900 



Abstract.— The social structure and 

 genealogical relationships of resident 

 killer whale pods in Prince William 

 Sound, Alaska, are inferred from asso- 

 ciation analysis and direct observation. 

 During 1984-95 a total of 2444 hours 

 of observation were made and 36,009 

 photographs were taken of identifiable 

 killer whales. Cole's association index 

 and a point correlation coefficient in- 

 dex were used to test the statistical sig- 

 nificance and strength of associations 

 between individuals, and a clustering 

 procedure was used to delineate group 

 structure. A total of 202 whales were 

 grouped into 9 pods. Genealogical rela- 

 tionships were inferred from the 

 strength of bonds among pod members. 

 Genealogical trees suggested that 

 intrapod groups were matrilineal in 

 structure. Splitting of one pod I AN podi 

 was observed during the study; how- 

 ever, there was no splitting of matri- 

 lineal groups. 



Association patterns and inferred genealogies 

 of resident l<iller whales, Orcinus orca, 

 in Prince William Sound, Alaska 



Craig O. Matkin 



North Gulf Oceanic Society 

 PO Box 15244 

 Homer, Alaska 99603 

 E-mail address comatkin'S xyz net 



Graeme Ellis 



Peter Olesiuk 



Department of Fisheries and Oceans 



Pacific Biological Station 



Nanaimo, British Columbia, Canada V9R 5K6 



Eva Saulitis 



North Gulf Oceanic Society 

 PO Box 15244 

 Homer, Alaska 99603 



Manuscript accepted 14 December 1998. 

 Fish. Bull. 97: 900-919 (1999). 



It is well established that individual 

 killer whales can be recognized by 

 unique marks, scars, and pigmen- 

 tation patterns (Bigg et al., 1987; 

 Heise et al., 1992). During the past 

 25 years a number of studies have 

 used this fact to investigate the life 

 history, behavior, population dy- 

 namics, vocalizations, abundance, 

 and movements of killer whales in 

 coastal waters of the eastern North 

 Pacific including Prince William 

 Sound, Alaska (Balcomb et al., 1982; 

 Bigg et al., 1990; Olesiuk et al., 

 1990; Ford, 1991; Matkin et al., 

 1994). As a result of these studies, 

 at least two forms of killer whale, 

 termed "resident" (fish-eating) and 

 "transient" (mammal-eating), have 

 been described in the nearshore 

 waters of Puget Sound, British Co- 

 lumbia, and southern Alaska ( Bigg, 

 1982; Leatherwood et al., 1990; 

 Morton, 1990; Ford et al., 1998; 

 Saulitis et al., in press). Although 

 resident and transient killer whales 

 occur sympatrically, they are con- 

 sidered separate populations owing 



to genetic variation in mtDNA hap- 

 lotypes (Stevens et al., 1989; Hoezel 

 and Dover, 1991 ) and microsatellite 

 variation in nuclear DNA (Hoezel et 

 al., 1998). Mating between the two 

 forms has not been observed and 

 resident and transient killer whales 

 have not been seen traveling to- 

 gether (Morton, 1990; Saulitis, 

 1993; Matkin et al., 1994). Studies 

 in British Columbia and Washing- 

 ton indicate that resident whales of 

 both sexes remain in their natal 

 pods for life, whereas transients 

 appear to have a more fluid social 

 structure (Bigg et al., 1990). 



The social organization of resi- 

 dent killer whales in Prince William 

 Sound has been previously de- 

 scribed from direct observation of 

 associations among individuals 

 (Leatherwood et al., 1984; Heise et 

 al., 1992). These studies name pods 

 on the basis of observation of groups 

 of individuals that repeatedly trav- 

 eled together. Each identifiable 

 whale is assigned an alphanumeric 

 code (Leatherwood et al., 1984; 



