Kvichak system. — Our most complete measure 

 of the distribution and abundance of juvenile 

 sockeye salmon in the Kvichak system is from 

 tow net samples from Iliamna Lake in 1962 (fig. 

 6). The contour lines of density in the figure 

 were determined by interpolation of the data on 

 rate of catch between stations sampled and are 

 not precise, but they illustrate the general distri- 

 bution of the two age classes. The distribution 

 and abundance changed during the summer and 

 was different for age and age I sockeye (fig. 6) . 



Most of the age fish were in the eastern two- 

 thirds of the lake from August 1 to September 

 15 — fig. 6 A (the major spawning areas are in 

 the eastern half of the lake) . 



Comparison of the distribution and abundance 

 of age I sockeye salmon in early and late summer 

 indicates a westerly movement, away from the 

 principal spawning areas (fig. 6B and 6C) . Age I 

 juveniles had two centers of maximum concen- 

 tration near the end of August — one in the west 

 third and one in the middle third of the lake. 

 These progeny of the 1960 escapement, there- 

 fore, occupied an area of the lake during their 

 second summer (1962) that is probably not 

 heavily used when most of the juveniles migrate 

 to sea as age I smolts in early summer. 



We found gradients in proportion of young-of- 

 the-year and yearling sockeye salmon in the tow 

 net catches in 1962 and in the average length of 

 each age group from the outlet of the system to 

 distant points in the system in catches made with 

 an Isaacs-Kidd midwater trawl in 1961 and tow 

 nets in 1962 (table 20) . Yearling sockeye salmon 

 were essentially absent from catches in 1961, but 

 in 1962 they made up about 90 percent of the 

 catches in the third of the lake closest to the out- 

 let, about 85 percent in the middle third, and 45 

 percent in the most distant third. None were 

 taken in Lake Clark. The average lengths of both 

 age classes were greater nearer the outlet and 

 less at the more distant points both years. 



A strong gradient was found at the end of 

 August 1962 in the mean length of age I sockeye 

 salmon from three areas of Iliamna Lake. The 

 average lengths in the east, middle, and west 

 thirds of the lake were 81.5, 88.1, and 99.6 mm., 

 respectively. Since spawning is most limited in 

 the west end of the lake, the gradient is thought 

 to be due to the more rapid or earlier migration 

 of the larger and possibly older fish. By the fol- 

 lowing spring (1963) the westward movement 



Table 20. — Age class composition and mean fork length of 

 juvenile sockeye salmon on September 1 in various areas of 

 the Kvichak system, 1961-62. The fish were captured with an 

 Isaacs-Kidd midirater trawl in 1961 and in a 6-foot-square 

 opening tow net in 1962. Original measurements were made 

 of live fish and these data converted to preserved fish equivalents 



1961 1962 



Young-of- Young-of- 



the-year Yearlings the-year Yearlings 



Aver- Aver- Aver- Aver- 



Fre- age Fre- age Fre- age Fre- age 

 Area quency length quency length quency length quency length 



Per- Per- Per- Per- 

 cent Mm. cent Mm. cent Mm. cent Mm. 



Iliamna Lake 



Igiugig 100 98.6 



Shoulderblade 100 96 3 



Grassy 81 56.6 19 90.5 



Middle Talarik.__ 24 56.1 76 91 5 



Chekok 100 93.1 



Dennis 20 57.8 80 80.2 



Upper Talarik_._ 100 59 7 18 54.7 82 85 3 



Ten-Mile Island 15 50.6 85 77 9 



Middle Island 8 .53.6 92 81.9 



Gibraltar 26 52 7 74 88.4 



Kakhonak Bay 3 97 89.1 



Intricate Bay 22 53 6 78 83.8 



Tommy Points. 100 64 3 25 50.7 75 79.2 



Iliamna 100 53.9 56 50.2 44 79.5 



Triangle Island.. 100 52.4 30 52 9 70 77 



Knutson Bay___ 100 46 2 66 52 5 34 76.7 



Pedro Bay 100 51 7 78 48.8 22 76 7 



Pile Bay. _ 100 50 2 75 48 4 26 74 6 



Lake Clark 



Lower lake 100 51.3 



Middle lake 100 46.3 



of these fish (now age II smolts) must have been 

 completed because most of them migrated in the 

 first few days after the ice in the lake broke up. 

 Naknek system. — The interlake migration of 

 sockeye salmon in the Naknek system during 

 their first summer influences the relation be- 

 tween the abundance of young salmon in each 

 lake at the end of the summer and the number of 

 adults in the parent escapement to each lake. 

 Data collected with tow nets in 1962 have been 

 used to estimate a systemwide average catch 

 per tow by day, weighted by area of each lake, 

 for July 11 to August 29. Sampling was not daily. 

 Missing data for an area were derived by averag- 

 ing the figures of the most recent preceding and 

 the first following sampling in that area. These 

 weighted figures were combined for all areas 

 each date to find the hypothetical catch per tow 

 for the entire system and were smoothed by a 

 moving average of three. The data (fig. 7) show 

 a gradual increase in abundance of age fish 

 while recruitment from the spawning grounds 

 exceeds mortality, a general leveling of the curve 

 from July 20 to August 10, and a decline in rate 

 of catch per tow after August 10. The curve 



SOCKEYE SALMON IN MAJOR RIVER SYSTEMS IN SOUTHWESTERN ALASKA 



433 



