FISHERY BULLETIN: VOL. 80, NO. 1 



Table 8.— Coho salmon stocks at the 13 cluster level of the 

 divisive cluster analysis. "Wild" denotes wild stocks. 



associated with the other southern Oregon coast 

 wild stocks, and the Rogue and Umpqua hatch- 

 ery stocks were in the cluster with other hatch- 

 eries that used wild stocks as egg sources. 



The second trend observed in the dendrogram 

 was geographical clustering. Three stocks from 

 Washington and California were dissimilar to 

 the Oregon stocks, and the Oregon wild stocks 

 clustered into two groups, northern and southern 

 coastal stocks. 



The third trend in the dendrogram was for 

 hatchery and wild stocks to cluster independent- 

 ly. One of the clusters was composed entirely of 

 wild stocks from the northern Oregon coast, and 

 another included all but one of the northern Ore- 

 gon coast hatchery stocks, in addition to two wild 

 stocks from the northern Oregon coast. The 

 hatchery stock excluded from this cluster (no. 5) 

 was from the Salmon River, a stock developed 

 from eggs of wild coho salmon; both brood years 

 of this stock were in the cluster of hatcheries that 

 used wild stocks as an egg source. The rest of the 

 northern Oregon coast hatcheries used return- 

 ing hatchery-reared adults for egg sources. The 

 two wild stocks in this cluster were from the 

 Trask and Nehalem Rivers. They are also simi- 



lar to the other wild stocks; however, because of 

 the mechanics of the group-average clustering 

 strategy, they both clustered first with the hatch- 

 ery stocks. The average Euclidean distance be- 

 tween the Nehalem wild stock and the other wild 

 stocks was less than that between the Nehalem 

 wild stock and the hatchery stocks of cluster 5. 

 The close relationships of the stocks in clusters 1 

 and 5 were also apparent in the results of the 

 canonical variate analysis, which showed these 

 two clusters to be continuous. 



The three trends in the clustering pattern indi- 

 cated that coho salmon stocks from similar en- 

 vironments had similar phenotypes. These 

 trends provide some guidance for the transfer of 

 coho salmon stocks. Geographical clustering in- 

 dicates that the phenotypic or perhaps genetic 

 similarity between stocks probably decreases as 

 the distance between stocks increases. Mc- 

 Intyre 8 showed a strong negative correlation for 

 the distance between stream systems and the 

 genetic similarity of the steelhead trout stocks 

 in those stream systems. If a similar relation be- 

 tween phenotype and distance exists among coho 

 salmon stocks, survival rate would be expected to 

 vary inversely with the distance that the stock is 

 transferred from its native stream. The crucial 

 question from the management standpoint, 

 assuming the relationships we found are real, is 

 how far stocks can be transferred before de- 

 creasing survival rate and the increasing genetic 

 impact on the native stocks reduce the practical- 

 ity of such transfers. 



Although geographical distance can be an im- 

 portant factor in selecting a donor stock, other 

 considerations must also be taken into account. 

 The difference between stocks from large and 

 small stream systems illustrates a problem in 

 basing stock transfers primarily on geograph- 

 ical distance. Stocks from large stream systems 

 were more similar to stocks in distant large 

 systems than to stocks in small stream systems 

 that were geographically close. Other environ- 

 mental variables may also differ, affecting the 

 phenotypes of geographically close stocks. 

 Characteristics such as time of peak spawning 

 or transferrin genotype may be closely related to 

 flow and temperature regimes or to disease 

 organisms present in the stream systems. These 

 characteristics and others not included in this 



"Mclntyre. J. D. 1976. The report of interbreedingof arti- 

 ficially propagated and native stocks of steelhead trout. 

 Oreg. Dep. Fish Wildl., Res. Sect, Steelhead Annu. Rep., 22 p. 



116 



