184 MARINE PRODUCTS OF COMMERCE 



in identifying the various species of salmon, are stationed above the counting 

 boards and record the numbers of fish on mechanical counters. An outstanding 

 example of this type of operation is at Bonneville Dam on the Columbia River, 



At some places in Alaska and British Columbia the small downstream migrating 

 salmon have also been counted by use of fine mesh weirs. At such installations 

 the fish are either counted manually by dipping them a few at a time with a net 

 from one part of a trap to another, or by measuring the volume of the fish and 

 estimating the number from sample counts of known volumes of fish. Another 

 method also used recently in salmon work is to enumerate the spawning adults 

 by observations from airplanes. These are either direct visual counts or counts 

 made from photographs taken while the plane is in flight over the spawning 

 stream. The method is particularly valuable in open country, such as that found, 

 for instance, in the spawning areas of the red salmon around Bristol Bay, Alaska. 



A technique applicable only to the Pacific salmons (which all die after spawn- 

 ing) is the counting of the dead spawned-out fish. Obviously, this method is 

 limited to shallow streams where the dead fish will not disappear from view. 



Catch per Unit of Effort. This statistical measure, mentioned briefly above, is 

 one of the oldest and most widely used methods employed for estimating the 

 abundance of marine fish populations. Although the collection of the necessary 

 records is often extremely complicated and diflBcult, the actual calculation is 

 simple, consisting of merely dividing the recorded total catch by the estimated 

 numbers of imits of effort expended. The calculation may be made for an entire 

 season or it may be done by weekly, monthly, or other convenient units. Various 

 sources of error, such as differences among the boats and nets, changes in eflB- 

 ciency, and lack of comparable units from season to season, must be measured 

 and adjusted. Assuming that the calculation has been accurately carried out, using 

 data from a representative sample, it still does not always represent the estimate 

 of the abundance of the fish, but only the availability of the fish to the fishermen. 

 Obviously, if fish are not available to fishermen, they cannot enter into the 

 calculated estimate of catch per unit. 



Availability is affected by many things, including the size of mesh in the 

 gear used, type of gear, places fished, and the natural variables which affect the 

 migration or local concentration of the fish. Calculations of catch per unit of 

 effort can often lead one far astray in attempting to estimate abundance if the 

 factor of availability is present and has been overlooked. 



Finally, even when the catch per unit calculation has been accurately made 

 and the availabihty effect is lacking, or has been adjusted, the resulting index 

 does not furnish a proper measure of abundance. This is due to the operation of 

 certain rather complicated mathematical relationships which will not be described 

 here. Suffice it to say, however, that in a good many instances catch per unit data 

 require further treatment by mathematical methods if they are to represent the 

 abundance of the fish. 



Age Analysis. It is possible to determine the age of individual specimens of 

 marine fish by counting certain rings on their scales, otoliths (small bones in the 

 ear), fin rays, or other hard parts, techniques employed by fishery biologists for 

 several decades. If these determinations are made on representative samples of 

 stock, a distribution may be made of the fish of each age in the stock known as 



