Table 1. — Median EEG response ratios of coho salmon' 

 from two different spawning sites to waters collected both 

 on and off their migration route. 



|_response to Ahnapee tributary below (Ab) J 



strongly if the water was taken from below the 

 trai)s (Kb) than from above the traps (Ka). 

 Furthermore, Kewaunee fish responded to Ah- 

 napee home-stream water more than to their 

 own (Table 1). Yet the Kewaunee fish did home 

 correctly — three of the six Kewaunee fish 

 carried fin clips unique to that site (the other 

 three were unmarked). 



Discussion 



Hara (1970) in his recent review of electro- 

 ])hysiological studies of olfactory discrimination 

 by homing salmon, concluded that "each spawn- 

 ing area has its own specific stimulant, or 

 special combination of stimulants, recognized 

 and responded to by the anadromous salmon." 

 He acknowledged both that EEG responses of 

 returning salmon adults can indicate a recog- 

 nition of a chemical factor from conspecifics 

 and that recent exposure of mature salmon to 

 nonhome-stream water can result in strong 

 EEG responses to that water (Oshima, Hahn, 

 and Gorbman, 1969). Therefore olfactory re- 

 sponses to the home stream could result from 

 (1) natal imprinting (a long-term memory), (2) 

 recent exposure to the water (a short-term 

 memory), (3) recognition of chemical factors 

 from consi)ecifics, (a genetic (?) memory), and 

 (4) the presence of stimulatory products not in 

 any way specific to the home stream, transient, 

 and incidental to the salmon's appearance, past 

 or present, in the stream (responses independent 

 of memory). 



Our work ajijiears to be the first to separate 

 responses to conditioned home-stream water 

 (other conspecifics were present) from responses 

 to "pure" home-stream water. Conditioning of 

 water by conspecifics appeared to double the 

 EEG response to the home stream. 



Perception of factors in the home stream 

 other than those contributed by conspecifics is 

 indicated in the greater responses of the Ahnapee 

 coho salmon to "pure" home-stream water (Aa) 

 than to other waters (R, M) not strongly condi- 

 tioned by conspecifics. But this result does not 

 necessarily imply natal imprinting (long-term 

 memory) or recent exposure (short-term mem- 

 ory). Perception of nonspecific odoriferous 

 products is indicated since (1) responses to 

 water from the home stream (Aa) and nonhome 

 river (R) by Ahnapee fish differ little, and (2) 

 Kewaunee fish responded more to "pure" water 

 from the Ahnapee tributary than to their own 

 home-stream water. 



The consistence of ranking of the four test 

 waters (Ab>Aa>R>L) by 11 fish indicates 

 some degree of reliability of the electrophysiolo- 

 gical technique. Nevertheless, there seems to be 

 an inherent lack of precision associated with 

 evoked bulbar recordings. Sutterlin and Sutter- 

 lin (1971) found that EEG responses varied 

 greatly among and within fish when compared 

 to simultaneous olfactory epithelium recordings. 

 Furthermore, our collection stations for water 

 samples were subject to seiches (reversed cur- 

 rent flow in the river and stream) and variable 

 rates of river and stream discharge that made 

 the water inconstant. 



In conclusion, our experiments with lake-run 

 coho salmon indicate the need to clearly dif- 

 ferentiate between responses to "pure" home- 

 stream water and water conditioned by pre- 

 sence of conspecifics. An intense and apparently 

 specific response to home-stream water is not 

 necessarily indicative of juvenile olfactory im- 

 printing. 



Acknowledgments 



Special thanks go to Ralph Schmiling, 

 Francis "Toots" Wenniger, and Jack Haegele 

 of Algoma, Wis., for help with the specimens, 

 boats, tools, housing, and storage. 



895 



