downstream. The greatest density of salmon 

 and trout was where a riffle entered a pool. 

 They were observed in a pool below a riffle at 

 km. 3.2 before and after the first release in 

 1963. Trout over 130 mm. long maintained 

 a central position below the swift current of 

 the incoming riffle. Large trout usually re- 

 mained close to the bottom and swam up quickly 

 for food. Coho salmon and trout yearlings 

 flanked the "large trout area" but were some- 

 what upstream in shallower water and distrib- 

 uted at all depths exceeding 18 cm. Thus the 

 distribution of coho salmon and trout in a pool 

 below the riffle appeared to be related directly 

 to size of the fish. The largest fish were in 

 a central location below the fast inflowing 

 water; smaller fish were in shallower water 

 upstream and shoreward from the larger in- 

 dividuals. At this same site, chinook salmon 

 that were maintaining their positions were 

 in still shallower water shoreward and some- 

 what upstream from the yearling coho salmon 

 and rainbow trout. 



The type of habitat occupied by sculpins 

 was related to species rather than to size. 

 Habitat preferences of the sculpins could not 

 be determined by visual observation (sculpins 

 hide among rocks and sometimes partially bury 

 themselves in substrate) but were indicated by 

 the results of electrofishing. Prickly and re- 

 ticulate sculpins usually were taken from water 

 of low velocity in pools near banks. Coast- 

 range sculpins, C. alenticus, appeared to prefer 

 riffles, and torrent sculpins, C. rhotheus, were 

 at the heads of riffles or throughout riffles of 

 moderate velocity. Densities of sculpins with- 

 in a "microhabitat" increased with increases 

 in cover. 



The distribution of hatchery chinook salm- 

 on within the river was so widespread that 

 some of them passed near all potentially pre- 

 daceous fish in the Elokomin River. Numbers 

 of chinook salmon fry maintained their po- 

 sitions throughout the stream below the hatch- 

 ery but were most numerous in pools below 

 riffles. Others were distributed in other sec- 

 tions of pools or in low-velocity riffles, espe- 

 cially where cover was available. Piscivorous 

 fishes, capable of eating chinook salmon fry, 

 had ample (but not necessarily equal) oppor- 

 tunities to do so. 



PREDATION ON 



CHINOOK SALMON FRY 



Chinook salmon were found in the stomachs 

 of salmon, sculpin, and trout. A brief account 

 is given regarding predation by salmon and by 

 trout. Predation by sculpins was studied more 

 thoroughly. Reported here are the variations 

 between species, the estimated numbers of 

 salmon eaten, and prey-dependent factors. 



By Sculpin 



Variation between species of sculpins. — 

 The numbers of salmon eaten (tables 2, 3, 

 4, and 5), the incidence (table 6) and rate of 

 predation (table 7), and maximum numbers 

 of salmon eaten by individuals (table 8) by 

 the four species of sculpins showed the same 

 trend. The prickly sculpin was the most in- 

 tense predator, torrent sculpin was second, and 

 reticulate sculpin was third; the coastrange 

 sculpin was of little importance as a predator. 



Reasons for variation in intensity of pre- 

 dation among species of sculpins may have 

 been due to differences in: (1) habitat, (2) 

 size of food items preferred, and (3) size of 

 the sculpin. The present data contribute no 



Table 2.— Number of salmon eaten by coastrange sculpins 

 in the Elokomin River after releases of chinook salmon 

 in May 1962 and 1963 



Year 



Sampling 

 date* 



Distance 



from mouth 



of river 



Total 

 sculpins 

 examined 



Sculpins 

 with salmon 

 in stomach 



Salmon in 

 stomach of 



sculpins 



1962 



Number 



1963A 



May 22 

 23 

 24 

 22 

 22 

 23 

 24 



3.2 

 3.2 

 3.2 

 6.4A 

 6.4B 

 6.4 

 13.0 



41 

 44 

 79 

 32 

 33 

 10 

 26 



