FISHERY BULLETIN: VOL. 87, NO. 4. 1989 



Table 2,— Patterns of homing displayed by adult coho salmon 

 from different experimental rearing regimes. Numbers listed 

 represent actual fish returning while numbers in parentheses 

 are the percentage of each experimental group returning to 

 that recovery hatchery. Groups 3 and 5 were reared and 

 released at Seward Park and Issaquah Creek hatcheries, 

 respectively. Groups 1, 2, and 4 were released into Lake 

 Union but had been reared at the University of Washington, 

 Seward Park, and Issaquah Creek hatcheries, respectively. 



Recovery 

 hatchery 



Experimental group 



1 



U. of Washing- 

 ton 34(100) 44(86) 15(88) 2(4) 



Seward Park 7(14)44(100) 



Issaquah Creek 2(12) 54(96) 



hatchery, generally returned to Issaquah Creek 

 (54/56 recoveries). The return of transported fish 

 (Group 4) was lower but they tended to enter the 

 UW hatchery (15/17 recoveries). Fifteen coho 

 salmon entered the NMFS facility in 1987 but 

 none were from any of the experimental treat- 

 ments. 



DISCUSSION 



The coded wire tagging data demonstrated 

 that salmon home almost without fail to the UW 

 and Iss hatcheries if they have been reared and 

 released at these sites. The return of all mem- 

 bers of Group 1 to the UW hatchery supported 

 the findings of many previous studies (reviewed 

 by Hasler and Scholz (1983)) that exposure to a 

 water source at the smolt stage or at the time of 

 release provides a sufficient basis for homing. 

 Similarly, Group 3, released from SP, returned 

 exclusively to SP. Fish from Group 2 had experi- 

 enced a gap in their migi'ation, relative to Group 

 3. They were reared at SP during the parr and 

 smolt stages but did not experience the route 

 from SP to the Lake Union release site, a dis- 

 tance of some 18 km. Most of these fish entered 

 the UW hatchery but 7 of 51 returned to SP. 



The Iss controls (Group 5) returned to that 

 hatchery and the salmon trucked to Lake Union 

 tended to enter the UW hatchery, though the 

 return rate of the experimentals was quite low. 

 The salmon held in Iss hatchery before being 

 trucked to Lake Union presumably learned the 

 characteristics of their hatchery but were unable 

 to detect them when they returned to Lake 

 Union as adults. Taken together, the results of 

 the experiments support Harden Jones' (1968) 

 hypothesis that salmon learn and subsequently 



retrace a sequence of odors. In situations where 

 the home odor travels relatively undiluted or 

 unchanged downriver, salmon artificially dis- 

 placed downriver might be able to home success- 

 fully. However, if the home water is diluted or 

 altered by passage through lakes (as may have 

 occurred in our experiments), salmon may only 

 return as far as their release site. 



It is possible that the differences between the 

 patterns of homing displayed by Groups 2 and 3 

 and Groups 4 and 5 could be due to differences in 

 the degree of smolting. For example, if there is a 

 very tight window for imprinting which is linked 

 to some subtle (or unknown) changes during 

 smoltification, then perhaps 7 of the 51 returning 

 fish from Group 2 had reached and ended the 

 imprinting phase prior to transport to Lake 

 Union. This would imply that all the returning 

 fish were able to detect SP water but that only 

 the above 7 responded to it. This explanation 

 seems unlikely, however, since these fish were 

 released during a relatively late phase of the 

 smolting process. If imprinting is linked to 

 events such as natural thyroid hormone peaks 

 and the onset of silvering and downstream mi- 

 gi'ation (Hasler and Scholz 1983), then all the fish 

 in Group 2, whether released at SP or trans- 

 ported to Lake Union, would have been ex- 

 pected to return to SP. 



The gap in experience that we provided was 

 relatively short in distance but gi'eat in effect on 

 homing, compared with experiments on the 

 Columbia River (e.g., Slatick et al. 1975) in 

 which much longer displacements did not affect 

 homing. However, extreme treatments, such as 

 displacement 574 km downriver from Dworshak 

 Hatchery to Bonneville Dam (Slatick et al. 1982), 

 did impair homing. Presumably, if salmon can 

 detect the upriver odor when they arrive at the 

 release site, little effect of displacement will be 

 recorded, regardless of the linear distance. 



The fish displaced downriver to Lake Union as 

 smolts tended to enter the UW hatchery even 

 though they had not experienced its water. We 

 hypothesize that these fish initially returned to 

 the release site in Lake Union and found it un- 

 suitable for spawning. The salmon could then 

 have been attracted to the odors of the 1,708 

 adult coho salmon which were in the UW hatch- 

 ery over the course of the season. By com- 

 parison, the equally proximate NMFS facility 

 contained only 15 adult coho. Adult coho salmon 

 can recognize waters conditioned by conspecifics 

 (Dizoii et al. 1973) and behavioral attraction to 

 such species-specific odors has been documented 



772 



