Robichaud et al.: Selection of Chionoecetes opilio and Hyas spp_ by Oadus morhua and Raja radiata 



671 



mined. All unidentified portions and material such as 

 vegetable matter, stones, and mucous were categor- 

 ized as "others". Empty stomachs, which accounted 

 for less than 9% of the total stomachs sampled, were 

 not used in the analysis. Crab prey were individually 

 identified to species, weighed, measured [carapace 

 width (CW)] and sexed. For each sample, total number 

 of individuals for each crab species, total weight of in- 

 dividuals, frequency of occurrence, and mean number 

 of crabs per stomach were determined. Size-frequency 

 distributions were constructed for each crab species. 

 Relationship between the carapace width of crab and 

 body length of cod or skate predator was also assessed 

 using scattergram plots. 



The bottom substrate was sampled using a Van Veen 

 grab. Substrate characteristics were also based on sedi- 

 ment information, collected also with a Van Veen grab, 

 during a juvenile snow crab survey done concurrently 

 with this study (Robichaud 1985, Robichaud et al. 

 1989). 



Results 



The average number of H. arcmeus and H. coarctatus 

 per cod stomach was 0.37 [variance (S) 1.04] and 0.42 

 (S 1.82), respectively. Predation by cod on the two toad 

 crab species followed similar patterns. The average 

 number of both toad crab species per cod stomach was 

 higher on gravel bottom [0.59 (S 1.54) and 0.53 (S 1.69) 

 for H. araneus and H. coarctatus, respectively] and 

 lower on sand and mud bottom [0.19 (S 0.56) and 0.33 

 (S 1.93) forH. araneus andH. coarctatus, respectively]. 

 In addition, the number of both toad crab species per 

 skate stomach was low [0.02 (S 0.02) and 0.06 (S 0.16) 

 for H. araneus and H. coarctatus, respectively]. Be- 

 cause there was no significant differences (P>0.05, 

 using i-test) in number per stomach between the two 

 toad crab species, and because of the large proportion 

 (23%) of toad crabs that was unidentifiable and cate- 

 gorized as Hyas species, all toad crabs were combined 

 for the stomach analysis. 



There were significant differences (P<0.05) in the 

 average size of cod between samples (Fig. 2). There 

 were no significant differences (P>0.05) in the mean 

 size of skate between samples (Fig. 2). 



Cod stomach analyses 



The stomach contents of cod caught by Danish seiners 

 or by trawls over deep (100-141 m) muddy or sandy 

 bottoms during the spring of 1981 and 1982 comprised 

 a large percentage by weight of snow crab (Fig. 3; 

 Table 1, samples 1-3). The percentage by weight of 

 snow crab found in cod stomachs ranged between 40.4 



and 66.2%. The mean number of snow crab per cod 

 stomach ranged between 0.13 and 0.39, respectively 

 (Fig. 3; Table 1). In the same samples, the percentage 

 by weight of toad crab (Hyas araneus AND H. coarc- 

 tatus, combined) ranged between 1.8 and 5.2%, and the 

 mean number of toad crab per cod stomach ranged 

 between 0.10 and 0.53 (Fig. 3; Table 1). The high per- 

 centage of the total weight of snow crab in cod stom- 

 achs during spring was mainly made up of large soft- 

 shelled males. Whole soft-shelled crabs were usually 

 found compacted into a single pliable ball in cod and 

 skate stomachs. On removal from the stomachs, they 

 could be unfolded and measured. Other prey taxa iden- 

 tified were invertebrates, such as amphipods, euphau- 

 siids, and molluscs, and minor quantities of fish (Fig. 

 3, samples 1-3). 



In May 1982, cod taken by jigging over gravel bot- 

 tom at shallow depth (55 m) contained a negligible 

 amount of snow crab (0.4% of overall weight) (Fig. 3, 

 sample 4). In the same sample, the percentage by 

 weight of toad crab was higher (27.8%) and the mean 



