Matkin et aL: Association patterns of Orcinus orca 



905 



summaries of encounters indicated that designated 

 subpods traveled together as a single pod over 50% 

 of the time. 



Determination of sex and age 



Sexually mature males were differentiated from fe- 

 males and immature males by their higher dorsal- 

 fin height-to-width ratio (HWR), which typically ex- 

 ceeds 1.4 by 15 years of age (Olesiuk et al., 1990). 

 Mature females were identified when they gave birth 

 and were accompanied by a new calf The sex of most 

 juveniles could not be determined except in cases 

 where the penis or the unique pigmentation pattern 

 of the genital region was observed (Bigg et al., 1990; 

 Olesiuk et al., 1990). 



Actual ages could be determined for whales born 

 during the study on the basis of their birth year. The 

 ages of whales that were immature at the beginning 

 of the study were estimated when they were first seen 

 on the basis of relative size of the whale and size of 

 the dorsal fin (Bigg et al., 1990). The approximate 

 year of birth for whales that matured during the 

 study was estimated by subtracting the mean age of 

 maturity ( 15 years for both sexes) from the year they 

 matured. Females were considered to have matured 

 in the year they gave birth to their first viable calf, 

 and males in the year in which their dorsal fin at- 

 tained an HWR of 1.4 (Bigg et al., 1990; Olesiuk et 

 al., 1990). The year of birth of males that were sexu- 

 ally but not physically mature at the start of the 

 study was estimated by subtracting the mean age of 

 physical maturity from the year their dorsal fin at- 

 tained a HWR of 1.6-1.8, indicating physical matu- 

 rity. Males that were physically mature and had a 

 dorsal fin HWR of 1.6-1.8 at the beginning of the 

 study were considered to be at least 21 years of age 

 at that time. The minimum age of females that were 

 mature at the beginning of the study was estimated 

 by subtracting 15 years from the estimated birth year 

 of their eldest offspring. This was a minimum esti- 

 mate because a female's elder offspring may have 

 died before the start of the study. Females that had 

 not given birth for a decade or more were considered 

 postreproductive (Olesiuk et al., 1990). 



Construction of genealogical trees 



Because statistical analysis indicated that whales 

 associated in stable groups (or pods), CAI values were 

 recalculated for all individuals within each pod and 

 displayed in a matrix to show the relative strength 

 -of association among pod members (Fig. 1). We pos- 

 tulated that associations within pods reflected ge- 

 nealogical relationships and used them in conjunc- 



tion with data on known relationships, sex and age 

 data, and direct obsei-vations to infer genealogical 

 relationships (Fig. 1). 



Possible maternal genealogical trees were con- 

 structed in three steps (Bigg et al., 1990). Individu- 

 als to be incorporated into the tree as offspring were 

 selected beginning with those born during the study, 

 followed by those that were juvenile at the start of 

 the study, and finally by those that were mature at 

 the start of the study. Second, their potential moth- 

 ers were identified. All mature females in the same 

 pod were considered, providing that they could have 

 been at least 12 years (minimum age of maturity) 

 older. An individual's own mature daughters were 

 excluded as potential mothers. Third, the relative 

 strength of bonds as indicated by CAI values between 

 the individual and all its potential mothers was ex- 

 amined. The potential mother with which the indi- 

 vidual was most closely bonded was assumed to be 

 its mother. An individual not strongly bonded to any 

 potential mother was not assigned a mother. Matri- 

 ces were cross-checked to insure that mother-off- 

 spring assignments created sibling groups that dem- 

 onstrated reasonable linkage by CAI values. Direct 

 observation was used to supplement statistical analy- 

 sis to construct genealogies in instances where indi- 

 viduals and groups were less frequently photo- 

 graphed and where there was some ambiguity in the 

 numerical analysis. 



Results 



A total of 2444 hours of direct observation of whales 

 was logged from 1984 to 1995 (Table 1), during which 

 36,009 frames of film were exposed that were suit- 

 able for use in statistical analysis of association pat- 

 terns (more than one individual appeared in the pho- 

 tographic sequence). A total of 202 whales photo- 

 graphed between 1984 and 1995 were included in 

 the association analysis. According to direct field 

 observations and inspection of photographs, these 

 had previously been assigned to 9 pods (Table 2; Heise 

 et al., 1992). Another 158 whales were tentatively 

 grouped into 5 pods by field observation and visual 

 inspection of photographs (pod AX, 54 whales; pod 

 AY, 11 whales; pod AS, 17 whales; pod AF, 48 whales; 

 pod AG, 28 whales). These pods were not included in 

 the association analysis because there were insuffi- 

 cient photographs. The individuals listed in Table 2 

 represent the cumulative membership of pods over 

 all years of the study. In all pods and most groups, 

 the numbers of individuals varied over the course of 

 the study as members died or were born during the 

 study. 



