Orsi and Jaemcke Marine distribution and origin of prerecruit Oncorhynchus tshawytscha 



495 



tudes owing to temperature and photoperiod regimes 

 (Taylor, 1990) but retained the early ocean-distribu- 

 tion pattern characteristic of their ocean-type lineage. 



Differences in length and spatial occurrence of 

 prerecruit ocean- and stream-type chinook salmon 

 in our study were similar to those found later in the 

 commercial troll fishery of southeastern Alaska. Our 

 stream-type chinook salmon were larger and more 

 prevalent in inside waters than were ocean-type 

 chinook salmon of the same ocean-age group. In the 

 southeastern Alaska troll fishery, stream-type 

 chinook salmon are also larger than ocean-type 

 chinook salmon of the same ocean-age group (Van 

 Alen and Wood, 1983; Van Alen et al., 1987), and 

 ocean-type chinook salmon are generally more preva- 

 lent than stream-type chinook salmon in outside wa- 

 ters (Van Alen and Olsen, 1986; Van Alen et al., 1987). 



Prerecruit chinook salmon were older and larger 

 in the northern regions. Age data from the commer- 

 cial troll fishery of southeastern Alaska similarly 

 reveal that chinook salmon are older in northern re- 

 gions (Van Alen and Olsen, 1986; Van Alen et al., 

 1987). In British Columbia, Healey ( 1986) found that 

 age 0.2 and age 0.3 troll-caught chinook salmon domi- 

 nated the catch, that larger age 0.3 fish were more 

 common in northern and central regions, and that 

 smaller age 0.2 fish were more common farther south 

 off Vancouver Island. Age 0.2 chinook salmon also 

 dominate the catch in the Strait of Georgia (Argue 

 et al., 1977; Carter et al., 1986) and off the Washing- 

 ton coast (Wright et al., 1972). However, increased 

 harvest of age 0.2 fish southward is also a reflection 

 of lower minimum-size limits in these fisheries. 



The 15-43 cm FL range of our troll-caught age -.0 

 chinook salmon during September in Alaska is con- 

 sistent with other studies. Hartt and Dell ( 1986) re- 

 ported a 16-36 cm FL range for age -.0 chinook 

 salmon caught with small-mesh purse seines fished 

 in the northeastern Gulf of Alaska from July to Oc- 

 tober. In another seine study, Fisher and Pearcy 

 (1995) reported a 9-32 cm FL range for age -.0 

 chinook in shelf waters off the Oregon and Washing- 

 ton coasts from May to September. Of the smaller 

 chinook salmon captured, Fisher and Pearcy ( 1995) 

 found that chinook salmon <15 cm FL were most 

 abundant in July and August and that chinook 

 salmon <13 cm FL (primarily age 0.0 fish) were 

 closely associated with the warm, low-salinity wa- 

 ters of the Columbia River plume. Moreover, Miller 

 et al. ( 1983 ) hypothesized that offshore movement of 

 chinook salmon is size dependent because few 

 chinook salmon <13 cm FL are found in waters >30 

 m. Our gear test with smaller hooks and lures did 

 indicate a lower length range for age -.0 fish sampled, 

 but fish <15 cm FL were not encountered. Therefore, 



the absence offish <15 cm FL in our study may have 

 been a function of the deeper water (>37 m) that we 

 sampled or of our later sampling time (September) 

 when small age -.0 fish were unavailable. In contrast 

 to the seine sampling, the greater upper length range 

 for this age group caught on troll gear may be attrib- 

 uted to seine avoidance by larger chinook salmon. 

 Chinook salmon occur deeper as they increase in 

 length (Orsi and Wertheimer, 1995). Consequently, 

 troll sampling may provide a more representative 

 sample of larger age -.0 chinook salmon within this age 

 group if larger fish were deeper than the seine nets. 



Fishery size limits and gear selectivity can influ- 

 ence length distributions. The 66-cm-FL Alaska size 

 limit for chinook salmon had little effect on the length 

 distribution of the age -.0 or -.1 chinook salmon 

 sampled in our study, because the range of length 

 for these age groups seldom exceeded the size limit. 

 However, the range of length for age -.2 chinook 

 salmon usually exceeded the size limit. In the south- 

 eastern Alaska troll fishery, Van Alen and Wood 

 (1983) have shown that the percent contribution of 

 age -.2 chinook salmon increases during summer. 

 Therefore, many larger age -.2 chinook salmon may 

 have been removed by the troll fishery before we 

 sampled in September. Healey (1986) reported that 

 the average length of age 0.2 chinook salmon in the 

 British Columbia troll fishery remained relatively 

 constant over the fishing season and attributed this 

 constancy to new recruits growing to legal size. Se- 

 lectivity by trolling gear also influences length dis- 

 tribution; Orsi ( 1987) reported significantly more age- 

 .2 chinook salmon caught in September-October on 

 standard commercial gear than on small trolling gear. 

 Other studies also document that large chinook 

 salmon are generally caught with large trolling gear, 

 such as plugs (e.g. Milne, 1955; Boydstun, 1972; Orsi 

 et al., 1993). Thus, fishery size limits and removals, 

 as well as gear selectivity, affected the length struc- 

 ture of age -.2 chinook salmon in our study. 



Mean fork length of stream-type chinook salmon 

 was consistently greater than that of ocean-type fish 

 of the same ocean-age group, but this difference was 

 less pronounced as ocean age increased. The greater 

 mean fork length of age -.0 stream-type fish is a func- 

 tion of greater freshwater age and an earlier time of 

 ocean entry. However, mean fork length of age -.2 

 ocean- and stream-type fish were nearly equivalent 

 in September. Although Healey ( 1991 ) reported that 

 rates of growth were similar between ocean- and 

 stream-type fish during their ocean life, we found 

 significantly higher growth rates for ocean-type fish. 

 The higher growth rates for ocean-type fish in our 

 study may explain why mean fork length of ocean- and 

 stream-type fish converged as ocean age increased. 



