In the second example, the tags (20,000) are put out in proportion to the distribution of the 

 population of 2 x 10 fish which, again, does not change. There is complete mixing of tagged fish 

 between the three groups and subsequent exchange of tagged fish between areas . The tagged fish 

 are randomly distributed within the population in each area. The recoveries are made in propor- 

 tion to the distribution of the population. A catch of 2,000 tons will yield 200 recoveries, but in 

 this example the number recovered in some areas does not meet the minimum requirement of 10. 

 In order to achieve this it will be necessary to increase the number of tags put out in the southern 

 California offshore area from 1,000 to 4,000 and also to increase the catch in the same area from 

 100 to 500 tons (thus increasing the recoveries in that area from 10 to 57.5). 



METHOD AND COST 



Practically all tagging experiments that have been conducted have depended upon the commer- 

 cial fishery as a source of fish to tag and as a source of recoveries . In general, such experiments 

 have not been designed to answer the type of question I have posed, but rather to gain information 

 on migrations, mortality rates or population sizes. In fact, considering the fact that most fisheries 

 are not carried out on actively spawning fishes, such experiments could not have been designed to 

 answer the subpopulation question. A notable, and possibly fortuitous, exception is the Pacific 

 herring tagging experim ent (Tester, 1949) . 



The U.S. sardine fishery is similar to most other fisheries i"n that the location of catch in 

 time and space is different from the location of spawning. Thus, the question of how fish would be 

 caught for tagging and for tag recovery naturally arises . Catching fish for tagging might be ac- 

 complished by the use of a lift-net similar to the one used on the M/V Yellowfin (Radovich and 

 Gibbs, 1954). Another and more likely possibility is the use of purse seiners (on a charter basis). 

 The recovery of tagged fish would almost certainly have to be made by seiners because of (1) the 

 quantity involved and (2) the necessity for working offshore in seas too rough to permit operation of 

 the lift -net. 



Two problems are involved here. First, is it possible to catch sardines with purse seines 

 on the spawning grounds ? There are some indications from the Mexican fishery that this is possible. 

 It would be necessary to firmly establish this (or demonstrate that it cannot be done) . Second, 

 granted that the fish could be caught on the spawning grounds in sufficient quantity, how could the 

 tags be recovered at a time when the U.S. processing plants are not ordinarily operating? Special 

 arrangements would have to be made with both U.S. and Mexican plants for their cooperation . 

 Furthermore, it would not be feasible to depend upon magnets in meal plants for the tag recoveries 

 as was done in previous sardine tagging experiments (see Fry, 1937). Instead, it would be desir- 

 able to use an electronic metal detector (see Dahlgren, 1936) with which individual tagged fish could 

 be recovered. 



The cost estimates for such a tagging experiment are discouragingly high. I estimate that, 

 including the necessary vessel charter, the first year's operation would cost in the order of $60, 000. 

 Assuming that feasibility and methods could be reasonably well worked out by the end of the first 

 year, full-scale operations in subsequent years would cost about $120, 000 per year. 



SOME LIMITATIONS 



Despite all the desirable aspects of a tagging experiment, there are limitations. These are, 

 however, limitations of implementation rather than of assumptions or inferences inherent in the 

 method. I indicated above that the examples given are of an oversimplified model, and indeed they 

 are. There are a number of obvious modifications that must be made before the model is reason- 

 ably realistic . Foremost among these is an accounting of the loss of tagged fish through mortality 

 and the "dilution" of tags in the total population by the recruitment of a new year -class into the adult 

 population . 



124 



