Table 8.— Maximum number of chinook salmon eaten by 

 the most predaceous specimen of each predator species 

 in the Elokomin River and Herman Creek 



food. For example, the larger food items were 

 consumed by the larger individual among mot- 

 tled sculpins (Daiber, 1956) and among prickly 

 and torrent sculpins (Northcote, 1954). Be- 

 cause salmon constitute relatively large food 

 items, only the larger sculpins are predators 

 on salmon fry. Prickly sculpins preyed most 

 heavily on salmon and were also the largest 

 species of sculpin. 



Most of the prickly sculpin collected were 

 over 110 mm. long — the largest was 221 mm. 

 (fig. 2). Few reticulate (fig. 3) or torrent 

 sculpins (fig. 4) were over 100 mm. long, and 

 a few coastrange sculpin were over 110 mm. 

 (fig. 5). The smallest prickly sculpin to prey 

 on salmon was 103 mm. in the Elokomin River 

 and 79 mm. in Herman Creek, compared to 



50 n 



x 40 H 

 go 



ELOKOMIN 



N =90 

 P = 20 



73 mm. for reticulate sculpins in the Elokomin 

 River and 66 mm. in Herman Creek, and to 

 73 mm. for coastrange sculpins and 68 mm. 

 for torrent sculpins in the Elokomin River. 



Body length of a sculpin does not entirely 

 determine its capacity for seizing and swallow- 

 ing large items of food because of different 

 mouth sizes among species (fig. 6). For ex- 

 ample, Northcote (1954) showed torrent scul- 

 pin to be more piscivorous and to feed on larger 

 food items at shorter lengths than prickly 

 sculpin and suggested that this difference was 

 related to the larger mouth of the torrent 

 sculpin. 



Body length-mouth gape relations were 

 studied for differences among species of 

 sculpin. Fish from Elokomin River, Herman 

 Creek, and other streams of the lower and up- 

 per Columbia River and Puget Sound drainage 

 were included to obtain broad geographic cov- 

 erage to lessen stream effects. After testing 

 for fit the regressions of mouth gapes that pro- 

 gressively increased with length, I concluded 

 that a straight line relation was best. 



The D 2 (distance function as described by 

 Rao, 1952) ' statistics were used to measure 



1 The correlation of mouth gape with body length 

 was removed by using body length as the first variable 

 for part of the analysis. Treatment of the data by 

 the D 2 method involves use of only the mouth gape 

 as the independent variable. 



HERMAN 



N = 116 

 P = 44 



CR. 



OOOQOOOOOOOOOOOO O 



r-cocno — c\jm^ir>«)r>-cocr>o— cm ro 

 CMCMCMCM 



I I I I I I I I 1 I I I I , , , ( 



cDr-cocr>o— cMfo^-mujr^cocrio— cm 



CMCM CM 



LENGTH (MM.) 



Figure 2.— Length frequencies of prickly sculpin collected after releases of salmon in the Elokomin River in 1962 

 and 1963 and in Herman Creek in 1962. Salmon predators are shaded, and nonpredators are unshaded. N = 

 total number of sculpins in sample of which P = number of predators. Total number of sculpins is smaller than 

 in Table 3 because Table 3 includes some sculpins taken prior to the release of hatchery reared salmon. 



