lakes as well as in the terrestrial environ- 

 ment, and bottom sannpling has shown that the 

 punnice is also evident at depths of 26,8 m. 



The climate of the area is moderated by 

 the Japanese Current, and the average annual 

 precipitation of about 152 cm. is fairly evenly 

 distributed throughout the year (Capps, 1937). 



Dissolved oxygen concentrations are suffi- 

 cient to sustain fish life at all times. For a 

 few days in the early spring, concentrations 

 at the lake bottoms may be slightly less than 

 3.0 p. p.m. (parts per nnillion), the figure 

 generally accepted as necessary for fish. 

 Values normally range from about 14.0 p. p.m. 

 at the surface in June to about 11.0 p. p.m. at 

 the surface in January (Parker and Vincent, 

 1956). During the spring and fall overturns 

 (generally in May and October), sufficient 

 oxygen is circulated to the lower depths of the 

 lakes to sustain fish through the summer and 

 winter stagnation, when the fornnation of a 

 thermocline or an ice cover prevents circula- 

 tion of oxygen- rich surface waters to the depths. 



Three lakes were selected for study. Little 

 Kitoi Lake furnished the brood stock for 

 planting the other lakes and also served as 

 one of the control lakes. Besides runs of 

 sockeye, coho, and pink salnnon ( O . gorbuscha) , 

 this lake maintains populations of threespine 

 sticklebacks ( Gasterosteus aculeatus), Dolly 

 Varden, sculpins, and a small number of rain- 

 bow trout (Salmo gairdneri) . Midarnn Lake, 

 the second control lake, maintains populations 

 of sticklebacks and Dolly Varden. Ruth Lake 

 was the experimental lake. It also contains 

 sticklebacks and Dolly Varden. 



METHODS 



Weirs with traps and holding boxes were 

 built at the outlets of all three lakes to facili- 

 tate accurate counting and sampling of down- 

 stream nnigrants. 



During the sunnnner of 1955 an application 

 of 5 percent rotenone killed resident fish pop- 

 ulations in Ruth Lake. A complete kill appar- 

 ently resulted because no fish other than the 

 introduced sockeye have been observed since 

 the treatment. 



Beginning in the fall of 1955, sockeye eggs 

 were taken from beach-spawning sockeye in 

 Little Kitoi Lake, fertilized, and placed in a 

 hatchery for incubation and rearing. Early in 

 July 1956, the first resulting fry were re- 

 leased in Midarm and Ruth Lakes. Fry were 

 reared in the hatchery and released yearly in 

 Ruth Lake until 1960; however, because of a 

 lack of sufficient brood stock to incubate 

 experimental eggs, no fry were released in 

 Midarm Lake after 1956. The number of fry 

 released each year and the density per sur- 

 face acre are shown in table 2, 



All fry releases from the Kitoi hatchery 

 were made as soon as the yolk sacs were 



Table 2. — Number and density per surface 

 hectare of hatchery-reared sockeye fry from 

 Little Kitoi Lake stock released in Midarm 

 and Rath Lakes, 1956-60 



absorbed and the fish were ready to feed 

 naturally. Each year some hatchery feeding 

 was required before all of the fry were ready 

 for sinnultaneous release. Foerster (1925) re- 

 ported that other investigations have shown 

 that only a few sockeye migrate to sea as fry; 

 he considered this behavior when he proposed 

 that fry be released from hatcheries into fresh 

 water as soon as possible. 



The 1958 plant offered an exception to the 

 release of fry from Little Kitoi stock into 

 Ruth Lake. In that year, to test the effect of 

 increased fry planting on smolt production, 

 and because only 55,000 fry were available 

 from Little Kitoi Lake, 55,000 fry from Hugh 

 Smith Lake near Ketchikan were added to 

 provide a total release of 110,000 into Ruth 

 Lake. Sockeye from Hugh Smith Lake were 

 similar to those from Little Kitoi in that both 

 races spawned in early October and both spawn- 

 ing lakes are adjacent to salt water. Hugh 

 Smith Lake lies at about lat. 55 N. in South- 

 eastern Alaska, and Little Kitoi Lake lies at 

 lat. 58° N. 



Salmon Smolts 



During their seaward migration from the 

 lakes, all sockeye smolts were counted, and 

 at least 100 were measured (fork length) each 

 day. They were anesthetized (during the early 

 years of the study with urethane and later 

 with tricaine methanesulfonate, MS 222) and 

 placed on millimeter graph paper with the tip 

 of the nose set on the zero line. A pin prick at 

 the end of the nnidline of the caudal fin re- 

 corded the length (fig, 2). Anesthetized fish 

 recovered in a holding pen and were then re- 

 leased into the outlet below the weir. 



Smears of smolt scales were placed on 

 glass microscope slides. During each week of 

 the downstreann migration, scales were 

 sampled from 10 smolts in each 5-mm. (milli- 

 nneter) size class, or from all smolts in size 

 groups that did not contain more than 10 fish. 



