similarity among the food habits of the three 

 fishes that preyed on benthic invertebrates was 

 calculated using commonly occurring prey (Table 

 3) and Horn's (1966) measure of niche overlap. The 

 overlap was largest between P. vetuliis and G. 

 zachirus (Cx = 0.40). This is because both fishes 

 fed on the same species of amphipods and the 

 polychaete Nothria iridescens. Parophrys vefulus 

 preyed on a very diverse array of invertebrate 

 taxa, while G. zachiriAs appeared to be more 

 selective in its feeding. 



The amount of food overlap among the other 

 species pairs was low (0.19 between P. vefulus: and 

 L. hilineaia and only 0.03 between G. zach irus and 

 L. hilineaia). These low values are explained by 

 the high occurrence of Ophiura lufkeni only in L. 

 bilineata. Also, L. bilineafa fed on members of two 

 scaleworm families, Aphroditidae and Polynoidae, 

 neither of which is represented in the other 

 flatfishes. 



The food habits of the flatfishes that we found to 

 be mainly piscivorous were also different. Eopsetta 

 jordani preyed on various fishes, including benthic 

 agonids, pleuronectids, and cottids, as well as a 

 benthic shrimp and crab, whereas C sordidus fed 

 almost exclusively on the pelagic Engraidis 

 mordax. 



Partitioning of the food resources among the 

 five flatfish species is obvious from our data-the 

 different, syntopic species fed upon different 

 organisms. According to MacArthur and Pianka 

 (1966), a more productive environment should lead 

 to a more restricted diet in terms of different 

 species eaten, but in a patchy environment this 

 does not apply to predators that spend most of 

 their time searching. If the bottom occupied by P. 

 retulus is inhabited by patches of invertebrates, 

 then this species might be such a scavenging 

 "generalist" predator. Rae (1969) documented a 

 food interaction similar to the one in this study 

 between the lemon sole, Microi^fomus kitt, and 

 witch, Glyptocephalu^ cynoglossus, off Scotland. 

 The witch, restricted to muddy bottoms, fed on a 

 more restricted fauna than the lemon sole, whose 

 diet included the hard-bottom species typical of its 

 habitat in addition to species from muddy-bottom 

 types. 



Differences in time of feeding could also account 

 for differences in the species composition of prey. 

 Diel changes in the habits of prey can serve to 

 increase or decrease their exposure to predators, 

 and hence their availability as food (Hobson 1965; 



Jones et al. 1973). More so than the other species, 

 the stomach contents of G. zachina^ were in a late 

 stage of digestion, suggesting that they had fed a 

 longer time before capture than other species. 



The diet of fishes is related not only to their 

 feeding behavior but also to their digestive mor- 

 phology and mouth structure. The size of the 

 mouth relative to body length correlated with the 

 size of food organisms for bothid flounders in 

 Georgia coastal waters (Stickney et al. 1974). 

 Symmetry of the jaws plays an important role in 

 the mode of feeding, as species with symmetrical 

 jaws generally take free-swimming food, while 

 those with asymmetrical jaws are mainly bottom 

 feeders (Yazdani 1969). Flatfishes that feed on 

 polychaetes and mollusks typically have smaller 

 stomachs, larger intestines, and smaller gill rakers 

 with fewer teeth than flatfishes that feed on other 

 fishes (DeGroot 1971; Tyler 1973). The mouths of P. 

 vetuh(^<, G. zachinii^, and L. bilineafa are small,'-' 

 the jaws and dentition are better developed on the 

 blind side (i.e., asymmetrical), the teeth are inci- 

 sorlike (bluntly conical in L. bilineafa), and the gill 

 rakers are without teeth. These morphological 

 adaptations correlate with the preponderance of 

 benthic invertebrates in their diets. The pisci- 

 vores, E. jordani and C. sordidus, on the other 

 hand, have larger mouths,'' nearly symmetrical 

 jaws with sharp teeth, and long gill rakers with 

 teeth. 



Acknowledgments 



We thank William Colgate, Allan Fukuyama, 

 and Valerie Hironaka for identifying amphipods, 

 mollusks, and ophiuroids, respectively. Discussions 

 with Michael Richardson and Robert Carney were 

 useful. This work is a result of research sponsored 

 [in part] by the Oregon State University Sea 

 Grant College Program, supported by NOAA 

 Oflice of Sea Grant, U.S. Department of Com- 

 merce, under Grant #04-5-158-2. 



Literature Cited 



Alverson, D. L. 



1960. A study of annual and seasonal bathymetric catch 



^Length of maxillary into head on ocular side is 4' 4-4y5, 4^/z-5%, 

 and 3'^-4'/5. respectively (Norman 1934); also see Norman for line 

 drawings depicting the relative mouth size of flatfishes discussed 



in this paper. , „,, , , o\ 



^Length of maxillary into head is about ^/s and 2% (nearly S), 

 respectively (Norman 1934). 



989 



