FISHERY BULLETIN: VOL. 85, NO. 3 



P < 0.01 during night for Dover sole. No signifi- 

 cant size differences (P > 0.05) were found during 

 evening for English sole and Dover sole. 



Of the three fishes, rex sole selected the largest 

 Capitella spp. during every time period, with me- 

 dian size ranging from 0.82 to 0.89 mm through- 

 out the diel cycle. Median size of Capitella spp. 

 selected by English sole and Dover sole ranged 

 from 0.62 to 0.66 mm and 0.58 to 0.75 mm, respec- 

 tively. 



Habitat Comparisons 



Differences in number of prey per stomach be- 

 tween English sole captured in habitats where 

 Capitella spp. were present and conspecifics cap- 

 tured in habitats where these polychaetes were 

 absent were highly significant (P< 0.001) at 

 night, but not significant (P > 0.05) during morn- 

 ing, afternoon, and evening (Fig. 4). The diel 

 trends of feeding intensity in the two habitats 

 were strikingly different. Where Capitella spp. 

 were present, feeding intensity increased from 

 afternoon to evening, and then peaked at night 

 (median = 53.5 prey per stomach). By contrast, in 

 habitats where Capitella spp. were absent, feed- 

 ing intensity declined from afternoon to evening. 



MORNING AFTERI^OON EVENING NIGHT 



SAMPLING PERIOD 



Figure 4. — Comparisons of values of feeding intensity for Eng- 

 lish sole between habitats with and without Capitella spp. using 

 the Mann-Whitney U-test. Number of stomachs examined is 

 presented above each bar. Significance level is given above each 

 pair of bars. ***P < 0.001, ns = P > 0.05. 



and reached a minimum (median = prey per 

 stomach) at night. 



DISCUSSION 



Although Capitella spp. accounted for only 

 25.7% of benthic individuals, their importance as 

 prey to English sole, Dover sole, and rex sole was 

 substantial. All three fishes exhibited significant 

 [P < 0.05) numerical and size selection of these 

 polychaetes during all or most of the diel cycle. 

 Based on literature accounts of the food habits of 

 these fishes, the observed importance oi Capitella 

 spp. as prey could not have been predicted di- 

 rectly. 



Most historical accounts of the food habits of 

 the three fishes do not identify Capitella spp. as 

 prey (e.g., Hagerman 1952; Kravitz et al. 1977; 

 Hulberg and Oliver 1978; Pearcy and Hancock 

 1978; Gabriel and Pearcy 1981; Allen 1982; 

 Hogue and Carey 1982). However, most of these 

 studies were conducted in areas where Capitella 

 spp. generally would not be expected to occur in 

 large numbers in the benthos (i.e., the continen- 

 tal shelf off Oregon and California). At least two 

 studies have found that one or more of these 

 fishes consume Capitella spp. Cross et al. (1984) 

 examined the food habits of English sole (n = 13) 

 and Dover sole (n =38) in areas influenced by 

 sewage discharges off Los Angeles, CA. Although 

 C capitata numerically accounted for 40-95% of 

 benthic assemblages, the dietary contributions by 

 this polychaete were small (i.e., 0% for English 

 sole and <10% for Dover sole). Toole (1980) found 

 that C. capitata was a major prey item of juvenile 

 English sole (66-102 mm TL, n = 45) captured on 

 an intertidal sand flat in Humbolt Bay, CA. How- 

 ever, because benthic assemblages were not sam- 

 pled, it is unknown whether these fish were prey- 

 ing nonrandomly on C . capitata . 



Of the three fishes sampled in the present 

 study, rex sole exhibited the greatest degree of 

 selective predation on Capitella spp. This species 

 was the only one to nonrandomly select Capitella 

 spp. based on both prey number and prey size 

 throughout the diel cycle. In addition, rex sole 

 selected the largest Capitella spp. of the three 

 fishes, and included these polychaetes in the 

 largest percentage of total diet during all four 

 time periods. The observed peak in feeding inten- 

 sity at night agrees with past descriptions of rex 

 sole as a nocturnal forager (Kravitz et al. 1977; 

 Allen 1982; Becker 1984). The concomitant peak 

 in percent dietary contribution oi Capitella spp. at 



476 



