WURSIG and WURSIG: BEHAVIOR AND ECOLOGY OF THE DUSKY DOLPHIN 



Table 3. — Estimated number of dusky dolphins and birds isee 

 text) in feeding bouts of different lengths iC-test of equality of 

 means when variances are assumed to be heteroscedastic). 



'ns = not significant. -P- 05; "P 01 



cellariiformes; the giant petrel, Macronectes 

 giganteus; and black-browed albatross, Diomeda 

 melanophris. Terns were usually the first birds to 

 begin flying over and diving into feeding dolphin 

 groups. Then gulls and finally larger birds aggre- 

 gated at the feeding area. 



The most striking behavior of dusky dolphins 

 we observed was their aerial displays. To find out 

 why these displays occurred, we noted the fre- 

 quency of this activity at different times. We saw 

 all aerial displays described by Norris and Dohl 

 (1980) for the Hawaiian spinner dolphin, S^e«e//a 

 longirostris. These were leaps, head-over-tail 

 leaps, backslaps, headslaps, tailslaps, spins, and 

 noseouts. The most "acrobatic" of these leaps was 

 the head-over-tail leap. The spin was performed 

 frequently by the Hawaiian spinner dolphin, but it 

 constituted < 1% of leaps in the present study. 

 When it was seen, it was also classified as acrobat- 

 ic, along with the head-over-tail leap, for further 

 analysis. 



Two other aerial displays seen in dusky dol- 

 phins were the headfirst reentry leap and its ap- 

 parent variant, "humping." In the headfirst reen- 

 try, the dolphin leaped clear of the water and then 

 arched its back strongly while flipping the tail to 

 make a headfirst reentry. While humping, the 

 same motion occurred except that the snout and 

 tail did not leave the water during the body arch. 



Leaps, head-over-tail leaps, backslaps, 

 headslaps, tailslaps, and spins usually occurred in 

 groups. That is, one animal started a particular 

 leap, and then continued it from 3 to about 20 

 times. Because we could often not be certain that 

 the same individual was performing the leaps dur- 

 ing a leap sequence, we do not have complete 

 quantification for this phenomenon. However, in 

 45 of over 1,000 leaps it was certain that the same 

 animal leaped throughout a sequence, because we 

 followed it visually while it swam below the sur- 

 face between leaps. In all 45 instances, the leap 

 type per sequence did not change, and a mean of 

 4 (SD = 2.2) leaps per sequence was performed. 



Although the first five or so leaps were performed 

 with "exuberance," as animals leaped clear of the 

 water and reentered forcefully, successive leaps 

 were not as high, possibly as the animal tired. 



The headfirst reentry and humping did not 

 occur in sequence, but were usually performed 

 only once by an animal within about 30 s. Because 

 animals stayed underwater for long times be- 

 tween leaps, we could never be certain that the 

 same animal leaped later on. Instead, the 30-s 

 estimate was derived from counts of total leaps 

 occurring in a particular group size, and must 

 therefore be treated with caution. The headfirst 

 reentry and humping were often performed in con- 

 cert with one or two others leaping in the same 

 manner at the same time. While all other aerial 

 behaviors left individuals close to the water sur- 

 face between leaps, the headfirst reentry and 

 humping took them farther below the surface, and 

 we saw them swimming out of sight at about a 75° 

 angle. 



As was described by Norris and Dohl (1980), 

 leaps, head-over-tail leaps, backslaps, headslaps, 

 tailslaps, and spins made noise as the animals 

 reentered the water. That is, they created sharp 

 bursts of sound when the animals slapped the 

 water with their flukes or body upon reentry. Hy- 

 drophones detected the sounds underwater at ap- 

 proximately 0.5 km distance, but not at 1.0 km 

 distance during four recording sessions under op- 

 timal (no wind or waves) conditions. Norris and 

 Dohl also mentioned that sounds made by these 

 leaps attenuate relatively rapidly, but gave no 

 distance estimates. 



Headfirst reentry and humping, however, made 

 little or no sound above or below the water (hydro- 

 phones did not detect sound 10 m from the activ- 

 ity). Dolphins slid out of and into the water along 

 their longitudinal body axis during these two leap 

 types. Because we believe that the noise made by 

 most leaps may be biologically meaningful, we 

 separated them into "plain noisy" (the leap, 

 backslap, headslap, and tailslap) and "acrobatic 

 noisy" (head-over-tail and spin) leaps, and distin- 

 guished them from noiseless or "clean" leaps 

 (headfirst reentries and humping). The noseout 

 did not make noise and was often difficult to see 

 from the boat or from shore. It was therefore not 

 quantified. 



We observed dolphins in feeding bouts for 145 h, 

 and observed them during periods when we saw no 

 feeding bouts for 309 h, or over twice as long. 

 Nevertheless, we saw a significantly higher fre- 



879 



