SPECIES ASSOCIATIONS AND COMMUNITY COMPOSITION 



OF MIDDLE ATLANTIC BIGHT 



CONTINENTAL SHELF DEMERSAL FISHES 1 



J. A. Colvocoresses and J. A. Musick 2 



ABSTRACT 



Cluster analyses of seasonal NMFS Groundfish Survey bottom trawl catches on the Middle Atlantic 

 Bight continental shelf revealed consistent species associations and faunal zones over a 9-year period. 

 Boundaries between faunal zones tended to follow isotherms and isobaths. During the late winter-early 

 spring, the following faunal zones were found: Northern inner shelf, northern mid-shelf, southern 

 inner- and mid-shelf, and outer shelf-shelf break. Five species groups were identified: A small 

 cryophilic group restricted to the first zone, a cold-water boreal group found in the first two zones, a 

 ubiquitous boreal/resident group containing the major dominants, a warm-temperate group confined to 

 the warmer southern and outer shelf waters, and a group of slope residents confined to the deepest zone. 

 During the fall, five faunal zones were found: Southern inner/mid-shelf, northern inner shelf, northern 

 mid-shelf, outer shelf, and shelf break. Five species associations were largely analogous to those in the 

 spring, with the following exceptions: The cryophilic group was absent, the ubiquitous group contained 

 mixed boreal and warm-temperate elements, and a second outer shelf group was recognized. The most 

 notable change in the distribution of groups from the spring was a general northward shift and a 

 sharply defined inshore movement of the temperate group. 



Communities of fishes on the continental shelf 

 have rarely been studied beyond the compilation of 

 species lists for given areas. This is enigmatic 

 when one considers the large amount of survey 

 data that has been collected from much of the 

 world's continental shelf waters in connection 

 with fishery exploration and monitoring. While 

 trawl survey data have traditionally been col- 

 lected with the primary aim of assessing commer- 

 cially harvestable stocks, they also provide an 

 excellent base for evaluating the interspecific re- 

 lationships among trawlable organisms. 



The few studies which have previously ad- 

 dressed community structure of open continental 

 shelf fishes have found clearly definable species 

 associations with distributions related to en- 

 vironmental parameters. Demersal fish species 

 assemblages found using objective mathematical 

 measures have been described for the continental 

 shelves in the Gulf of Guinea ( Fager and Long- 

 hurst 1968), northwest Pacific coast of the United 

 States (Day and Pearcy 1968), and Campeche 

 Bank off Mexico (Sauskan and Ryzhov 1977). 



Since 1967 the National Marine Fisheries Ser- 

 vice (formerly Bureau of Commercial Fisheries) 



Contribution No. 1151 from the Virginia Institute of Marine 

 Science of the College of William and Mary. 



2 Virginia Institute of Marine Science and School of Marine 

 Science, College of William and Mary, Gloucester Point, VA 

 23062. 



Manuscript accepted September 1983. 

 FISHERY BULLETIN: VOL. 82, NO. 2, 1984. 



has conducted a semiannual bottom trawl survey 

 of the continental shelf waters from Nova Scotia to 

 Cape Hatteras (Grosslein 1969). This program has 

 produced a data base which offers a unique oppor- 

 tunity for the analysis of the composition and var- 

 iability of the fish communities in this region. 



In the present study, that portion of these data 

 collected in the Middle Atlantic Bight (Cape Cod 

 to Cape Hatteras) during the cruises from fall 

 1967 through spring 1976 were analyzed with the 

 aim of defining the composition of fish communi- 

 ties present within this area and how they vary 

 geographically, thermally, and seasonally. 



METHODS 



Sampling 



Groundfish Survey cruises were conducted by 

 the National Marine Fisheries Service during the 

 fall and spring from fall 1967 through spring 1976, 

 aboard either the RV Albatross IV or RV Delaware 

 II. The survey area extended from the 15-fathom 

 (27 m) contour offshore to 200 fathoms (365 m). A 

 stratified random sampling design was utilized, 

 based on depth and geographical zones (Fig. 1). 

 Catch data from strata 1-12 and 61-76 (Middle At- 

 lantic Bight) were analyzed in the present study. 

 Sampling intensity in each stratum was allocated 

 according to the geographic area of each stratum 



295 



30 



