670 



Fishery Bulletin 89(4), 1991 



Figure 1 



Location of study area and sampling sites 

 during 1981-82. Numbers in parentheses 

 indicate the number of fish sampled, and 

 numbers outside parentheses are sample 

 identification numbers. Data from sites 

 with the same sample number were com- 

 bined for the analysis. 



Margaree I ^ 



I 



skate stomachs were collected off the northwestern 

 coast of Cape Breton I. (Fig. 1). Sampling was con- 

 ducted on an opportunistic basis from various fishing 

 vessels and during research cruises; consequently, 

 different gear types were used (Danish seines, jigs, 

 gillnets, trawls) (Robichaud 1985). Data from samples 

 collected at adjacent sites, on similar bottom, during 

 the same week period and using the same sampling 

 gear were combined and assigned the same sample 

 number (Fig. 1). 



Fork length of cod and total length of skate were 

 measured at sea. However, in samples 3 and 8 where 

 large numbers of cod were caught (207 and 160 cod, 

 respectively), all stomachs were removed but, due to 

 limited time, only a random subsample of cod was 

 measured. Possible differences in mean sizes of cod and 

 skate between samples were tested by a "one-way" 



analysis of variance combined with a Duncan test 

 (Duncan 1955) after transforming the data with Log(x). 



Stomach removal was performed by (1) tying off the 

 anterior part of the stomach with plastic "spaghetti" 

 tubing identified by a specimen number, and (2) 

 separating the esophagus in front of the tubing and cut- 

 ting the posterior part of the pylorus in order to pre- 

 vent the loss of stomach contents. Stomachs were then 

 preserved in 4% formaldehyde and seawater and trans- 

 ported to the laboratory for analysis. For each cod and 

 skate, data on capture location, depth, bottom type, and 

 gear type were recorded. 



In the laboratory, stomach contents were examined 

 and prey remains identified to the lowest taxonomic 

 level that the state of digestion permitted. Food in each 

 taxon was weighed and its contribution, in terms of 

 percent by weight of the stomach contents, was deter- 



