FISHERY BULLETIN: VOL. 77. NO 4 



quency of noisy and clean leaps while dolphins 

 were surface feeding (Table 4). More noisy leaps 

 were made during and after surface feeding than 

 before, and those few noisy leaps which occurred 

 before surface feeding were most often "plain" or 

 nonacrobatic. During surface feeding, plain and 

 acrobatic leaps occurred with about equal fre- 

 quency, while after surface feeding most leaps 

 were acrobatic. The noiseless or "clean" head-first 

 reentry leap and humping behavior occurred more 

 often before and during surface feeding than after 

 surface feeding. 



Our general views of behavior introduced above 

 were as follows: When not surface feeding, dol- 

 phins usually moved in small (<20 individuals) 

 groups (Table 2) with sporadic but relatively in- 

 frequent aerial behavior (Table 4). They usually 

 moved slowly (about 6 km/h) at this time. Im- 

 mediately previous to feeding bouts, they moved 

 more rapidly for short periods (mean time = 6.2 

 min,SD = 4.53),oftencreating whitewater asthey 

 surged through the water at speeds -10 km/h 

 (Table 1). We gained the impression that between 

 such surges, they stayed underwater for longer 

 periods than their normal diving times (mean dive 

 time = 21 s, Wiirsig in press), often disappearing 

 from sight for over 60 s. This pattern of movement 

 lasted from a few minutes to as long as 1 h. When it 

 stopped and dolphins were again found more often 

 near the surface, they moved slowly and stayed in 

 basically the same location. At this time, noiseless 

 (clean) and noisy (plain) leaps began (Table 4). 

 When we were near this activity in a boat, we were 

 able to see a fish school usually 2-3 m in horizontal 

 diameter, and 0.5 m vertical height, near the 

 water surface. 



Every time (65 occasions) the fish were observed 

 they were southern anchovy 6-15 cm long. When 

 we spotted the fish school near the surface, we also 

 saw terns beginning to dive for the fish, and gulls 

 coming from the vicinity. Plain noisy and clean 

 leaps continued, and acrobatic leaps began to ap- 

 pear (Table 4). During surface feeding, many of 

 the clean displays were composed of humping be- 

 havior. During humping, dolphins rapidly moved 

 singly or in pairs through the fish school after 

 coming almost vertically from deeper water, 

 caught one to five fish in their mouths, and then 

 descended again at a steep angle. Dolphins were 

 also seen near and around the feeding bout nu- 

 cleus, chasing and feeding on individual fish not 

 part of the tight school. 



If the surface feeding dolphins were not joined 

 by a nearby group or groups within several min- 

 utes, the feeding bout died down. We lost sight of 

 the fish, either because the school had been re- 

 duced by dolphins and birds or because it moved 

 away from the surface. Often, however, other dol- 

 phin groups in the vicinity converged on the feed- 

 ing bout, moving rapidly in a straight line towards 

 the feeding birds and dolphins from as far away as 

 8 km, measured by theodolite on one occasion. A 

 more usual distance was 2-3 km. As a result of this 

 movement, the feeding bout grew larger — up to an 

 estimated 300 dolphins and thousands of birds — 

 and lasted for a correspondingly longer time (Ta- 

 ble 3). Surface feeding appeared to stop when dol- 

 phins rapidly moved away from the activity, or 

 when they began deep dives and clean leaps once 

 again. In either case, we no longer saw the fish 

 school. Birds stopped flying and diving in a con- 

 centrated area and settled on the water or followed 



Table 4. — Observed incidences of aerial behavior in dusky dolphins. Numbers represent number of 15-min periods during which a 

 particular type of leap was seen. Frequency of leaping within that period has not been quantified. "Noisy leaps, in general" represent 

 leaps which made noise but were not separated into "plain" or "acrobatic" during data gathering. All significance testing used the 

 equality of percentages, arc sine transformation for the statistic tg. 



abed 

 Time of leaping 



A During the 15 mm before feeding 

 B During feeding (average time of 15 mm) 

 C During tfie 15 mm after feeding 

 D Witfiout feeding 

 Total 



Comparisons 



Amount of noisy leaps in general, a. associated witti feeding, A-C (83%) versus witfiout feeding. D (17%). significant difference, P 001 

 Amount of noisy leaps, in general, a, before feeding. A (22%) versus after feeding, C (78%); significant difference. P 0001. 

 Amount of noisy leaps before feeding. A, plain, b (100%) versus acrobatic. C (0%): significant difference, P 0001 

 Amount of noisy leaps after feeding, C, plain, b (4.8%) versus acrobatic, D (95 2%); significant difference. P 001 

 Amount of plain noisy leaps, b. before feeding. A (94.4%) versus after feeding. C (5 6%); significant difference. P 001 . 

 Amount of acrobatic noisy leaps, c, before feeding, A (0%) versus after feeding, C (100%) significant difference, P 001 

 Amount of clean leaps, d, associated v»itfi feeding, A-C (76°t,) versus witfiout feeding, D (24%); significant difference, P 001 

 Amount of clean leaps, d, before feeding, A (74%) versus after feeding, C (26%); significant difference P 01 



880 



