Lake). The interlake and interbasin migrations 

 of presmolt fish in the Naknek system caused 

 large numbers of fish to accumulate in downsys- 

 tem lake areas, which have comparatively lim- 

 ited spawning grounds. Fyke net catches from 

 Coville, Grosvenor, and Brooks Rivers and tow 

 net catches from the lakes indicate that Naknek 

 Lake, except for the relatively isolated North 

 Arm basin, serves as a rearing area for fish 

 originating in other lakes. 



Chignik S\ stem 



Interlake migration of age sockeye salmon 

 from Black LaketoChignik Lake was detected by 

 fishing fyke nets in Black River in 1961 and 1962. 

 Scales of age and age I salmon from Chignik 

 Lake indicate that some Black Lake progeny use 

 Chignik Lake as a rearing area, but scales from 

 spawning populations in Black Lake indicate 

 that most fry remain in Black Lake to age I and 

 then migrate directly to the ocean (Narver, 

 1963). Extensive sampling of Chignik Lagoon 

 with tow nets and seines in 1962 and 1963 re- 

 vealed that some of the progeny of spawners of 

 Chignik Lake migrate into Chignik Lagoon soon 

 after they emerge from the gravel and use the 

 lagoon as a rearing area. 



Karluk System 



Interlake movements of presmolt sockeye 

 salmon have not been investigated in the Karluk 

 system. Because Karluk Lake is the only rearing 

 area of significant size, these movements could 

 not be important. 



ABUNDANCE AND GROWTH OF FISH IN 

 PELAGIC AREAS 



The capacity of a system to produce juvenile 

 sockeye salmon is measured ultimately in the 

 number of smolts produced. We have discussed 

 our studies of the capacities of the spawning 

 grounds. To locate and define factors that affect 

 production of smolts in the rearing areas, we 

 investigated the relation of the abundance of 

 adult sockeye salmon to the abundance and 

 growth of juveniles and associated species in 

 the lakes. 



Sockeye Salmon 



The abundance of juvenile sockeye salmon in 

 the lakes was measured with the pelagic tow net 

 developed by Johnson (1956) and improved by 



Burgner (1958; footnote 9). In this study the 

 standard unit of gear was a conical tow net 10 

 feet (3.1 m.) in diameter and 25 feet (8.3 m.) 

 long. This net was pulled behind a pair of boats 

 at night and fished with the center of the opening 

 at two depths — 1.6 and 5 m. Modifications of 

 the standard tow net were ( 1 ) a square opening 

 of 9 feet (2.7 m.) and a length of about 25 feet 

 (8.3 m.), used in the Wood and Kvichak systems; 

 and (2) a square opening of 6 feet (1.8 m.) and 

 a length of about 19 feet (5.8 m.), used in the 

 Chignik system. The rate of catch is expressed 

 in terms of the standard round opening 3.1-m. 

 net pulled 1,500 feet (457 m.) in about 6 

 minutes. 



Because we expected the abundance and 

 growth of juvenile sockeye salmon and associ- 

 ated fishes would vary within lakes, we divided 

 each lake into several sampling areas. The areas 

 were not generally of uniform size, even within 

 a lake. When we wish to discuss the average 

 abundance or size of fish in a lake we must con- 

 sider the relative contribution of each sampling 

 area to the total population. We do this by a 

 process of weighting in which we consider the 

 portion of the total lake area in each sampling 

 area and the relative abundance and size of the 

 fish in the sampling area.'* This process is re- 

 peated to compile an average for a system of 

 more than one lake. 



The mean catches of age and age I and older 

 sockeye salmon in tow nets (surface and deep 

 tows combined) in the Wood, Kvichak, Naknek, 

 Chignik, and Karluk systems, August 16 to Sep- 

 tember 15, are given in tables 17 and 18. Dif- 

 ferences in abundance between systems within 

 years were large (for example, 1961) as were 



'* The following: formulas yield the weighted average catch per tow 

 and the weighted average length of fish. 



( Ic, 



i 1 A 



T _ . - , (c,a,) L 



i=l 



(c. 



where C„ ^ weighted catch per tow for entire lake 



c, ^= average catch per tow for sampling area i 



A = surface area of entire lake 



a^ = surface area of sampling area i 



L^^ = weighted average length of fish in the lake 



I = average length of fish in sampling area i 



SOCKEYE SALMON IN MAJOR RIVER SYSTEMS IN SOUTHWESTERN ALASKA 



429 



