FISHERY BULLETIN: VOL. 83. NO. 4 



Table 3.— Canonical variable (CV) loadings for fall 1976 and spring 1978 gradient analyses, with canonical correlation coefficients (Re), 

 amount of variation explained by each canonical axis (% variation), and total variation in species distribution explained by the environmen- 

 tal data. 



but total effort on Georges Bank increased several 

 times (Fig. 7). The assemblage trends examined in 

 this paper can be linked to these high levels of effort. 

 During this time period, seasonal bottom trawl 

 surveys monitored trends in finfish abundance over 

 the area from the Gulf of Maine to Cape Hatteras. 

 This survey proved invaluable to fish stock assess- 

 ment work because changes in the relative abun- 

 dance of most of the commercial species were follow- 

 ed closely and were highly correlated with commer- 

 cial catch, effort, and other indices (Clark 1979). 

 Other species of ecological, perhaps not commercial 

 importance, were also routinely and closely 

 monitored over this time The spring and fall bot- 

 tom trawl survey provided an excellent means for 



assessing community or assemblage responses over 

 this time period. 



Cluster analysis, with the Bray-Curtis dissimilarity 

 index and group average fusion method, proved 

 helpful for defining demersal fish assemblages on 

 Georges Bank. Recent studies confirm the value and 

 applicability of the Bray-Curtis index (Bloom 1981). 

 This method provided a means for collapsing the 

 multidimensional nature of the spring and fall 

 Georges Bank survey cruises into smaller, more 

 easily interpreted, units. It was then possible to 

 investigate not only long-term temporal and spatial 

 persistence questions, but also intraspecific 

 responses within the particular assemblage of 

 interest. 



518 



