FISHERY BULLETIN: VOL 72, NO. 4 



Table 4. — The total number of Panulirus interruptus roaming 

 during precontrol (Homarus absent), experimental (Homarus 

 present), and postcontrol (Homarus absent) periods for shelter 

 and no shelter conditions. Values for the experimental period are 

 one-half the 10-day total so that the totals shown for the three 

 periods are comparable. The maximum number of roaming lob- 

 sters possible for each group, during each period and for each 

 condition (shelter or no shelter) was 2,250. 



in Table 5. There was a significant correlation 

 (r^ = +0.90, P = 0.05) between the total roaming 

 time of each H. americanus {N = 5) and the num- 

 ber of //omaras-initiated behavioral actions. 

 That is, the greater the total roaming time, the 

 greater was the number of behavioral actions. 



Table 5. — Total i/omarus-initiated behavioral actions and 

 roaming time for shelter and no shelter conditions, 

 //omarus -initiated behavioral actions include social contact, 

 Homarus approach, Homarus threat, and Homarus attack. So- 

 cial contact includes somePa?ia/in^s-initiated actions; however, 

 the great majority of the total is due to/Zomart^s-initiated ac- 

 tions. 



the shelter condition was not significantly corre- 

 lated with the number of //omarws-initiated 

 behavioral actions (Table 5;r, = +0.60, P >0.05). 

 The amount of P. interruptus roaming without 

 shelter was significantly greater than the amount 

 of roaming with shelter (Table 4; P = 0.028, 

 Mann-Whitney U test). Similarly, the amount of 

 H. americanus roaming without shelter was also 

 significantly greater than the amount of roaming 

 with shelter (Table 5; P = 0.004, Mann-Whitney 

 U test). 



Panulirus vs. Cancer with No Shelter 



The number of P. interruptus roaming during 

 the three observation periods was: precontrol 

 532, experimental 534, and postcontrol 549 

 (Table 6). There were no significant differences in 

 the amount of roaming during the three observa- 

 tion periods (P > 0.05, Kruskal-Wallis One-Way 

 Analysis of Variance by Ranks). 



Table 6. — The total number of Panulirus interruptus roaming 

 and within 30 cm of the wall during the precontrol (Cancer 

 absent), experimental (Cancer present), and postcontrol (Cancer 

 absent) periods for the Panulirus vs. Cancer no shelter condition. 

 The maximum number of roaming lobsters for each group during 

 each period was 1,000. The maximum number of wall + no wall 

 positions for each group during each period was also 1,000. 



I 



Precontrol 



Experimental 



Postcontrol 



Homarus vs. Panulirus with Sheher 



Distribution of Panulirus 



The numbers of P. interruptus roaming during 

 the three observation periods were: precontrol 

 705, experimental 566, and postcontrol 365 (Table 

 4). The differences between the three test periods 

 were not significant (P > 0.05, Kruskal-Wallis 

 One- Way Analysis of Variance by Ranks). 



The majority of roaming that occurred in both 

 P. interruptus and H. americanus took place 

 during the 1510-1700 h observation period as in 

 the no shelter condition. Seventy-seven percent 

 of P. interruptus roaming and 78% of H. ameri- 

 canus roaming occurred during this period. The 

 amount of roaming done by H. americanus in 



Homarus vs. Panulirus with No Sheher 



The displacement effect H. americanus had on 

 P. interruptus is clearly shown in Table 7. During 

 the precontrol observation period (Homarus 

 absent) P. interruptus was most frequently found 

 within 30 cm of the wall. In all five replicate 

 experiments the subjects spent the majority of 

 their time near the wall. Group 5 was observed 

 within 30 cm of the wall three times more fre- 

 quently than it was observed away from the wall. 

 The other groups exhibited even greater wall — 

 no wall differences. For example. Group 3 was 



1156 



