FISHERY BULLETIN: VOL. 81, NO. 1 



field locations. Contingency table analyses indi- 

 cated no significant differences between loca- 

 tions in 8 of 9 tests (P>0.05, Table 1). Therefore 

 the catches from both field locations were com- 

 bined according to treatment. The number of 

 trap hauls for each stock species was made equal 

 by randomly deleting observations. The hypothe- 

 sis that the CPUE of C. irroratus, C. borealis, 

 and H. americanus is not affected by the pres- 

 ence of other animals inside traps was tested 

 by comparing the total catch of each species in 

 stocked traps with the total catch in control 

 traps. Catches obtained after 24 h immersion 

 time were compared using a x 2 goodness of fit 

 test (Zar 1974). 



In traps containing 8 or 3 lobsters, the total 

 catch of C. irroratus, C. borealis, and H. ameri- 

 canus was significantly reduced (x 2 <2> = 277.8, 

 35.1, 18.2, respectively, P<0.001) (Table 2). In 

 addition, the catch of both species of crabs was 

 significantly lower in 8-lobster treatments than 

 in 3-lobster treatments (C. irroratus, \ 2 w = 22.9, 



Table 1.— x 2 values for 3 X 2 contingency tables comparing 

 strings of each treatment type for Homarus americanus (Ha), 

 Cancer irroratus (Ci), and C. borealis (Cb) between locations. A 

 separate contingency table was made for each species caught. 

 * = P<0.05, @ = expected frequency of one cell was <5. 



P<0.001; C. borealis, x \d = 6.1, P<0.025). The 

 catch of lobsters was not affected by the density 

 of stocked lobsters (x 2 u> = 2.42, P>0.05). The only 

 effect of stocking traps with crabs was to increase 

 the catch of C. borealis in traps stocked with 

 either 3 C. borealis or 3 C. irroratus (for both 

 treatments, x 2 (d = 8.6, P<0.005). Stocking traps 

 with crabs had no effect on the catch of lobsters 

 (P>0.05). 



The average size of animals captured did not 

 differ between treatments for any of the species 

 (Student's t test, P>0.05) (Table 3). 



The results of the laboratory experiments in 

 which lobsters were stocked concurred with 

 those from the field. The catch of both C. irrora- 

 tus and C. borealis was significantly reduced 

 when H. americanus was in the parlor (Table 4). 



Behavior 



Location Within Trap 



The spatial distribution of animals caught in a 

 trap may be affected by behavioral interactions 

 among the trap occupants. To test this hypothe- 

 sis, the proportion of the catch found in the entry 

 section, or "kitchen," in control traps was com- 

 pared with the proportion in the kitchen in 

 stocked traps. All comparisons of proportions 

 were made using the normal approximation for 

 differences between two proportions (Zar 1974). 

 Stocked animals were placed in the parlor. 



In both field and laboratory experiments, a 



Table 2.— Total numbers of Cancer irroratus, C. borealis, and Homarus americanus caught after 24-h immersion time in field 

 experiments. Catch per trap haul is indicated in parentheses; control = empty baited traps; treatment refers to species stocked; 

 n = no. of trap hauls for each treatment level. 



Table 3. — Average size (mm) and standard deviation (SD) of Homarus americanus. Cancer bore- 

 alis, and C. irroratus caught in all traps, locations combined. Size of crabs is carapace width; size 

 of lobsters is carapace length. 



54 



