A (killer) whale of a mystery.

In just three years, 14 killer whales vanished from a hand of 36 summer "residents" -- known as the AB pod -- of Alaska's Prince William Sound. Biologists assume that a resident gone from its pod a year or more is dead, notes Craig O. Matkin of the North Gulf Oceanic Society in Homer, Alaska. Because a pod this size typically loses only one or two members annually, the size of AB's losses "is absolutely unprecedented," notes Marilyn E. Dahlheim of the National Marine Fisheries Service in Seattle. It's also highly suspicious, the pair note. Matkin has observed the AB pod since 1984 and says the first sign something had gone wrong-- seven missing whales -- occurred just six days after the Exxon Valdez spill.

Biologists have distinguished two characteristic types of killer whales -- transients and residents. Although transients sport a somewhat wider, more pointed dorsal fin than residents, the distinction rests primarily on social structure. Resident killer whales are unique among cetaceans in forming lifelong communities containing one or more mothers and their offspring, Matkin notes. The unusual stability of resident pods permits fairly accurate lifespan estimates for these longlived, long-distance travelers.

Though pods as large as AB occasionally divide into smaller units, those splits invariably occur along matrilineal lines, not within a family unit, Dahlheim points out. The 13 whales apparently lost over 1989 and 1990 came from six of the pod% seven maternal families. Since there is virtually no emigration from resident pods, and since Matkin's team has had so many encounters with AB that they can recognize individual whales by their unique dorsal-fin patterns, "we are very confident that the animals missing represent mortalities," he says.

Though the team spotted six killer whale pods in the spill during 1989, they observed unusual losses only in AB. However, Dahlheim says, AB appears to have come into contact with the oil earliest, when some of its more volatile - and toxic constituents may still have been present. As whales usually sink when they die, the biologists have no bodies to autopsy for confirmation of oil-related syndromes. Nevertheless, Dahlheim speculates that the 1989 losses may have resulted from direct inhalation or skin exposure to oil, and the following year% disappearances from effects of chronic exposure. However these deaths occurred, Dahlheim says, "the whole social structure of this pod has been totally disrupted."