the 

 study period appeared to have no major effect on 

 the composition and distribution of fish com- 

 munities in the area other than the latitudinal 

 division between the inshore site groupings. The 

 results of the present study are very much in ac- 

 cordance with the conclusions of Taylor et al. 

 (1957) and Colton (1972) who found that while the 

 ranges and distributions of certain species did 

 shift with a changing thermal regime, there were 

 no obvious overall changes in faunal composition. 

 This is understandable when one considers that 

 the average change encountered (about 2°C) is 

 relatively small compared with the temperature 

 tolerances of the species involved and the seasonal 

 and geographic temperature variation encoun- 

 tered. 



The five fall site groups can best be described as 

 southern inner- and mid-shelf (I), extending out to 

 about 60 m from Cape Hatteras to Delaware Bay 

 and containing the area of warmest temperatures; 

 northern inner shelf (II), extending northward 

 from group I along a similar depth regime and 

 containing cooler waters; northern mid-shelf (III), 

 extending from group II out to about 90 m and 

 occupying the area of the cold pool; outer shelf 

 (IV), occupying the area between groups I and III 

 and about 150 m; and shelf break (V), >150 m. 

 While, again, with these groups there is some 

 overlap (particularly with groups I and II as dis- 

 cussed above), their definition is fairly good con- 

 sidering the rapidly changing environmental con- 

 ditions and migratory activity offish during this 

 period. 



The fall species associations, as noted above, 

 have much in common with those noted in the 

 spring. The small cryophilic group is absent, but 

 the terms applied to the other four spring groups 

 may be applied here as well. An exclusively 

 boreal-resident group (B) persists on the northern 

 inner- and mid-shelf, including four members of 



310 



