Pooling of Within-Season Cruises 



Size of the data matrix was too large for simul- 

 taneous clustering of either of the two multiple- 

 year seasonal data sets. However, the results of the 

 cluster, nodal, and dominance analyses of the 

 individual cruises revealed a high degree of 

 within-season repetition in the composition and 

 distribution of species groups and in the faunal, 

 geographic, and hydrographic attributes of site 

 groups. Major repetitive species groups were rec- 

 ognized for each season, and site groups for each 

 year were referred to generalized seasonal site 

 groups. The validity of these groups was examined 

 by subjecting the pooled seasonal data sets to 

 nodal and dominance analyses based on these 

 groupings and comparing these results to those for 

 the individual cruises. 



FISHERY BULLETIN: VOL. 82, NO. 2 



RESULTS AND DISCUSSION 



Thermal Regime 



The geographic patterns of bottom-water tem- 

 peratures were variable among years within both 

 of the sampling seasons, although these differ- 

 ences were minor compared with the seasonal vari- 

 ation within a given year. Variability among years 

 within a season can be attributed to two sources: 

 Climatic differences among years and sampling 

 artifacts (i.e., differences in the dates and duration 

 of the sampling periods, and stochastic differences 

 arising from the location of stations and the tem- 

 poral sequence in which they were done). 



The spring cruises were conducted in March and 

 April, the period during which water tempera- 

 tures in the Middle Atlantic Bight are at a mini- 



"T 



^TfOO 



- 



FIGURE 2.— Bottom isotherms for spring 1969 (A) and 1976 



298 



