FISHERY BULLETIN: VOL. 86, NO I 



Northern right whales feed by both surface- 

 skimming and subsurface feeding, apparently 

 holding the jaws agape below the surface much 

 like they do when at the surface (Watkins and 

 Schevill 1976, 1979). Surface-skimming by north- 

 ern right whales that feed on copepods and juve- 

 nile euphausiids occurs at speeds of about 3 knots 

 with relatively little turbulence (Watkins and 

 Schevill 1976, 1979). The one southern right 

 whale which we observed surface-skimming, and 

 which was also mentioned in Goodall and 

 Galeazzi (1986), swam at about 8 knots, with an 

 immense amount of turbulence. This whale un- 

 doubtedly was feeding on Euphausia superba. 

 The extremely high speed of this whale when 

 surface-skimming may be related to type of prey. 

 Euphausia superba is a large euphausiid that 

 swims up to 30 cm/second, and it is quite adept at 

 avoiding scuba divers and other dark objects 

 (Hamner et al. 1983; Hamner 1984). It may be 

 that right whales adjust the speed of their surface 

 feeding runs to overcome the swimming speed of 

 the particular prey that they pursue. Fin whales 

 also exhibit differences in breathing and diving 

 rhythms when feeding on euphausiids and on 

 schooled fish (Watkins et al. 1984). 



The southern right for which we have a 3.5-h 

 record in 1986 fed exclusively below the surface. 

 It swam for about an hour at the surface almost 

 directly NNW toward the only patch of krill that 

 we had located previously via the echo sounder 

 during repeated transects across the mouth of 

 Andvord Bay. It is not understood how baleen 

 whales locate prey. They might echolocate, they 

 might listen for krill noises, or they might re- 

 member where they had prior success and return 

 to the general vicinity and then begin to hunt 

 randomly. Although our data do not allow us to 

 choose among these possibilities, it appears that 

 this particular southern right whale did navigate 

 directly to patches of prey, as suggested by 

 Watkins and Schevill (1979) for northern right 

 whales. However, some subsurface searching 

 probably occurs as well. The whale that we fol- 

 lowed for the longest period of time made a rea- 

 sonably long dive of 210 seconds in the middle of 

 its surface swim toward the krill patch. It appar- 

 ently did not find anything to eat (nor did the 

 sonar indicate the presence of krill), and it made 

 no more significant dives for about 30 minutes. 



When the whale reached an area with small 

 scattered krill schools, its dives were generally 

 short, and the temporal spacing and duration of 

 dives and surface intervals were erratic. When 



the whale reached a large school of krill it stopped 

 swimming between dives, dives lasted longer, 

 surface intervals between dives became more reg- 

 ular, and successive dives were progressively 

 shorter. After 10 of these dives the whale floated 

 without any swimming movements at the surface 

 for about 12 minutes while occasionally rattling 

 its baleen and slightly lifting its upper jaw. 



The whale clearly anticipated its dives. It hy- 

 perventilated for about 90 seconds prior to all of 

 the long dives. Hyperventilation is a common pre- 

 cursor to diving in air-breathing vertebrates 

 (Kooyman et al. 1981), but this respiratory pat- 

 tern apparently has not been reported previously 

 for a baleen whale. 



The data that we present in support of this feed- 

 ing pattern are admittedly modest, consisting of 

 only one continuous 3.5-h data sequence for one 

 individual. Nonetheless, this data set is truly 

 unique and it includes a series of statistically sig- 

 nificant changes of behavior at the surface and at 

 depth which are correlated with prey distribu- 

 tion. Since the changes of behavior are real, they 

 merit interpretation. One hypothesis to account 

 for the increase and then the decrease in dive 

 length after encountering prey is that the whale 

 dove to its full capacity in order to maximally 

 exploit an opportunity to feed. With repeated 

 dives its dive capacity may have diminished, pro- 

 ducing first long, then shorter and shorter dives 

 (Fig. 1). In contrast, this breathing pattern has 

 not been observed in grey whales, which exhibit a 

 sustained capacity for repetitive, nearly 

 monotonic diving while feeding (Obst et al. in 

 prep.). It is possible that right whales, being more 

 specialized for surface-skimming, become ex- 

 hausted more readily during sustained diving. 



An alternative hypothesis to account for this 

 pattern of behavior is that the whale captured 

 krill with such ease on the first long dive of 348 

 seconds that a "digestive bottleneck" (Kenward 

 and Sibly 1977) developed; subsequent dives 

 would be progressively shorter because the stom- 

 ach (or some storage chamber) filled more rapidly 

 than the food could be processed. The long surface 

 interval of immobility followed by a second feed- 

 ing bout might then represent the time taken to 

 clear such a chamber (e.g., the first of the three 

 stomach pouches) in preparation for the next bout 

 of feeding. If this hypothesis is correct, it implies 

 that once a right whale has found a particularly 

 favorable school of krill, it fills up very rapidly. 

 This interpretation contrasts markedly with a 

 popular impression of great whales as continuous 



148 



