GABRIEL and PEARCY FEEDINC SELECTIVITY OF DOVER SOLE 



plaice, Pleuronectes platessa, lies between 1.5 and 

 2.5; and stomach enzyme activity in plaice may 

 cease at pH levels above 5.5 (Bayliss 1935). How- 

 ever, optimal pH levels in Dover sole stomachs and 

 effects of food on local pH levels are unknown. 



Relatively indigestible, low caloric ophiuroid 

 arms were surprisingly abundant in the stomachs 

 of Dover sole. Rae (1956) reported frequent occur- 

 rences (up to 309^ of stomachs sampled) of 

 ophiuroids in the guts of lemon sole, Microstomus 

 kitt. Ophiuroids may be easy to capture and read- 

 ily available. Alternatively, they could provide 

 some required nutrient unavailable in other food 

 sources. Finally, different rates of stomach evacu- 

 ation and digestion could affect our results. The 

 importance of ophiuroids may be easily overesti- 

 mated because ophiuroids may remain in the 

 stomach longer than the soft-bodied polychaetes, 

 small molluscs, or crustaceans. Their arms were 

 frequently tangled in a bulky, inflexible mass 

 which may move slowly through the digestive 

 tract. If a stomach contained only one or two food 

 items, ophiuroid arms were usually present. Less 

 digestible food items in the diet of fishes often have 

 slower gastric evacuation rates ( Fange and Grove 

 1979). This would result in a longer "residence 

 time" for ophiuroids than other prey taxa and 

 suggests that the diet of Dover sole is principally 

 composed of polychaetes. De Groot (1971) 

 categorized M. pacificus as a polychaete-mollusc- 

 (echinoderm) feeder, characterized by a small 

 esophagus and stomach and complicated intesti- 

 nal loop, an adaptation characteristic of flatfishes 

 which feed on polychaetes which are often con- 

 taminated with indigestible items, e.g., tubes. 



On first inspection, it appears that Dover sole do 

 not feed more selectively on energetically more 

 profitable prey when these prey types are more 

 common. For example, the value of E of 

 polychaetes at SG29 was lower than at SGIO (0.13 

 vs. 0.33, Table 5) while the abundance of 

 polychaetes at SG29 was higher than at SGIO (886 

 vs. 396/m^ 669c vs. 279^ of the total numbers of 

 benthic animals in the box cores (Table 7)). How- 

 ever, the frequency of occurrence of principal prey 

 often changed with fish size at SG29, indicating 

 prey selection, whereas most prey species occurred 

 at statistically equal frequencies over the entire 

 size range offish sampled at SGIO. 



These size-related changes in selectivity at the 

 two locations may be related to availability or 

 energetic advantage of the prey. Prey body size 

 was significantly correlated with size of fish at 



which a prey increased in importance in diet at 

 SG29 (Figure 4). Large fish were apparently more 

 successful than small fish at capturing large prey. 

 Small fish may be limited to smaller, slower mov- 

 ing or weaker prey by mouth size or body strength. 

 These predator-prey size relationships are consis- 

 tent with those observed by Schoener (1971) for 

 Anolis lizards: as predator size increased, average 

 prey size increased. Ross (1978) also reported that 

 mean size of prey increased with fish size for the 

 leopard sea robin Prionotus scitulis, >90 mm. 



Depth of prey in sediment is also significantly 

 correlated with size offish at which a prey species 

 begins to occur more significantly (SG29; Figure 

 6). Although prey depth and prey body size were 

 not statistically correlated, the small-bodied prey 

 found deep in the sediment were usually not the 

 same species which increased in frequency in 

 larger fish. Thus, large fish apparently are physi- 

 cally capable of extracting large-bodied, deeply 

 buried prey from sediment while smaller fish are 

 not. Learning as well as extraction capability may 

 be important in successful extraction of large 

 polychaetes. 



Since the distribution of prey species was shal- 

 lower at SGIO than at SG29, the physical advan- 

 tage afforded large fish in the exploitation of prey 

 buried deep in the sediments may be eliminated at 

 SGIO. When depths of species common to both sta- 

 tions were compared, nearly all species were found 

 closer to the surface at SGIO, although the differ- 

 ences are not often statistically significant be- 

 cause of small sample size at SGIO. 



Two potential instances of increased selectivity 

 in the face of increased abundance of a profitable 

 or preferred prey are suggested in this study. First, 

 the abundance of polychaetes, a preferred taxon, 

 was lower at SGIO than SG29 (Table 7). Few sig- 

 nificant changes in the frequency of occurrence of 

 prey occurred with fish size at SGIO where large 

 fish may have had to consume any polychaete en- 

 countered, regardless of size and/or location, to 

 meet their energetic requirements. In other 

 words, the energetic advantage arising from size- 

 selective specialization may disappear as abun- 

 dance of preferred food items decreases, as found 

 for bluegills by Werner and Hall ( 1974). 



Second, size-related availability results in dif- 

 fering effective prey densities to larger vs. smaller 

 fish at SG29. As fish size increases, a wider range 

 of prey may become available and so prey densities 

 are effectively higher for larger fish. Selectivity 

 increases with fish size (Figure 5). Since body size 



761 



