320 



Fishery Bulletin 98(2) 



intuitive and there is little direct sup- 

 portive evidence. To test this idea, we 

 designed our study with the following 

 objectives: 1) to examine emigration 

 timing and growth of tagged individ- 

 ual juvenile pink salmon in relation 

 to secondary production and water 

 temperature and 2) to investigate 

 the relation between timing of emigra- 

 tion, early marine growth, and intra- 

 and interannual variability in marine 

 survival. 



Methods 



This research focused on the Auke 

 Creek pink salmon population. Auke 

 Creek is a small, lake-fed stream 

 that empties into Auke Bay in south- 

 east Alaska (Fig. 1). The National 

 Marine Fisheries Service maintains 

 a permanent two-way counting weir 

 (where both seaward-migi'ating and 

 returning salmon are counted) at the 

 confluence of Auke Creek with Auke 

 Bay. The pink salmon run at Auke 

 Creek is bimodal, that is it has a dis- 

 tinct early and late spawning compo- 

 nent. Salmon in the early run begin 

 spawning in late July through August 

 and salmon in late run spawn from 

 September through mid-October. The 

 emigration of fry in the spring lasts 

 from late March to mid-May and is 

 relatively unimodal. The peak emigration usually 

 occurs in late April (Taylor, 1980) 



The downstream emigration of wild pink salmon 

 juveniles from Auke Creek into Auke Bay generally 

 begins in early March, peaks between mid-April and 

 early May, and ends by mid-May. From 1986 to 1989, 

 53,526, 17,249, 38,149, and 42,599 juvenile pink 

 salmon emigrated from Auke Creek respectively. 

 Each day all captured emigrants were counted and 

 samples were measured to the nearest millimeter 

 fork length. 



Each year a portion of the emigrants was marked. 

 Marking involved the excision of the entire adipose 

 fin after a 0.5-mm binary coded wire was injected 

 into the snout. Tag codes were assigned in lots of 

 10,000; therefore to use all the tags in a particular 

 code-lot, tagging was conducted over several days 

 each week. The 1985 brood was tagged in five-day 

 emigration periods when sufficient numbers of fish 

 were captured. The 1986 through 1988 broods were 



Figure 1 



Map of Auke Bay, Alaska, showing beach seine sites (circles) and surface trawl tran- 

 sects (dashed lines). 



tagged in two to three-day emigration periods. After 

 having been tagged, the fish were held for one day 

 ( 1985 and 1986 broods) or three days ( 1987 and 1988 

 broods) to assess mortality and tag loss. Dead fish 

 and those missing tags were deducted from the total 

 tagged in each group. Release dates were similar 

 each year. 



Nearshore areas of Auke Bay were sampled for 

 juvenile pink salmon from late March to early July 

 with a 37-m long x 3-m deep beach seine. All captured 

 juvenile salmon were identified to species, counted, 

 and checked for evidence of tagging at the capture 

 site; recaptured tagged fish were retained for tag 

 recovery and size measurements. Random samples of 

 up to 100 unmarked fish from each beach seine from 

 sites on Spuhn and Indian Islands in Auke Bay were 

 retained for length and weight measurements and 

 stomach analysis. Other species (nonsalmonid) that 

 were captured in the beach seines were enumerated 

 and measured to the nearest millimeter (fork length 



