LONG-TERM OLFACTORY "MEMORY" IN 



COHO SALMON, ONCORHYNCHUS 



KISUTCW 



Many experiments have correlated the impor- 

 tance of olfaction and the precise homing of sex- 

 ually mature salmon. As juveniles, the fish are 

 presumably imprinted on the natural odors of 

 their natal-area water (Hasler, 1966). The 

 odors apparently serve as cues to guide the 

 adult's return. Thus, some type of "odor mem- 

 ory" must persist from the time of the down- 

 stream journey of the smolt to the return of the 

 sexually mature adult. For introduced Lake 

 Michigan coho salmon, Oncorhynchns kisiitch, 

 this is either i/^ year (precocious males) or II/2 

 years. 



The existence of long-term olfactory memory 

 persisting over this time period has only been 

 inferred. Idler et al. (1961) and Fagerlund 

 et al. (1963) found that already homed salmon 

 made unconditioned behavioral responses to 

 home water. Hara, Ueda, and Gorbman (1965) , 

 Ueda, Hara, and Gorbman (1967), and Oshima, 

 Hahn, and Gorbman (1969a) found specific EEG 

 (electroencephalographic) responses to home- 

 stream water. Hara ( 1970) in his review of this 

 EEG technique states: "This electric response 

 [from the olfactory bulbs] is specific in the sense 

 that it cannot be evoked by water from spawning 

 sites of other groups of breeding salmon." The 

 EEG and behavioral studies strongly suggest 

 long-term memory of the juvenile's stream ex- 

 perience. However, since these workers used 

 homed adults that had recently experienced home 

 water, the data are evidence only of an odor 

 memory lasting from the time of removal from 

 the home stream to the time of testing. 



We tested coho salmon that were exposed to 

 a synthetic odoriferous substance for 1 month 

 during smoltification and then removed from any 

 conceivable influence of this substance for 10 



f \]^ Yt°^^ '^^^ completed under financial support 

 -Tc^^v. National Science Foundation (Grant No. GB 

 7bl6X) to Prof. Hasler and the University of Wisconsin 

 bea Grant Program which is a part of the National Sea 

 Grant Program maintained by the National Oceanic and 

 Atmospheric Administration of the U.S. Department of 

 Commerce. 



months. Ten months later these fish and con- 

 trols were examined for olfactory bulbar EEG 

 responses (after Hara et al., 1965) to the im- 

 printing substance. 



Materials and Methods 



On 7 April 1970, approximately 2,500 hatch- 

 ery-raised coho salmon smolts (II/2 years old) 

 were put into each of two contiguous 25-m sec- 

 tions of a raceway at a Wisconsin State fish 

 hatchery at Crystal Springs. We marked the 

 fish to eventually distinguish the upper section 

 control subjects from the lower section exper- 

 imentals. A small drop (1/3 m) prevented water 

 in the lower section from reentering the upper 

 section. Immediately below the drop a dilute 

 concentration of morpholine was introduced by 

 infusion pump at a rate to maintain a steady- 

 rate concentration of 10-^ ppm. This value is 

 one order of magnitude above an avoidance 

 threshold of unconditioned coho salmon finger- 

 lings (Wisby, 1952). On 5 May 1970, 1 month 

 after initiation of the morpholine treatment, all 

 but 50 fish from each raceway section were 

 trucked to Lake Michigan and released as part 

 of another experiment. The 100 remaining fish 

 were moved to a hatchery near Madison, Wis., 

 and held together in a single outside raceway for 

 10 months prior to EEG tests. 



Our testing procedure was generally similar 

 to that used by Hara et al. (1965) to examine 

 olfactory bulb responses to home-stream water. 

 The subject was paralyzed with gallamine trie- 

 thiodide (2 mg/kg), restrained, and the gills 

 perfused with tap water. One of the olfactory 

 bulbs was exposed, and an electrode (Transidyne 

 General, model 415') was placed on the surface 

 near the rear margin. The responses evoked by 

 perfusion of the ipsilateral naris were amplified 

 (Bioelectric Instruments, model DS2c) and re- 

 corded on a two-channel oscillograph (Hewlett 

 Packard, model 7712B) for later analysis. This 

 oscillograph was equipped with an integrating 

 preamplifier for efficient quantification of bulbar 

 activity. Therefore, all responses reported later 



Reference to trade names does not imply endorse- 

 ment by the National Marine Fisheries Service, NOAA. 



315 



