RUGGERONE and ROGERS: ARCTIC CHAR PREDATION ON SOCK EYE SALMON SMOLTS 



consumed/char per 24 h. Other variables such as 

 the number of smolts migrating during the day- 

 light, the percent of smolts migrating during the 

 daylight, the presence of adult sockeye, days after 

 ice-out, river temperature, incident solar radia- 

 tion, light intensity during the evening as a func- 

 tion of cloud cover, and river depth as a function of 

 lake level did not add any new information. 



The model describing char consumption of 

 smolts has the following form: 



N = a + 6(1 - e cP ) + de w + f(l - e~« c ) + hL 3 



02 



where a, b, c, d, /', g, h = empirical constants 



TV = number of smolts con- 

 sumed/char per 24 h 

 P = number of smolts dur- 

 ing previous migration 

 W = smolt weight 

 C = number of smolts dur- 

 ing day of capture 

 migration 

 L = char length. 



A graphical interpretation of the empirically 



40 60 80 100 120 140 



NUMBER OF SMOLTS MIGRATING 11000'SI 



160 



5.0 60 7.0 80 



AVERAGE SMOLT WEIGHT 9 



90 



100 



12r 



300 



340 



380 420 



CHAR LENGTH Imml 



460 



500 



FIGURE 3. — The partial effect of the previous day's migration, 

 the day of capture migration (A), average smolt weight (B), and 

 Arctic char fork length (C) on the number of smolts consumed/ 

 char per 24 h. Smolt consumption was calculated by setting the 

 alternate variables to their mean value. 



derived model is shown in Figure 3. Confidence 

 intervals about each curve are difficult to interpret 

 because they do not consider the concurrent value 

 of other parameters (Draper and Smith 1981) and 

 are not shown. Instead, a plot of smolts consumed 

 by char versus the number of smolts migrating in 

 the previous day's migration demonstrates the ini- 

 tial variability and the basis for the model (Fig. 4). 

 The average predicted consumption rate was 0.8 

 smolts/char per 24 h and the maximum was 5.6 

 smolts/char per 24 h. These predictions were simi- 

 lar to the observed average and maximum con- 

 sumption rates of 0.8 and 6.0 smolts/char per 24 

 h. 



The functional response of char was best de- 

 scribed as a Type II response and was separated 

 into two curves: consumption of smolts versus 

 smolt abundance during the day of char capture 

 and consumption of smolts versus smolt abun- 

 dance during the previous day (Fig. 3 A). Two 

 curves were needed because most char digestion 

 times were longer than 24 h. Thus, the predicted 

 number of smolts consumed/char per 24 h was an 

 average based on 2 successive days of feeding. The 

 maximum partial effect of smolt abundance on the 

 char consumption rate was about 3.7 smolts/char 

 per 24 h when other variables were held at their 

 mean values. 



Smaller smolts and larger char were associated 

 with increased consumption rates by char. Con- 

 sumption of smolts increased exponentially with 



- N 



Q 



LU 

 5 



X 



Z 



O 

 o 



cc 



Ul 

 CD 



2 

 D 

 Z 



2h ' 





JL 



J_ 



_l_ 



J_ 



_L 



_L 



_1_ 



2 « 6 8 I0 12H 16 18 



NUMBER IN PREVIOUS MIGRATION (X10 4 ) 



FIGURE 4. — Smolt consumption/Arctic char per 24 h in compari- 

 son with smolt abundance during the previous day's migration. 



405 



