GABRIEL and PEARCY: FEEDING SELECTIVITY OF DOVER SOLE 



Terebellidae (SG29: Pista disjuncta), Sterna- 

 spidae (SG29: Sternaspis fossor), and Opheliidae 

 (SGIO: Travisia foetida). Most frequently con- 

 sumed polychaete species were generally smaller 

 bodied, including lumbrinerid (primarily at 

 SG29), nephtyd (primarily at SGIO), paraonid, 

 and capitellid polychaetes (Table 2). Principal 

 prey species are listed in Tables 3 (SG29) and 4 

 (SGIO). 



Pelecypods were the most frequently consumed 

 molluscs, constituting most of the mollusc biomass 

 eaten (Table 1). Macorna spp., Adontorhina cyclia, 

 and Axinopsida serricata were most important at 

 SG29 and Crenella decussata. Huxleyia munita, 

 and Odontogena borealis were common at SGIO. 

 Among crustaceans, amphipods (SG29: Har- 

 piniopsis excavata, H. fulgens; SGIO: Melphidippa 



amorita, Ampelisca macrocephala, Nicipe tumida ) 

 occurred more frequently and in larger numbers 

 than cumaceans, and were consumed more often 

 than molluscs at the deeper station. 



Changes in Diet with Predator Length 



A chi-square test (Tyler") was used to determine 

 the dependency of diet on fish size, based on differ- 

 ences in frequency of occurrence of prey items 

 consumed by predators belonging to different 

 length intervals. The chi-square test showed many 

 size-related variations in the diet of Dover sole 

 taken at SG29 but few in the diet at SGIO. At 



Tyler, A. V. 1969. Computer program.s for analysis of feed- 

 ing heterogeneitie.s related to predator bodv size. Fish. Res. 

 Board Can. Tech. Rep. 121, 49 p. 



Table 3. — SG29: Changes in selectivity with fish size considered by principal prey taxa. Range offish size i 

 heterogeneity between size intervals for each principal prey taxon as described in text. 



divided to maximize 



Taxon 



For fish 

 length 

 (cm) 



Ivlev 

 index of 

 electivity 



Chi-square 

 value 



Signifi- 

 cance' 



For fish 



2 length 



(cm) 



Ivlev 

 index of 

 electivity 



Chi-square 

 value 



Signifi- 

 cance' 



-.07 



38 

 92 



859 



6935 

 37 430 



NS 



'NS = not significant: ■■•?■ 0.001: •• P- 0,01 :"P 0,05, 



^All members of family pooled, 



^All members of genus pooled (species identifications not possible) 



"Members of two species pooled based on taxonomic uncertainty (Nothria elegans and N. iridescens). 



^No members of this genus present in box core, 



^All members of genus pooled (species identifications tentative or uncertain). 



'All nonpnncipal or unidentifiable pelecypods found in fish compared with all pelecypods found in box core. 



^Axinopsida serricata and Adontorhina cyclia poo\e6. 



^No members of this genus present in box cores, value for Harpiniopsis fulgens represents all Phoxocephalidae. pooled. 



'"Based on frequency of occurrence in fish compared with numbers of ophiuroids found in box core. 



NS 



NS 

 NS 



NS 



NS 

 NS 



NS 



NS 



NS 



NS 



753 



