FLUCTUATIONS IN ABUNDANCE OF MARINE FISHES 183 



up until 1944-1948, when a drop to only 130,000 tons occurred in the 1947-48 

 season. Although the complete story will not be told until the catches of the next 

 few decades are on record, it is already apparent from the increased catches in 

 the 1948-1949 and 1949-1950 seasons that the fishery is beginning to recover. Be- 

 fore any intelligent study of the fluctuations in the abundance of marine fisheries 

 can be initiated, it is necessary to find some method of measiiring abundance. 

 This procedure constitutes the subject of the next section. 



Measurement 



Selection of Method and Unit. In measuring the abundance of marine fish it 

 must first be decided whether this abundance is to be expressed in terms of 

 the number or the total weight of fish. Both methods have been used, with perhaps 

 somewhat more emphasis on the former. It is also necessary to define the "stock" 

 of fish. Very few fishery investigators consider this to be the total of all fish of a 

 given species in an area studied at any given time. It is much more common and 

 somewhat more practicable to measure the commercially available part of the 

 stock, comprised normally of the larger and sexually mature members of the 

 population. The division between "precommercial" or "juvenile" and commercial- 

 sized fish is never sharply defined, and usually depends upon size at matiu-ity and 

 the amount of size selection exercised by the gear used. Thus, gill nets, for in- 

 stance, are very selective in their fishing action, whereas seines and traps are 

 much less so. Although illustrations of the fluctuations in the catches of marine 

 fish have been given, it must be emphasized that the actual catch is very seldom 

 a measure of the abundance of the fish. 



The catch data, however, when used in connection with data on the amount of 

 fishing effort expended (amount of gear and time of operation), provide a 

 measure of catch per unit of effort, which is roughly indicative of abundance. 

 This and other methods of estimating abundance will be discussed, but it is 

 pertinent to note that in all cases the sampling problem is of particular importance. 

 Except for the actual enumeration of the total population the estimate depends 

 upon some type of sample, whether it is one of scales or lengths of fish, of catches 

 from the stock or of fish eggs. In making estimates from these samples it is assumed 

 that they are representative of the population as a whole, but unfortunately this 

 assumption is often unwarranted. There are numerous opportunities for statistical 

 error, and only by the greatest care and critical examination of the biostatistical 

 methods used can gross inaccuracies be avoided. It is of the utmost importance 

 to face the sampling problem during the planning stages of any investigation 

 and to make sure that, while feasible with the facilities available, it is one which 

 will give representative results. 



Actual Counting. Simple enumeration is, of course, the most accurate method 

 of determining the abundance of the stock at any particular time, but this method 

 is applicable to very few marine fishes because of their inaccessibility. In fact the 

 only ones that can be treated in this way are the anadromous species which mi- 

 grate up streams and rivers during the spawning period. The method has been 

 used particularly in connection with investigations of Pacific and Atlantic salmons. 

 Fences or "weirs" are actually thrown across streams so as to completely block 

 upstream migration. One or more gates, equipped with white boards over which 

 the fish may be seen clearly as they pass through, are provided. Talliers, trained 



