FISHERY BULLETIN: VOL 72, NO. 4 



areas (areas within 30 cm of the wall). There 

 was a reversal of the precontrol wall position of 

 four out of five P. interruptus groups during the 

 experimental period (Table 7). When H. ameri- 

 canus was removed for the postcontrol period, 

 three of the four groups which had reversed 

 their precontrol position during the experimental 

 period moved back to the wall position. Group 

 2 remained in the no wall position during the 

 postcontrol. The exact reason for this is unknown, 

 but perhaps this group's encounters with H. 

 americanus were somehow more intense and thus 

 the enounters had a more lasting effect on the 

 behavior of Group 2. The individual used in this 

 group was by far the most active H. ameri- 

 canus tested for both the shelter and no shelter 

 condition and initiated the second highest number 

 of behavioral actions (Table 5). 



Group 5 remained in the wall position more 

 than the no wall position during all three test 

 periods. Thus, the H. americanus introduced to 

 this group had little effect on its distribution 

 within the tank. The reason for this is probably 

 the low level of activity and aggression shown 

 by this individual. It showed the least amount of 

 roaming (approximately one-half the value shown 

 by the next least active individual) and the lowest 

 number of behavioral actions of any lobster in 

 the no shelter condition. It directed only 16 

 attacks at P. interruptus, less than one-half of 

 those shown by the next least aggressive H. 

 americanus. 



In contrast to the above, C. antennarius had 

 no effect on the distribution of P. interruptus in 

 the absence of shelter. All five groups of P. 

 interruptus spent nearly all their time near the 

 wall during each of the three test periods 

 (Table 6). That is, P. interruptus groups were 

 never observed spending more time in the 

 no wall position than in the wall position when 

 paired with C. antennarius. 



Douglis (1946) found that when//, afnericanus 

 was present in a large aquarium tank with one or 

 more blue crab, Callinectes sapidus, spider crab, 

 Libinia emarginata, or hermit crab, Pagurus 

 polycarus, "it tended by pushing and fighting to 

 keep the crabs on the opposite side of the tank 

 from itself." Thus, although details were not 

 given, it appears that H. americanus displaced 

 crabs that, by virtue of their claws, would 

 seem to be much better prepared to cope with 

 H. americanus than is P. interruptus. 



The presence of more shelter than was neces- 

 sary for the number of lobsters in a tank sub- 

 stantially decreased the number of P. interrup- 

 tus displaced from the wall position. This was 

 primarily due to the fact that when shelter was 

 present, all lobsters spent the majority of their 

 time in the shelter. This interpretation is sup- 

 ported by comparing data on the roaming time 

 for the shelter and no shelter conditions. The 

 presence of shelter significantly reduced the 

 amount of roaming in both P. interruptus and 

 H. americanus. 



The greater frequency of agonistic interactions 

 observed between//, americanus and P. interrup- 

 tus during the 1510-1700 h observation period, 

 as compared with the 0800-1000 and 1200-1400 h 

 observation periods, was also no doubt due to 

 the higher level of locomotion during this period. 



A great amount of variability in the activity 

 and aggression of individual H. americanus was 

 also observed in our study. Some individuals 

 were very active and aggressive, while others 

 were neither. In our study these variables were 

 not related to the sex or size of//, americanus. 

 In the shelter condition, a female exhibited the 

 greatest number of behavioral actions, while in 

 the no shelter condition, it was a male. Of the 

 10 H. americanus tested, these lobsters ranked 

 fourth and seventh in size (Table 5). 



CONCLUSIONS 



Although we cannot predict with certainty the 

 effect a large introduced population of//, ameri- 

 canus would have on P. interruptus and other 

 decapod crustaceans native to southern Cali- 

 fornia, our data suggest that an adverse outcome 

 from such an introduction could occur. The types 

 of behavioral interactions we observed in the 

 laboratory between H. americanus and P. inter- 

 ruptus would most likely also occur in the field. 

 This conclusion is strengthened by the studies 

 and observations of other workers who have com- 

 pared and found close agreement between the 

 laboratory and field behavior. Our data show that 

 a large percentage of the behavioral actions of 

 H. ainericanus toward P. interruptus probably 

 would involve aggressive actions. Assuming that 

 individuals of these two species came in con- 

 tact with one another in nature, these aggres- 

 sive actions could have direct or indirect effects 

 on the distribution and abundance of P. inter- 

 ruptus. Thus, our evidence suggests that it would 



1158 



