432 



Fishery Bulletin 90(3), 1992 



photographs of 95 individuals from the western Gulf 

 of Alaska (von Ziegesar and Matkin 1989), 18 from cen- 

 tral California collected during 1977-85 (Perry et al. 

 1988), and 225 individuals from central California iden- 

 tified during 1987-88 (Calambokidis et al. 1990). This 

 comparison provided no evidence of movement by in- 

 dividual whales between these three feeding regions. 

 Two whales previously identified in both southeastern 

 Alaska and Prince William Sound (Baker et al. 1986) 

 were not reidentified in southeastern Alaska in 1986, 

 suggesting that their immigration to southeastern 

 Alaska may have been temporary. 



The identification and reidentification of individual 

 animals across years lends itself to the estimation of 

 abundance using capture-recapture formulae (e.g., 

 Hammond 1986). Table 1 summarizes abundance esti- 

 mates of the southeastern Alaska feeding herd from 

 a pair-wise comparison of all yearly samples using the 

 Petersen estimate with Bailey's correction (Caughley 

 1977). The yearly estimates range from a low of 269 

 (1983-84) to a high of 606 (1985-86). The weighted 

 mean of the Petersen estimate (i.e., the Schnabel 

 estimate; Seber 1982) across the 8-year study indicated 



Table 2 



Observed and expected frequency of yearly identifications for 

 579 adult humpback whales Megaptera novaeangliae in 

 southeastern Alaska during the years 1979-86. 



Note: Expected frequencies were calculated from the zero- 

 truncated Poisson distribution according to the methods 

 described by Caughley (1977). 



Table 3 



Calving rates of mature female humpback whales Megaptera novaeangliae in southe;istem 

 Alaska, based on reproductive histories of 41 individuals identified in two or more sum- 

 mer seasons (see Baker et al. 1987). 



Identification year 



1980 1981 1982 1983 1984 1985 1986 Sum 



Females identified 

 Total calves 

 Calves/female 



2 

 0.25 



33 

 9 



0.27 



33 

 15 



0.45 



12 



3 



0.25 



31 



15 



0.48 



21 



5 



0.24 



'Includes one calf thought to have died during the summer. See text for details. 



that this regional subpopulation has included 547 

 animals (95% CL: 504-590). 



Possible inequalities of individual reidentification 

 probabilities were examined by calculating the iden- 

 tification frequencies for individual whales across the 

 8 study years (Table 2). The observed frequency 

 distribution showed fewer 2- or 3-year reidentification 

 records and more single identifications and reidentifica- 

 tion records of extreme frequencies than expected 

 when compared with a zero-truncated Poisson distribu- 

 tion calculated according to Caughley (1977). The 

 significant departure of the observed from the expected 

 distribution (x.^ [4] 291, p<0.001) suggests that all in- 

 dividual whales were not equally available for reiden- 

 tification during the study period. Possible causes of 

 this unequal 'catchability' include births, deaths, and 

 permanent emigration across the 8-year study, as well 

 as heterogeneity of reidentification probabilities due 

 to local habitat preferences and the limited range of 

 surveys. 



Reproductive rates 



Among the 238 adults individually identified in 1986, 

 there were 32 cows accompanied by calves assumed to 

 be less than a year old. Using this census information 

 we estimated the crude birth rate in 1986 to be 0.125, 

 calculated as the total number of identified cows (n 32) 

 divided by the total number of identified whales of all 

 classes (n 257, including only identified calves). This 

 estimate, however, may have been biased by the 

 greater visibility of cow/calf pairs and by additional 

 effort directed towards individually identifying mem- 

 bers of this age/sex class. 



An alternate estimate of annual reproductive rates 

 was calculated using the identification histories of in- 

 dividual females known to be reproductively mature 

 prior to the 1986 surveys (Baker et al. 1987). Of the 

 41 mature females previously 

 identified by Baker et al. (1987), 

 24 were reidentified during the 

 1986 surveys and 9 were accom- 

 panied by a calf, yielding an esti- 

 mate of 0.375 calves/mature fe- 

 male-year ^ The addition of 

 the 1986 identifications provides 

 an updated estimate of the long- 

 term calving rates for 41 females 

 previously discussed by Baker et 

 al. (1987) (Table 3). Between 

 1980 and 1986, these 41 females 

 were observed with 58 individual 

 calves across 162 seasonal iden- 

 tifications. Although annual calv- 

 ing rates appeared to alternate 



24 

 9* 

 0.38 



162 



58 

 0.36 



