spawn also in the tributaries to, and on beaches 

 of, small lakes. Some of the fry from these small 

 lakes migrate to nursery areas in Iliamna Lake 

 or Lake Clark. 



The distribution of the escapements to large 

 streams, lake beaches, and all other spawning 

 grounds of the Kvichak system in 1959-62 is 

 shown in table 11. When the escapement was ex- 

 tremely large in 1960, the beaches of Iliamna 

 Lake received a larger proportion of the total 

 escapement and the large streams received a 

 smaller pi'oportion than in years when the es- 

 capements were smaller. The importance of the 

 other spawning grounds remained nearly con- 

 stant from 1959 to 1962. 



Table 11. — Dislribulion of sockeye salmon escapements in 

 spawning grounds of the Kvichak systevi, 1959-62 



Escapements to large streams have varied by 

 a factor of 10, whereas those to the system 

 varied by 20 and those to the lake beaches by 

 about 50. Thus, the cyclic nature of the produc- 

 tion of sockeye salmon in the Kvichak system 

 could be reinforced or pei'haps caused by the 

 cyclic nature of the production from the beaches 

 alone. 



We found marked contrasts in age character- 

 istics — both ocean and fresh-water — among 

 spawning groups in the Kvichak system in 1962. 

 Stream spawning groups (except those of the 

 Copper River) had a significantly lower pro- 

 portion of 3-ocean fish than beach spawning 

 groups. ' Differences also existed in the average 

 fresh-water age of various spawning groups in 

 the Kvichak system in 1956-58 (Koo and Smith, 

 1960). 



Smith (1964) found little or no difference in 

 the time the various spawning groups entered 

 the Kvichak system in 1957, 1958, and 1959. 

 Adult sockeye salmon were tagged as they en- 

 tered Iliamna Lake, and the tagged fi.sh were 



• Kim. Wan Sou, and Robert L. Demory. 1963. Kvichak salmon 

 studies. 1962 spawning ground studies. Univ. Wash., Fish. Res. Inst.. 

 Circ. 195. 26 pp. 



recovered on the spawning grounds. The migra- 

 tion into the lake lasted about 4 weeks. Fish 

 tagged in any one of these weeks were recovered 

 in about the .same proportion in large and small 

 streams, on lake beaches, and in spring-fed 

 ponds, regardless of whether spawning in these 

 places occurred early or late in the season. 



Naknek system. — The principal spawning 

 grounds of the Naknek system are : (1) Brooks 

 and Grosvenor Rivers (rivers between lakes) ; 

 (2) American and Hardscrabble Creeks (tribu- 

 tary streams) ; (3) Naknek River (outlet or 

 trunk river) ; and (4) beaches of Grosvenor 

 Lake. Some sockeye salmon also spawn on 

 beaches in Iliuk Arm and South Bay of Naknek 

 Lake. The proportions of sockeye salmon that 

 spawn on the several major spawning grounds 

 change markedly from year to year. In some 

 years a single major ground may receive more 

 than 60 percent of the escapement to the system 

 and in other years the same ground may receive 

 less than 10 percent. 



Each of the major groups of spawning sock- 

 eye salmon in the Naknek system in 1962 was 

 characterized by diflferent proportions of com- 

 binations of fresh-water and ocean ages ( Straty, 

 1966) . Most major spawning grounds contained 

 a smaller proportion of adults that migrated 

 seaward as age II smolts than did the total es- 

 capement to the Naknek River. This anomaly 

 suggests the existence of spawning groups as 

 yet undetected. 



Tagging in the Naknek River in 1961 and 

 1962 demonstrated that all major spawning 

 areas within the system received fish from all 

 parts of the run about in proportion to the 

 sizes of the spawning groups (Straty, 1966). 

 Some slight indication appeared, however, that 

 Brooks Lake and Brooks River received more 

 of the earlier fish than the other areas. 



Chignik si/stem. — Sockeye salmon in runs to 

 the Chignik system are predominantly 3-ocean 

 fish. The greatest contrast in age composition 

 in this system is in the time the young live in 

 fresh water : most of the smolts migrate at age I 

 from Black Lake and at age II from Chignik 

 Lake. 



Tagging and studies of scales have established 



that adults returning to Chignik and Black 



Lakes pass through the fishery at different times 



(Narver, 1963). The early run (in June) has 



•primarily sockeye salmon bound for Black Lake, 



SOCKEYE SALMON IN MAJOR RIVER SYSTEMS IN SOUTHWESTERN ALASKA 



421 



