590 



Fishery Bulletin 90(3). 1992 



the animal spent at the surface ("surface interval"), 

 and (4) time spent below the surface during each dive 

 ("dive time"), defined as the period of submergence 

 following typical sounding behavior (i.e., a prominent, 

 high arching of the back and tail, often followed by 

 bringing the tail flukes above the water surface). The 

 surface interval during which the biopsy strike took 

 place was excluded from our analyses of both the pre- 

 and postbiopsy respiratory variables. We determined 

 (using chi-square goodness-of-fit tests) that the distribu- 

 tion of the observed data was not significantly different 

 from a normal distribution, thus differences between 

 the pre- and postbiopsy values were compared using 

 two-tailed paired (-tests (Zar 1984). Surface-interval 

 to dive-time ratios were calculated for each whale 

 during pre-biopsy and postbiopsy focal samples, and 

 compared using a Wilcoxon sign rank test (Zar 1984). 

 The surface-interval/dive-time ratio integrates several 

 respiratory values into a single measure of the res- 

 piratory "strategy" for each individual. 



LORAN-C positions, which have an error of ~30m 

 in the study area (Day 1983), were used in estimating 

 the net rate of movement of a whale as defined by its 

 surfacings. LORAN positions were recorded at the 

 start and end of each focal sample. With each LORAN- 

 C reading, the bearing to the whale to the nearest 5 

 degrees and the visually-estimated distance from the 

 vessel to the whale were also recorded. This informa- 

 tion was used to correct the LORAN-C data if the 

 whale was >30m from the vessel, which was likely 

 given the limitations of vessel movement during focal 

 samples described above. To estimate the net move- 

 ment rate, the distance between the first and last cal- 

 culated whale positions within each focal sample period 

 was divided by the elapsed time (30 minutes), yielding 

 a net movement rate in knots. Actual swimming speed 

 could not be determined due to uncertainty about the 

 direction of a whale's movement underwater or the 

 linearity of its track. Results between pre- and post- 

 biopsy periods were compared using a Wilcoxon sign 

 rank test (Zar 1984). 



Photographs of the dorsal fin and tail flukes of in- 

 dividual whales were taken upon approach for biopsy. 

 Each whale was identified using the catalog of Gulf of 

 Maine humpback whales kept at the Cetacean Research 

 Unit, where individuals are assigned a two-letter, one- 

 number file code. If the animal could not be identified 

 in the catalog, it was assigned a three-digit code. When 

 possible, each animal was assigned to one of the follow- 

 ing age groups: juvenile (1-3 yr), adolescent (4-6 yr), 

 or adult (>6 yr). Age classifications were based on 

 previous and/or subsequent repeated annual observa- 

 tions of the same individuals by the authors from calf 

 year (used for juveniles and adolescents), sightings of 

 the individual as an initially small and subsequently 



larger animal (juveniles and adolescents), or annually 

 repeated sightings of an individual with no appreciable 

 growth over several years (adults). If an animal was 

 sighted only during the year in which it was biopsied, 

 it was not classified by age-class. 



During focal samples collected in 1985, a total of 30 

 behavior types were observed and analyzed. Behavior 

 types were defined using an ethogram for humpback 

 whales developed by the Cetacean Research Unit prior 

 to this study (unpubl. data). The probability that any 

 given behavior was displayed by more or fewer animals 

 in the pre- vs. postbiopsy focal samples was tested 

 using the binomial distribution (with the probability of 

 each period containing an occurrence of the behavior 

 assumed as 0.5), while the change in frequency of each 

 behavior in individuals, given that the behavior was 

 observed at all, was compared using Wilcoxon signed 

 rank tests (Zar 1984). All 30 behaviors were tested for 

 variation. Many of the behaviors did not show any vari- 

 ation between control and response periods, and there- 

 fore are not described in detail. These were belly-up 

 rolls, breaches, bubble clouds (bubble clouds followed 

 by obvious surface feeding), bubble cloud behaviors 

 (bubble clouds not followed by obvious surface feeding), 

 defecations, flipper flares, flipper flicks, flipper in air, 

 flukes, half flukes, hangs, high flukes, high head-ups 

 ("spyhops"), lobtails, logging, low flukes, quarter rolls, 

 single bubbles, snakes (a twisting of the body), surges, 

 tail breaches, and passing under a boat. Definitions of 

 those behaviors which either varied significantly in 

 frequency or showed some notable variation in the fre- 

 quency of display following the biopsy procedure are 

 the following: 



Back rise Animal breaks surface while swimming, 

 with no accompanying exhalation. 



Belly-up lobtail Animal, ventral side up, elevates tail 

 into the air, then slaps the water surface with the 

 dorsal surface of its flukes. 



Hard tail flick Animal rapidly and forcefully flexes 

 tail up and down one time during otherwise normal 

 swimming behavior; much spray can be thrown; 

 flukes clear surface. The hard tail flick is faster and 

 presents a less regular arching movement of the tail 

 than a lob-tail. 



Low head-up Animal lifts head into air at 30-45° 

 angle to surface. 



Sounding dive Animal arches its back in a typical div- 

 ing posture but does not bring its tail flukes above 

 the surface. 



Tail rise Animal slowly straightens its caudal pedun- 

 cle at the surface during normal swimming. 



Tail slash Animal moves tail forcefully from side to 

 side, flukes at or just below the surface; creates white 

 water frothing. 



