FISHERY BULLETIN: VOL. 70, NO. 3 



some of the transition activities are performed 

 by the smaller fishes first. Thus, for example, 

 the increased tendency of many fishes to aggre- 

 gate, as seen especially in acanthurids and scar- 

 ids, occurs first in the smallest species of the re- 

 spective groups: Acanthunis nigrofuscus 

 among the acanthurids, and Scams taeniurus 

 among the scarids. Consider too the plankton- 

 feeding damselfishes, which during the transi- 

 tion period progressively descend closer to the 

 sea floor and finally take cover on the reef. The 

 smallest species, Chromis vanderbilti, is first to 

 seek cover, and the next smallest, C. leuciirus, 

 is second. The other, larger damselfish species 

 follow later. Similarly, the smallest of the sur- 

 geonfishes, Acanthunis nigrofuscus, is the first 

 of that family to go under cover, and the others 

 follow in order roughly corresponding to in- 

 creasing size. Most acanthurids on the reef are 

 of about the same size and therefore show si- 

 milar timing in seeking cover ; nevertheless, the 

 larger species, like A. dussumieri, are the last 

 to settle down. The same pattern occurs among 

 the filefishes, where the smaller Cantherines 

 sandivichiensis takes cover long before the lar- 

 ger C. dumerili; and among the parrotfishes, 

 where the smallest, -Scarws taeniurus, is no longer 

 seen when such larger relatives as S. rubrovio- 

 laceu^ and S. perspicillatus are still active. The 

 effect operates within species, just as it does be- 

 tween species; thus, for example, large adults 

 of Scarus ruhroviolaceiis are still active long 

 after all the small juveniles of this species have 

 gone under cover. 



In the morning the situation in these same 

 species is reversed. The larger individuals are 

 the first of the diurnal fishes to emerge from 

 cover, and the others appear in order roughly 

 according to decreasing size to perform their 

 characteristic transition activity. 



It remains uncertain to what extent fish size 

 relates to the transition activity of nocturnal 

 fishes. Holocentrus sammara, Apogon menese- 

 mus, and A. snyderi, which are consistently the 

 first nocturnal species to leave shelter in the 

 evening and the last to seek shelter in the morn- 

 ing, are the largest of their respective groups 

 common in Kona; however, additional data are 

 lacking. 



Increased Activity During Twilight 



One senses a marked increase in activity 

 among fishes at two times during the twilight 

 periods: in the evening, from just before sun- 

 set to about the time that the diurnal fishes 

 abandon the water column, and in the morning, 

 from the time the diurnal fishes reoccupy the 

 water column to just after sunrise. This im- 

 pression of heightened activity derives mostly 

 from a combination of the following: 1) fishes 

 migrating from one location to another; 2) in- 

 creased territorial aggression (evening only) ; 

 and 3) an increased tendency to aggregate in lo- 

 cations more visible to the observer. 



Migrations. — The migrations are discussed 

 above. Their contribution to a general impres- 

 sion of increased activity is obvious. 



Territorial aggression. — A minor element of 

 the increased activity during evening twilight 

 is increased territorial aggression. During the 

 day there are frequent territorial disputes 

 among many of these fishes, often, if not gen- 

 erally, relating to feeding areas; this is a fre- 

 quent source of conflict among the acanthurids 

 (see, for example, Jones, 1968). Aggression 

 during evening twilight, which was noted only 

 in diurnal fishes, seems unrelated to feeding; 

 rather, it may express conflict over resting spots. 

 It is significant not so much that such aggression 

 occurs, but that there is so little of it. With 

 such a vast number of fishes settling into resting 

 positions on the reef at nightfall, one might ex- 

 pect individuals with similar requirements to 

 compete strongly for optimum spots. That rel- 

 atively little aggression of this sort occurs sug- 

 gests that some mechanism establishes spatial 

 distribution without overt conflict. One possi- 

 bility is that each individual has a well-estab- 

 lished resting spot; this is suggested by the two 

 individual Thalassoma duperrey that were ob- 

 served to be strongly attached to particular lo- 

 cations. However, this hypothesis is not widely 

 supported. Overall, I recognized only a rela- 

 tively few cases where what seemed to be the 

 same individual returned to a particuar spot on 



734 



