Matkin et al : Association patterns of Oranus orca 



917 



photogi'aphs of this pod, it was determined that AB 

 pod nearly always traveled as a unit; however, in the 

 dendrogram linking intrapod groups (Fig. 2) it ap- 

 peared as three separate groups (PCC>0). This find- 

 ing indicated that although they were traveling to- 

 gether, the subpods tended not to mix with the rest 

 of the pod. 



Communities were described by Bigg et al. (1990) 

 as closed populations of pods that associated with 

 one another. They described two geographically dis- 

 tinct communities of resident killer whales (north- 

 ern and southern residents) that had separation in 

 range near the middle of Vancouver Island, British 

 Columbia. We found no separation of pods into com- 

 munities by this definition in our area, although our 

 study discerned matrilineal (intrapod) groups, 

 subpods, and pods. Resident whales from AF and AG 

 pods photographed regularly in southeastern Alaska 

 were observed swimming with the pods described in 

 this study (Matkin et al., 1997). One of the pods de- 

 scribed in this paper, AD5 pod, was photographed 

 off Kodiak Island. It thus appears that geographic 

 boundaries do not delineate communities for resident 

 killer whales from southeastern Alaska to Kodiak 

 Island. 



The association analysis for individuals within 

 pods strongly suggested that strength of bond was 

 directly correlated to degree of relatedness. There 

 was only one case in which statistical analysis indi- 

 cated that offspring born during the study did not 

 maintain the strongest bond with its mother. The 

 mother AB8 and her sibling AB18 both died at the 

 time of the Exxon Valdez oil spill. AB8 left her year- 

 old offspring. AB41 (born 1988). Association analy- 

 sis indicated that AB41 was more closely linked 

 (CAI=58) to its mother's apparent sibling, AB18, than 

 to its mother, AB8 (CAI=50); the young whale AB41 

 died several years later ( 1993-4). 



Five whales that were juveniles at the beginning 

 of the study were most closely associated with whales 

 other than their mothers. Three were females, AK07, 

 AN08, and ANll, each of which produced offspring 

 during the study. Following these births, they were 

 more closely bonded to their offspring rather than to 

 their own mothers. This finding demonstrates that 

 new mothers, when they produce calves, develop dis- 

 tance from their own mothers (Table 4). The tendency 

 of females with offspring to travel farther from their 

 mothers than the distance prior to first reproduc- 

 tion suggests a process of separation that may also 

 be basic to new pod formation. 



Our proposed genealogical trees suggest that 

 jntrapod groups are matrilineal groupings of moth- 

 ers and their descendants (Figs. 1 and 3-12). There 

 was no immigration or emigration of male or female 



offspring from these natal groups over the course of 

 our study. These extremely stable matrilineal groups 

 appear to be the foundation of resident pod social 

 structure in Prince William Sound. This is similar to 

 results from studies in British Columbia and Wash- 

 ington State (Bigg et al., 1990). 



We were most confident in genealogical trees for 

 pods that were most frequently photographed, such 

 as AE and AK pods, and less confident in trees for 

 the much less frequently observed ADS and AD16 

 pods. The large number of mortalities in AB pod also 

 made construction of genealogical trees more diffi- 

 cult. The greatest potential source of error in genea- 

 logical assignment was the death of the mother of a 

 young whale prior to the study, in which case the 

 young whale would likely travel with its closest fe- 

 male relative. However, since the annual natural 

 mortality rate for reproductive females is extremely 

 low (0.0048 according to Olesiuk et al., 1990), this 

 source of error was probably insignificant except fol- 

 lowing disasters such as the high loss of reproduc- 

 tive females after the Exxon Valdez Oil Spill (Matkin 

 etal., 1994). 



In all but one of the resident pods we examined, 

 the total number of whales increased over the pe- 

 riod of the study, indicating that a majority of matri- 

 lineal groups have been growing or dividing (or do- 

 ing both) over the past decades. The exception was 

 AB pod, which declined during this period from 35 

 to 23 whales (28 deaths; 16 births). Six of the mor- 

 talities occurred during 1985 and 1986 when there 

 was a conflict with the sablefish (Anaplopoma fim- 

 bria) fishery (Matkin et al., 1994). During this pe- 

 riod apparent bullet wounds were observed in 16 

 whales. 



Fourteen of the mortalities in AB pod occurred in 

 the year and a half following the 1989 Exxon Valdez 

 oil spill (Matkin et al., 1994). Some of the matrilin- 

 eal groups in the pod are nearly extinct as a result of 

 these deaths. An adult male, AB03, appears to be 

 the last member of a once-large matrilineal group 

 linked by the apparent sisters AB06 and AB07 (Fig. 

 1; AB03 has since died). Another large matrilineal 

 group ( matriarch AB09) has been reduced to a single 

 orphaned 5-year-old, AB45. Many of the mortalities 

 have been those of juveniles (13) or reproductive fe- 

 males (4) and have severely reduced the reproduc- 

 tive potential of the pod. 



The statistical support for the existence of intrapod 

 groups and the apparent lack of emigration or immi- 

 gration into these groups allows for determination 

 of demographic changes, both natural and induced, 

 within resident killer whale populations. The de- 

 tailed delineation of social structure in resident killer 

 whales provides a unique opportunity for monitor- 



