HOBSON: ACXrVlTY OF HAWAIIAN REEF FISHES 



the last individual of T. duperrey active in 

 my immediate area take shelter (Figure 2). 

 Each of these individuals swam in and out of 

 coral crevices several times before finally slip- 

 ping into a hole where it remained. On sub- 

 sequent close inspection I found each one lying 

 on its side in the hole, and intermittent checks 

 over the next hour disclosed each still in the same 

 position. Following one such observation, a pre- 

 dawn inspection the next morning found the fish 

 in the same spot. 



There is evidence that, on a given evening, an 

 individual is intent on resting in a particular 

 spot. Two incidental observations give credence 

 to this view. One evening I lay prone and im- 

 mobile across a large coral head (Porites) , 

 watching the many T. duperrey around me, and 

 waiting to note the last one that took cover. 

 The behavior of one individual was unusual: it 

 circled close by regarding me more intensely 

 than seemed normal, then swam away. Several 

 times this same fish reappeared, repeated this 

 behavior, then swam away again. Curious, I 

 abandoned my position and followed when the 

 fish reappeared. After being led in a circle, I 

 found myself back at the original location, where 

 the fish slipped into a coral crevice in the spot 

 where I had been lying. Only then did it occur 

 to me that this fish had behaved abnormally 

 probably because I had been blocking entry to its 

 nocturnal resting spot. Other individuals of T. 

 duperrey were still active in the area when this 

 incident occurred, making it uncertain whether 

 the fish actually had been delayed in attaining 

 cover. A similar incident later was more con- 

 clusive. Again, I rested motionless, watching 

 the many individuals of this labrid during the 

 evening transition period, this time leaning 

 against a vertical wall of coral (Porites). One 

 individual swam close by, then moved to a low 

 ledge about 2 m away, where it then swam back 

 and forth, watching me continuously. Several 

 times it momentarily left the ledge, approached 

 me, and then darted back to the ledge. Recalling 

 the earlier incident, I did not move from my 

 station until long after all other fish of this spe- 

 cies had gone under cover. Finally, 12 min after 

 sunset (light 0.8 ft-c), with the behavior of the 

 labrid under observation still unchanged, I 



100 



10 - 



o 



I 



I 

 o 



0-. • 



0.1 



10 5 5 10 15 



TIME RELATIVE TO SUNSET (min) 



20 



Figure 2. — Some characteristics of certain labrids dur- 

 ing evening twilight. Each lettered symbol represents a 

 species: A. Thalassoma duperrey — when the last indi- 

 vidual took cover on five evenings. B. Labroides phthir- 

 ophagus — when one individual took cover on five eve- 

 nings. The circle represents mean values of both time 

 and incident light when the event occurred ; the vertical 

 line, the range of incident light values; and the hor- 

 izontal line, the range in time relative to sunset. The 

 three diagonal lines represent: the maximum light 

 values for each minute relative to sunset during all 

 evening twilight observation periods (top line) , the min- 

 imum values for each minute (bottom line), and the 

 mean values (heavy median line). 



moved away. Immediately, it swam to my form- 

 er location and slipped into a crevice. Later, 

 using a light, I could see this wrasse lying on its 

 side, wedged back in the crevice. (These two 

 incidents are not included among the data rep- 

 resented in Figure 2, which includes only situ- 

 ations judged normal.) 



These observations indicate that a fish can 

 have a strong afl[inity for a specific resting spot. 

 However, I have no evidence that this is more 

 than a short-term phenomenon. Many nocturnal 

 resting spots of T. duperrey, and other fishes, 

 were discovered incidentally during other phases 

 of the work in Kona when I closely inspected 

 crevices in the coral during the night. When 

 circumstances permitted, followup observations 



719 



