near. Then they made short, quick lunges at the 

 prey. Little predation occurred during the day and 

 I never ol)served a sculpin catching a salmon. 



Stocks of fry used for testing appeared normal, 

 healthy, and vigorous. A reserve of fry of a species 

 was maintained in holding aquaria during and 

 after testing without mortality— in fact dead or 

 inferior fry were never observed over 2 yr. Pretest 

 salmonids held in stream aquaria often main- 

 tained positions in the faster moving water. In- 

 dividuals that may have been inferior as indicated 

 by use of slow water shallows, by swimming at the 

 downstream end, or by impingement on the outlet 

 screen were never observed. 



Results showed variation between rates of 

 predation on a prey species, size of prey species, 

 and on temperature and length of daylight during 

 testing. Predation by torrent sculpin was least on 

 coho and chinook salmon, intermediate on sockeye 

 salmon and steelhead trout, while practically 

 complete on pink and chum salmon (Table 1). Chi- 

 square analysis showed significant differences 

 between all species except for 2 of the 10 com- 

 binations tested: sockeye salmon-steelhead trout 

 and pink-chum salmon. The number of survivors 

 per test varied considerably for the chinook salm- 

 on and steelhead trout. Steelhead trout were 

 relatively deep bodied but shorter than salmon fry 

 and among the salmon, chum and pink were thin 

 bodied (Table 1 shows lengths; body volume de- 

 terminations indicated chinook, coho, and sockeye 

 salmon had as much as twice the displacement of 

 the other species). Testing of chum salmon and of 

 steelhead trout was a month or two later in the 

 spring when temperatures were higher (Table 1) 

 and duration of daylight was longer than for other 

 species. 



Innate Predator Avoidance of Species 



Differences in rates of predation on the study 

 species are not well explained by observed 

 differences in behavior, size of prey, ambient con- 

 ditions, or predator related effects but may be due 

 to innate behavior after emergence of fry from 

 the gravel. The only species with greatly divergent 

 behavior in the stream aquaria was the steelhead 

 trout. Remaining near the water surface during 

 day effectively removes them from the influence 

 of sculpin predators; however, they may settle to 

 the substrate at night, a time when sculpins are 

 more effective predators (Patten 1971b). 



The larger prey species, those having the lon- 



gest body lengths and being relatively deep 

 bodied, were not always those with the higher 

 survival. Chinook, coho, and sockeye salmon were 

 the largest. Chinook and coho had the highest sur- 

 vival but the sockeye salmon, the largest prey, had 

 survival similar to the steelhead trout, the smallest 

 prey. Chum and pink salmon were slim and as long 

 as coho and longer than steelhead trout, but their 

 survival was lowest of the species studied. If size 

 of prey or satiation of predators from greater food 

 volumes influenced rates of predation, these fac- 

 tors were apparently less important than other 

 effects on a species level. 



Length of day or temperature had no apparent 

 effect on rate of predation. Sculpins are most 

 predaceous on salmon at times of marginal light 

 intensity (Patten 1971b), which might suggest 

 they are more serious predators at times of shorter 

 day lengths. Trends between intensity of sculpin 

 predation on fry and temperatures observed dur- 

 ing this and other studies have never been ob- 

 served. 



The data show strong interspecific variations of 

 the study species in vulnerability to predation by 

 the torrent sculpin. I suspect a difference in innate 

 behavior exists; some species are better able to 

 evade predation. Furthermore, the early life his- 

 tory and behavior of the study species may be 

 linked to their predator avoidance abilities. Chum, 

 pink, and sockeye salmon quickly migrate from a 

 stream environment to the sea or a lake where 

 they form schools (Mason 1974, has observed chum 

 salmon forming loose aggregations in estuaries). 

 Schooling may aid these species in avoiding 

 predation (Shelbourn 1966). Chinook and coho salm- 

 on and steelhead trout on the average form loose 

 aggregations in streams during a period of growth 

 before migrating to the sea. Forming loose 

 aggregations would increase feeding opportuni- 

 ties in streams. Density of predators may be high 

 in this situation (Patten 1971a) and survival is at- 

 tained by a well-developed avoidance response for 

 chinook and coho salmon. 



Steelhead trout fry had a comparatively high 

 mortality among stream resident species that may 

 have been related in part to their behavior during 

 tests, to their small size or an inferior predator 

 avoidance response. Their survival, at least during 

 the early fry stage, may be increased by 

 unavailability through selection of a protective 

 habitat. Hartman (1965) described the microhabi- 

 tat of recently emerged steelhead trout and coho 

 salmon in the Chilliwack River, British Columbia, 



933 



