RICHARDS ET AL.: BEHAVIORAL INTERACTIONS OF AMERICAN LOBSTER AND CANCER CRABS 



Table 4. — Total number of Cancer irroratus or C. borealis 

 caught in 10 laboratory trials of each treatment and catch spe- 

 cies. ** = P<0.001, \ 2 goodness of fit test. 



greater proportion of the crab catch was found in 

 the kitchen of 8-lobster treatments than of con- 

 trols (Tables 5, 6). Stocking traps with 3 lobsters 

 had no effect on the distribution of crabs, and lob- 

 sters were unaffected by either stock density of 

 lobsters (P>0.05). 



Interspecific interactions between C. irroratus 

 and C. borealis apparently influenced the distri- 

 bution of these species inside traps. In traps 

 stocked with either 3 or 8 C. irroratus, the pro- 

 portion of the C. borealis catch found in the 

 kitchen was significantly greater than in con- 

 trols (Z = 2.50, P<0.01). In traps containing 3 C. 

 borealis, the proportion of the C. irroratus catch 

 found in the kitchen was significantly greater 

 than in controls (Z = 2.50, P<0.01), but no effect 

 was seen in traps stocked with 8 C. borealis 

 (P>0.05) (Fig. 2). 



o 



Q_ 



o .02 

 rr 



*" 



C. irroratus 





a 



5^ 



C 4) 



C. borealis 



SPECIES CAUGHT 



Figure 2.— Proportion of Cancer irroratus and C. borealis 

 found in the kitchen of traps stocked with congeners in field 

 experiments. All data obtained after one setover day are in- 

 cluded, n = number of trap hauls, ** = significant difference 

 (P<0.01) between treatment and control, using normal ap- 

 proximation for differences between two proportions (Zar 

 1974). 



Table 5.— In field experiments, spatial distribution of Cancer 

 irroratus, C. borealis, and Homarus americanus catch in traps 

 stocked with H. americanus (Ha). All data obtained after one 

 setover day are included. Proportion of catch found in the 

 kitchen of stocked traps was compared with controls using 

 normal approximation for differences between two propor- 

 tions (Z) (Zar 1974). n = number of trap hauls; * = P<0.05, 

 ** = P<0.001. 



Table 6. — In laboratory experiments, spatial distribution of 

 the Cancer crab catch in traps stocked with Homarus ameri- 

 canus (Ha). Proportion of catch found in the kitchen of stocked 

 traps was compared with controls using normal approxima- 

 tion for differences between two proportions (Z) (Zar 1974). 



Competition Inside Traps 



To further investigate how the location of ani- 

 mals in a trap is affected by behavioral interac- 

 tions, competition for preferred areas in the trap 

 was studied in the laboratory. Frequency of occu- 

 pation was used as an index of preference and 

 was measured as the number of times a given 

 position was occupied when censused every 15 

 min. The observed distribution of animals was 

 compared with an expected uniform distribution 

 using a x 2 goodness of fit test. For lobsters and for 

 each crab species in the absence of lobsters, the 

 preferred position in the parlor was underneath 

 the entry head (C. irroratus, x 2 w = 202.0, P< 

 0.001; C. borealis, x \v =51.8, P<0.001; H. ameri- 

 canus, x 2 (4> = 744.2, P<0.001). When lobsters 

 were present, the number of crabs in the parlor 

 decreased sharply, so comparisons between lob- 

 ster-stocked and control traps were made using 

 proportions. In the presence of lobsters, the 

 preference of both crab species changed (C. irro- 

 ratus, Z = 2.26, P<0.01; C. borealis, Z = 5.97, 

 P<0.001). Cancer irroratus occupied the middle 

 of the parlor, and C. borealis occupied the cor- 

 ners most frequently when H. americanus was 



55 



