FISHERY BULLETIN VOL 77. NO 4 



and background defeated good photography. Fi- 

 nally, the bay is relatively small, 3.2 km across its 

 mouth and indented about 2.5 km deep. Its entire 

 area was visible from the clifFtop, and usually vis- 

 ibility was good enough that one could see schools 

 well beyond its confines (Figure 1). 



These unusual circumstances allowed us to 

 gather new information about spinner dolphin be- 

 havior, especially about the diurnal cycle and pat- 

 terns of movement, though some difficult observa- 

 tional problems remain. 



Spinner dolphins proved exceptionally interest- 

 ing, observational subjects. Not only do they "spin" 

 or leap from the water and revolve rapidly around 

 their longitudinal axis, but they also perform 

 other aerial behavior that can be observed from a 

 considerable distance. These bits of aerial be- 

 havior and the sequence in which they occurred 

 proved to be a key to what one might call the 

 emotional or activity level of the school. This level 

 in turn is closely correlated with a number of fea- 

 tures of dolphin life, especially the regular se- 

 quence of activities during a daily cycle. Aerial 



Figure l. — Kealakekua Bay, Hawaii. Shown are observation 

 posts on cliif that backs the bay, shallow-water areas (in me- 

 ters), and approximate areas frequented by resting schools of 

 spinner dolphins. Also indicated are arbitrary bay sectors used 

 in analysis of arrival and departure directions. 



behavior, once understood, becomes a predicator 

 of daily activity patterns. 



This work, performed in 1970-73, represents a 

 beginning analysis of natural spinner dolphin 

 behavior, a field still in its infancy. Previous re- 

 ports of the behavior of wild dolphin schools have 

 been mostly single or very fragmentary observa- 

 tions (see Norris and Dohl in press for a review). 

 A few detailed studies exist and allow comparison 

 with this work, such as the work performed by 

 Saayman and Tayler (1971); Saayman, Bower, 

 and Tayler (1972); Tayler and Saayman (1972); 

 and Saayman, Tayler, and Bower (1973), and 

 more recently by Wiirsig and Wiirsig (1977); 

 Shane (1977); Wursig (1978); Wells et al. (in 

 press). Saayman and Tayler have analyzed the 

 daily movements of the bottlenose dolphin, Tur- 

 siops aduncus, and the Indopacific humpback 

 dolphin, Sousa teuszii, and their feeding forma- 

 tions and strategies for fish crowding and cap- 

 ture. Wiirsigs' studies concentrated on Argenti- 

 nian populations of Lagenorhynchiis obscurus 

 and T. truncatus, while Wells et al. work dealt 

 with T. truncatus in Florida. 



There are parallels in these works with the be- 

 havior patterns described here. The recent work 

 by the Wiirsigs on the group size, composition, and 

 stability of bottlenose dolphin schools bears simi- 

 larity. Like ours, much of this work was performed 

 from clifftop observation posts using natural scars 

 and marks to identify individuals. Shane's study, 

 also on bottlenose dolphins, utilized natural scars 

 and marks. Wells et al. carried out extensive tag- 

 ging studies on Florida bottlenose dolphins. All of 

 these studies and that reported here show re- 

 markable fluidity in school composition and size 

 over short periods of time. The dolphin popula- 

 tions that have been studied, it seems, are not 

 composed of discrete schools of modest size but 

 instead of highly fluid groups that may range con- 

 siderable distances and may be found associated in 

 very variable combinations of individual animals. 



Morphology 



Hawaiian spinner dolphins are moderate-sized, 

 slim-bodied, and long-beaked odontocete ceta- 

 ceans. Adults reach at least 2 m TL (total length), 

 and about 55-62 kg (Perrin 1975). They are hand- 

 somely marked animals with a dark gray cape 

 over the dorsal surface, a light gray lateral field 

 (using the terminology of Perrin 1972) sharply 

 demarcated from the cape above, and the white 



822 



