FISHERY BULLETIN: VOL. 86, NO. 2 



Table 2. — Diet overlap (McCabe et al. 1983) based on number (n) 

 and weight (W) of prey by sampling period for juvenile salmon in 

 Porcupine Creek estuary, 26 May-7 July 1981. The number of 

 stomach samples is in Table 1 . 



Table 3. — Comparison of mean number of the 12 most important 

 prey and total of all prey per salmon stomach from Porcupine Creek 

 estuary, 26 N^ay- 1 June 1 98 1 . Means followed by the same letter 

 are not significantly different (Kruskal-Wallis analysis of variance, 

 P>0.05) compared within a row and across columns. The number 

 of stomach samples is in Table 1 . 



Prey item 



Pink salmon Chum salmon Coho salmon 



Total 



115a 



216 a 



b 

 4 b 

 6 b 



19 b 



Table 4. — Mean length (mm) of prey items in salmon stom- 

 achs from Porcupine Creek estuary, 26 May-7 July 1981. 

 Mean prey length within prey taxa was significantly greater 

 for coho salmon than for pink or chum salmon (sign test, 

 n = 12 means, P=0.03 and P =0.003, for coho salmon vs. 

 pink and chum salmon, respectively). The number of stom- 

 ach samples is in Table 1 . 



T — I — I — I — I I I "1 — I — *T — I — T — I — I — r 

 0.5 2.5 4.5 6.5 8.5 10.5 12.5 >1 4 

 Prey Length (Interval Midpoint, mm) 



Figure 3. — Relative frequencies of length of prey eaten by pink, 

 chum, and coho salmon in the Porcupine Creek estuary in 1981. 

 Total prey measured were 687 in 11 pink, 5,634 in 63 chum, and 

 1,179 in 53 coho salmon. 



Table 5. — Spearman rank correlations between number of prey 

 items and fork length of juvenile salmon. Because of the large 

 number of correlations tested, significance levels were adjusted by 

 multiplying the probability P by the number of tests for each salmon 

 species. The number of stomach samples is in Table 1 . 



Prey item 



Pink salmon Chum salmon Coho salmon 



0.05 

 -0.02 



0.10 

 -0.03 

 -0.34* 



0.04 

 -0.37- 



0.36* 

 -0.19 

 -0.50* 

 -0.18 



-0.04 

 -0.11 



-0.04 



0.35* 



'None present. 



'None present in any stomachs. 

 *Adjusted probability P < 0.05. 

 **Adjusted probability P < 0.01. 



after adjusting for differences in coho salmon FL 

 (Table 6). Thus, changes in diet were mainly re- 

 lated to increasing size of coho salmon as they 

 migrated through the estuary. 



DISCUSSION 



Both size and diet can affect predation and com- 

 petition among juvenile salmon in an estuary. A 



218 



