represented. Also, unusually larye catches tend to bias presentations based 

 upon pooled data. 



Although substantial efforts were made to conduct each survey at 

 approximately the same time each year, differences affecting distribution did 

 occur. These were due to: (1) unavoidable differences in calendar 

 scheduling; (2) differences in climatic conditions; (3) the size of the 

 stocKs; and (4) the availability of food items. Since the data have been 

 pooled, annual differences in distribution cannot be discerned from the 

 graphics presented in this report. Also, for the same reason, anomalous 

 distributional patterns may appear as normal when all years are plotted 

 together. For these reasons interpretation of the distribution plots is 

 important and has been done for each species. In some cases two spring 

 distribution plots have been included—one with the entire series from 1968 to 

 1979 (Figure 4) and another with data from 1976 to 1979 (Figure 5). Several 

 spring cruises prior to 1976 were conducted as much as four to six weeks later 

 than the normal spring survey period. For some species this difference 

 occurred during a critical period of their migration and the plots from the 

 longer series reflected what appeared to be an anomalous distribution caused 

 by this temporal bias. By deleting data prior to 1976 the temporal bias was 

 eliminated and resulting plots can be considered a truer representation of 

 spring distribution. Figure 6 is a plot of all trawl stations made during the 

 autumn series 1967-79. 



The results of the LSTB and LSTL analyses have been weighted to take the 

 differing sizes of the strata into account. The area of each stratum (in 

 square nautical miles) was used as the weighting factor. 



