Sockeye salmon were widely distributed and relatively 

 abundant throughout the system. The distribution of the 

 two species ot sticklebacks approximated that ot juve- 

 nile sockeye salmon, and in some areas sticklebacks were 

 more abundant than juvenile salmon. Pygmy whitefish 

 and pond smelt were abundant in some basins but never 

 abundant together — pond smelt were abundant in Co- 

 ville Lake and pygmy whitefish in Brooks Lake and 

 South Bay. 



Our best measure of the actual abundance of juvenile 

 sockeye salmon came from sampling with tow nets in the 

 pelagic areas of each basin (Ellis 1974). The use of this 

 gear to estimate abundance of juvenile salmon is docu- 

 mented by Johnson (1956), Ruggles (1966), and Pella 

 (1968). Ellis found that the abundance of age sockeye 

 salmon decreased after midsummer in Coville and 

 Grosvenor Lakes, increased in Iliuk Arm and South Bay, 

 and was so low or erratic in Brooks Lake, Northwest 

 Basin, West End, and North Arm that changes appear to 

 have no meaning (Fig. 2). A decrease in abundance of age 

 sockeye salmon after midsummer would be expected 

 because normally mortality would exceed recruitment 

 from spawning grounds and littoral areas. Some basins, 

 Coville and Grosvenor especially, had very few age I fish, 

 although they were common to abundant in others (Fig. 

 3). 



Water flows from one main-stem basin to the next 

 (Coville Lake, Grosvenor Lake, Iliuk Arm, South Bay, 

 and West End) and a summer migration of age sockeye 

 salmon results in migration of large numbers of fish from 

 Coville Lake through the sequence of basins (Ellis 1974). 

 By late August the abundance of age sockeye salmon in 

 Coville Lake had markedly declined and had increased 

 in Iliuk Arm and South Bay (Fig. 2). The three external 

 basins. North Arm, Northwest Basin, and Brooks Lake, 

 are not included in this sequence of events but the role of 

 West End in the summer migration is uncertain. Rela- 

 tively few age sockeye salmon migrate from Brooks 

 Lake to South Bay in the fall." The apparent abundance 

 of age fish in the external basins is relatively constant 

 and so low during the summer that no great number 

 could be migrating out of the external basins or in from 

 South Bay or West End. 



To determine relative numbers of the major species in 

 the pelagic areas, we summarized the tow-net data for 

 August 1961-63 (Table 3). The catches of species other 

 than sockeye salmon were much more variable than the 

 catches of salmon. (Pygmy whitefish were rarely taken in 

 tow nets. Indication of their considerable abundance in 

 Brooks Lake and South Bay comes from otter trawls and 

 gill nets.) 



A reduction in growth of juvenile sockeye salmon in 

 some systems has been associated with large numbers of 

 sockeye salmon, alone or with other species (Foerster 

 1944; Johnson 1958; Krogius 1961; Burgner 1964; Narver 

 and Dahlberg 1964), but this association does not appear 



-Hartman. W. L., W. R. Heard. C. L. Strickland. 1962. Red salmon 

 studies at Brooks Lake Biological Field Station. 1961. Unpubl. manuscr.. 

 53 p. Northwest & Alaska Fisheries Center Auke Bay Fisheries Labora- 

 tory. Natl. Mar. Fish. Serv.. NOAA, P.O. Box 15.5. Auke Bay. AK 99821. 



COVILLE LAKE 



NORTHWEST BASIN 



BROOKS LAKE 



Figure 2.— Weighted mean number of age (I sockeye salmon per stan- 

 dard tow by early (before 2(1 July), middle (2(> July-ID AuRust). and 

 late (after HI August) time periods in each lake of the Naknek Ri\er 

 system. 19fil-(;:i. From Kllis (1974). 



COVILLE LAKE 



NORTH ARM 



NORTHWEST BASIN 



BROOKS LAKE 



Figure .3. — Weighted mean number of age I socke,\c salmon per stan- 

 dard tow by early (before 2*i •Jul.\ i. middle (2(i July-H) .August), and 

 late (after 10 .August) time periods in each lake of the Naknek Kiver 

 system. I%l-(i:!. From Kllis (IIITl). 



in the Naknek system (Ellis 1974). The three basins with 

 the greatest numbers of all species of fish in the tow-net 

 catches were those with age sockeye salmon of average 

 or greater than average length — Northwest Basin, West 

 End, and Coville Lake (Table 3). It appears that compe- 

 tition for food or space did not result in a reduction in 

 growth of juvenile sockeye salmon in the Naknek system. 

 The mean fork lengths ot age sockeye salmon from the 

 Naknek system on 1 September were generally equal to 

 or greater than those of two other major river systems of 

 Bristol Bay (Wood and Kvichak Riversi tor which we 

 have comparable data (Hurgner et al. 19691. 



The rankings of abundance of age sockeye salmon in 

 the various basins in early summer relative to the rank- 

 ings of area of potential spawning grounds (Table 4) 



