BUREAU OF FISHERIES XIX: 



the increase in landings of haddock at New Engiand ports, the variety 

 most nsed in the filleted and packaged form, which amounted to 37 

 per cent over 1926 and 84 per cent over 1922, when this method of 

 preparation was first employed. The shore fisheries of the Middle 

 Atlantic States appear to be declining in productiveness of certain 

 important species, judged by a comparison of 1926 statistics with 

 those for 1921, the last previous canvass in this section. The catches 

 of bluefish, scup, and squeteagues suffered decreases of 72, 37, and 36 

 per cent, respectively, but the catch of butterfish increased 18 per 

 cent. The shellfisheries registered moderately increased production. 



The fish-canning industry of the Atlantic coast in 1927 produced 

 the largest pack of shrimp on record and an increased oyster pack. 

 The sardine pack of Maine was 26 per cent smaller than in the 

 l^revious year. The menhaden industry experienced its third poor 

 year in the last four years, with an output only half as large as has- 

 been considered normal in recent j^ears. 



On the Pacific coast the canned-salmon pack of 1927 was curtailed 

 considerably through scarcity of fish. The California sardine pack 

 continued to grow, being six times as large in 1927 as in 1921, when 

 production was at a low level during the postwar depression. The 

 canned-tuna output increased, chiefly because more of the striped and 

 yellowfin varieties were packed. The halibut catch was slightly 

 higher in 1927, in spite of the depleted condition of this fishery, but 

 the increase probably can be accounted for by the greater intensity^ 

 of fishing and its extension to the more westerly grounds. 



COLLECTION OF STATISTICS 



The statistical program of the bureau is most intimately related to 

 its primary function — discovering the depletion of food fishes, its 

 causes and possible remedies. Inasmuch as statistical information is 

 fundamental in the science of fisheries conservation, the bureau has 

 striven to improve this branch of its work, which is admittedly inade- 

 quate at present. During the past four decades such statistics have 

 been collected by canvassing the fisheries of nine sfeographical sections 

 of the country, taking one at a time and completing the work as 

 rapidly as possible with the personnel available. In recent years it 

 has been possible to canvass each section about once every five years. 

 This method has tAvo fundamental defects. In the first place, the 

 fishermen are reached a considerable time after the previous year's 

 business has been closed, and unless they have kept a record of their 

 operations and catch (a rare circumstance) the information received 

 is but an approximate estimate rather than a definite record. Sec- 

 ondly, the fortunes of a fishery fluctuate so widely from year to year 

 that data at five-year intervals are likely to be misleading, for they 

 may represent a poor or a good year rather than a normal one. 

 Annual statistics are essential to indicate accurately trends in fishery 

 matters. 



While the bureau's facilities are not adequate to collect annual 

 statistics for the entire country, fortunately this may be done through 

 cooperation with State agencies. Since early times the fisheries in 

 State waters have been under State control, and several States have 

 included the rendering of statistical returns as an obligation of com- 



