ll 

 If a total mortality rate between years of 50 percent is assumed^ , then obviously, of the 



20,000 tags put out in the initial season, only 10,000 will remain in the first season after tagging. 



Similarly, the total population will be reduced from 2 x 10" to 1 x 10^ fish. If, however, the 



entering year-class approximates the size of the population to which it is added (again, see Clark 



and Marr, 1955:22), the total population will again consist of 2 x 10 fish. Thus, while the total 



population maintains the same size, the number of tagged members has been reduced by one -half . 



As a consequence, either the number of tags originally put out would have to be doubled or the 



catch would have to be doubled (or some appropriate combination of increase in the number of 



tags put out and in catch) . 



As stated above, a total spawning population size of 2 x 10^ fish was assumed in the simple 

 examples given. Any increase in total population size would require a corresponding increase in 

 the number of tags put out and/or the catch. The most recent estimates of total population size 

 (Clark and Marr, 1955:19) indicate a population about four times as large as that used in the 

 examples. 



In the discussion thus far the assumption has been made that if there is gene flow throughout 

 the population it is by the mechanism of adults spawning in two or more spawning areas in the 

 same or successive seasons. There is an alternative mechanism; namely, that the fish produced 

 in a given area may themselves spawn in a different area. If the first mechanism is found to 

 obtain, the second may be ignored. If, on the other hand, evidence is negative for the first, then 

 it will be necessary to investigate the second. This will involve determining the smallest size at 

 which sardines may be successfully tagged and being able to associate such fish with an area of 

 origin . 



CONCLUSIONS 



Of all the methods now known, a tagging experiment is capable of yielding the most 

 definitive answers. However, even in simple situations such an experiment is very costly. And, 

 as the simple examples are made more realistic, all tendencies are toward increasing the magni- 

 tude, and thus the cost, of the task. 



Until the possibilities of biochemical methods are fully explored, it is reasonable to defer 

 a full-scale tagging experiment. Results obtained by other methods, however, may eventually have 

 to be checked by a full-scale tagging experiment. There is the further possibility that results 

 obtained by other methods may delimit more or less discrete problems which can be resolved by 

 considerably less than a full-scale tagging experiment. 



2/ A not unreasonable assumption; see Clark and Marr (1955:26). In this instance it matters 

 not what fraction of the total mortality is from natural causes and what part is from the regular 

 fishery. Tags recovered from the fishery will have no bearing on the solution of the subpopulation 

 problem, although they may be of extreme value in permitting estimates of mortality rates and 

 population size and also in the study of movements. 



125 



