PROGRESS IN BIOLOGICAL INQUIRIES. 1937 13 



indicates that the exploratory trip in 1935 did not provide a repre- 

 sentative sample (in respect to age and size composition) of the 

 Georges Bank population, but that the 1936 data were much better. 

 The work clearly demonstrates that the commercial fishery does not 

 sample all sizes of marketable haddock in equal proportions. Owing 

 to the price differential between large and scrod haddock, thei former 

 generally are over-samj)led. It is concluded that field work of the 

 kind carried on from the Atlantis is absolutely essential to supple- 

 ment data from the commercial catch. Such woi'k makes possible an 

 estimation of the relative abundance of 1-, 2-, and 3-year haddock 

 which are not sampled b}" the commercial fishery ; provides informa- 

 tion concerning the haddock populations in areas not fished by the 

 trawling fleet because the fish are too small or too scattered; and 

 gives some check on the accuracy with which the commercial catch 

 samples the population. In addition, many valuable data are ob- 

 tined on growth rate, sex ratio, and distribution. 



Considerable progress Mas made in combining and summarizing 

 length-frequency data collected in 1936 and 1937. One-month and 

 4-month summaries by subareas have been prepared so that the data 

 now are ready to be broken down into age classes. Analysis of scale 

 samples was confhied to the Atlmitis collections and contributed to 

 the results discussed above. During the work on length frequencies, 

 a careful analysis was made to determine the most satisfactory 

 method for sampling the commercial catch and for combining the 

 samples. The results of this investigation were reported in a paper 

 read before the American Statistical Association. 



Prospects for 1938. — In Progress in Biological Inquiries, 1936, it 

 was stated that in 1937 a moderate decline in catch per unit of 

 effort on Georges Bank might be expected while the total catch 

 should increase somewhat. The catch per unit actually decreased 

 about 20 percent (8 months' data) while the total catch increased 10 

 percent. For the Nova Scotian banks it was predicted that there 

 would be a considerable decline both in catch per unit of effort and 

 in total catch. The catch per unit actually decreased about 26 per- 

 cent (8 months' data) while the total catch decreased about 25 

 percent. 



In view of the facts that analysis of 1937 data is not yet completed 

 and that data on the abundance of the 1936 class (which will reach 

 marketable size in 1938) are lacking, it is impossible to make definite 

 predictions for the 1938 season. The catch per day both in the 

 Georges Bank region and on the Nova Scotian banks will undoubtedly 

 be considerably less than in 1937 unless there is a great influx of 

 scrod during tlie summer and fall ( 1936 class on Georges Bank and 

 1934 class on the Nova Scotian banks). Unfortunately, the relative 

 abundance of these classes on the grounds mentioned is not known, 

 for the members of these groups were too small to be taken in signifi- 

 cant ]umil:)ers by the commercial trawlers during 1937 and no boat 

 was available for special trawling trips .such as wei-e made on the 

 Atlantis in 1935 and 1936. Unless the 1936 and 1934 classes api)ear 

 in uuusual abundance, it appears that the decline in catch per day 

 will be sufficient to cause a considerable drop in total haddock catch 

 in spite of the increase in the size of the fleet. 



