FISHERY BULLETIN: VOL. 81. NO. 1 



present (C. irroratus, x \v = 82.3, P<0.001; C. 

 borealis, x 2 (4, = 52.5, P<0.001) (Table 7). 



Space inside the trap was partitioned into ver- 

 tical strata. Both crab species showed a signifi- 

 cant increase in occupation of the top part of the 

 trap when lobsters were present (C. irroratus, 

 0.47 vs. 0.79, Z = 4.87, P<0.001; C. borealis, 0.21 

 vs. 0.38, Z = 1.76, P<0.05). This contrasts with 

 99% occurrence of lobsters in the bottom portion 

 of the trap. 



from the parlor did not increase in lobster- 

 stocked traps for either species (C. irroratus, Z= 

 1.37, P>0.05; C. borealis, Z = 0.37, P>0.05). 



DISCUSSION 



Trap Efficiency 



The results of the field and laboratory experi- 

 ments demonstrate that the presence of lobsters 



Table 7.— Laboratory-observed frequency and relative frequency of occupation of positions in the 

 parlor by Cancer irroratus, C. borealis, and Homarus americanus. Counts were weighted to com- 

 pensate for unequal availability of positions due to trap design. * = significant (P<0.01) x 2 val- 

 ues for frequency of occupation and preferred positions; + = significant (P<0.01) differences in 

 occupation of a particular position in lobster-stocked traps and controls; ctl = control; lob = 5 lob- 

 sters stocked. 



Trap Entry and Escapement 



Laboratory observations revealed that C. ir- 

 roratus and C. borealis respond differently to 

 traps stocked with H. americanus. The presence 

 of H. americanus did not affect the number of C. 

 irroratus entering the kitchen (39 vs. 33, x 2 u> = 

 0.35, P>0.05); however, significantly fewer 

 C. borealis entered when H. americanus were 

 stocked (35 vs. 8, x 2 (d = 18.2, P<0.001). 



The proportion of C. irroratus which moved 

 from the kitchen to the parlor was significantly 

 reduced in lobster-stocked traps (0.81 vs. 0.23, 

 Z= 2.73, P<0.0001). The proportion of C. bore- 

 alis entering the parlor did not decrease signifi- 

 cantly when H. americanus was present (0.53 

 vs. 0.31, Z = 0.58, P>0.05); however, the number 

 of C. borealis that had entered the kitchen was 

 relatively low. 



The proportion of both C. irroratus and C. bo- 

 realis which escaped the kitchen increased sig- 

 nificantly in the presence of H. americanus (C. 

 irroratus, 0.23 vs. 0.55, Z = 2.86, P<0.005; C. bo- 

 realis, 0.26 vs. 0.63, Z= 1.97, P<0.025). Escape 



reduces the CPUE of crabs, and provide a pos- 

 sible explanation for the inverse relationship be- 

 tween lobster and crab catches seen in other 

 studies (e.g., Stasko 1975; Krouse 1978; Fogarty 

 and Borden 1980). This effect appears to be den- 

 sity-dependent since fewer crabs were captured 

 when a large number of lobsters were present. 

 Factors other than behavioral interactions 

 could cause negative correlations between lob- 

 ster and crab catch rates. Cancer irroratus is 

 often spatially separated from C. borealis and H. 

 americanus in Narragansett Bay (Jeffries 1966; 

 Fogarty 1976). Such discontinuous distributions 

 could result in inverse catches of C. irroratus and 

 H. americanus, or of C. irroratus and C. borealis, 

 but do not explain the differences seen in the 

 catch of adjacent traps in this study. Other fac- 

 tors known to affect catchability (e.g., size, sex, 

 reproductive condition, molt stage) were held 

 constant among stocked animals used in the dif- 

 ferent treatments. Temperature changed little 

 over the course of the study (average surface tem- 

 perature, 21.9°±2.15°C). This and other environ- 

 mental variables would have affected all treat- 



56 



