246 
BULLETIN OF THE BUREAU OF FISHERIES 
In Shuyak Strait the schools taken up to 1927 were made up almost entirely of 
large mature herring, many of them attaining a length of 14 inches. In 1928 only 
schools of small herring appeared. Herring of the same large size occur in Kachemak 
Bay in August, but a month later they are mixed somewhat with smaller sizes. 
At Unalaska in 1928 the herring were very large, many of them being almost 15 
inches. A very few persons have mentioned to the writer that they measured central 
Alaska herring that were 16 inches in length, but these were apparently exceedingly 
rare. 
INDEPENDENCE OF AREAS 
METHODS OF STUDY 
Alarming changes in abundance have occurred in many localities. For instance, 
in Yes Bay and Kootznahoo Inlet in southeastern Alaska, herring were exceedingly 
abundant during the early years of the fishery, but soon became scarce and have 
remained so for many years. What caused the failure of such fisheries? 
In studying these changes in abundance the first question raised is whether they 
are due to depletion or to variations in the migratory habits. Do the herring of the 
Alaska coast belong to one population that moves about at will, striking the coast 
wherever or whenever conditions impel it, or is each locality inhabited by a local 
stock that mingles but slightly, if at all, with the populations of adjacent areas? 
In the first case, fishing at any point on the coast would affect the supply at every 
other point, and if depletion occurred it would be general. In the second case, 
intensive fishing in one locality would not endanger the supply elsewhere, but the 
danger of local depletion would be highly intensified. The movement of the herring 
schools therefore becomes of great importance. 
A study of these movements by the direct method of tagging the individual fish 
and obtaining records from the fishermen of the places of recapture has been suc- 
cessful in the cases of the plaice, cod, mackerel, salmon, halibut, and other fishes, but 
never in any of the clupeoid fishes. The ease with which impounded herring become 
infected with fungous growths, even when only a few scales have been lost, makes it 
appear probable that any wounds made in tagging might easily lead to death. Detec- 
tion of the tagged individuals among the great numbers of the species taken at one 
time would be difficult enough without such heightened mortality. In the case of the 
Pacific herring it was felt that if attempted, tagging would never be successful except 
perhaps in a few localities. In southeastern Alaska and Prince William Sound, where 
the principal fisheries are conducted, a very large proportion of the herring are used 
for reduction purposes, and, since these fish are not handled individually, the chances 
of detecting a tagged fish, even if recaptured, are infinitely small. Because of these 
reasons, tagging was not seriously considered as a feasible means of studying migration 
in the herring and indirect methods were concentrated upon. However, owing to 
the concrete results to be attained by successful tagging we were unwilling to relin- 
quish this method of investigation without giving it some trial. 
Halibut Cove, Kachemak Bay, was chosen as the most favorable location for a 
test. In this district practically the entire catch is pickled, insuring the detection of 
any recaptured fish, since in this method the fish are handled one by one. Accordingly 
in the spring of 1927, 3,071 herring of pickling size were tagged with No. 3 monel 
metal strap tags attached to the caudal peduncle, as in the salmon. 3 The results, 
however, were negative. 
; A full description of this type of tag, with illustrations, is given by Gilbert and Rich (1925). The No. 3 tag, [used on the her- 
ring, weighs[o.5[gram, is 3.3 millimeters in width, and about 14 millimeters In length when clinched. 
