Figure 1. — Middle Atlantic Bight boundaries used in analysis of bottom-water temperature data. (Dots represent typical distribution of ran- 

 domly selected bottom temperature stations.) 



each cruise are based on the percentage of survey area 

 represented by 2°C temperature contour intervals or 

 temperature class intervals (TCI's) (Figs. 2, 3). The 

 midpoint temperature of each TCI was multiplied by 

 the percentage area within that TCI and their total 

 product was divided by 100 to obtain a mean index for 

 each cruise (Tables 1-4). The derived indices are based 

 solely on an areal basis, but the stratified random 

 sampling design allowed representative sampling by 

 depth and geographic areas during each cruise. Since no 

 attempt was made to determine absolute temperature 

 indices, and the same geographic areas were randomly 

 sampled during each cruise, the method chosen for 

 determining relative temperature variations seems 

 justified. A more precise estimate of relative bottom 

 temperatures and elimination of possible bias associ- 

 ated with depth may have been analyzed. However, 

 analytical data was not adequate for all the cruises and 

 my major objective was to determine general tempera- 

 ture indices for rather large geographical areas during 

 the entire seasonal time-series of data collection. 



In order to determine year-to-year variation, the 

 observed indices were adjusted by basing individual in- 

 dices on a time-series mean sampling or reference date 

 for each season (Table 5). Based on charts from Colton 

 and Stoddard (1973) these seasonal reference dates 

 closely approximate the timing of the annual minimum 

 and maximum bottom-water temperature conditions 

 expected in both subareas. Plots of annual bottom-water 

 temperature cycles based on average temperature data 

 by 10-day intervals at Nantucket Shoals and Five 

 Fathom Bank Lightships during 1956-70 from Chase 

 (1972) were used to calculate adjusted temperature 

 values for the Northern Bight and the Southern Bight, 

 respectively. The mean temperatures observed on each 



cruise were adjusted by adding or subtracting the dif- 

 ference in the mean temperatures at the appropriate 

 lightship on the seasonal reference date and on the mid- 

 date of each cruise. The magnitudes of the adjustments 

 were usually ±0.2°C for the Northern Bight and ±0.5°C 

 for the Southern Bight, but differences as great as 2.8°C 

 were noted in 1970 when three cruise mid-sampling 

 dates were approximately 1 mo earlier or later than 

 their respective seasonal reference dates (Tables 5, 6, 8; 

 Figs. 4, 7, 8). These large differences between observed 

 and adjusted data clearly indicated the need for the 

 adjustments in order to make meaningful comparisons 

 of annual temperature variations. Since a compre- 

 hensive analysis of climatological and/or hydrographi- 

 cal conditions which might have revealed the factors re- 

 sponsible for the large differences between observed and 

 adjusted data was beyond the scope of my investi- 

 gation, these differences are attributed to the asynoptic 

 seasonal sampling in certain years, e.g., 1970 (Table 5). 



The observed indices may be considered in situ 

 measurements of the thermal environment and are 

 particularly relevant to synoptic biological observa- 

 tions. The primary objective of the adjusted indices is to 

 measure the relative annual variations that have occur- 

 red during each spring and autumn in order to reveal 

 any warming or cooling trends that may have occurred. 



Species distribution charts and catches by numbers 

 and pounds landed per l A h tow observed during the 

 same research cruises that temperature data were 

 collected since 1967 were examined to determine any 

 possible effects of the observed temperature changes. 

 Examples of several of these relationships are briefly de- 

 scribed in the Discussion section. Further analyses of 

 these observations are currently being conducted by 

 other members of the Woods Hole Laboratory staff. 



