The National Marine Fisheries Service (NMFS) has come up with a "plan" for the recovery of southern resident killer whales — a.k.a. Puget Sound orcas — but the agency doesn't really know what the problems are, and it doesn't really know how to solve them. Is there a kids' joke that begins, "When is a plan not a plan?"

Yes, it's "killer" whales. The feds use the K word, even though "killer whales" as a term has been largely abandoned out of political correctness. You rarely hear that phrase these days. We know that people started thinking differently about the animals when they got acquainted with captive Namus and Shamus and their relatives in the 1960s and 70s, and that the whole species got a big public relations boost from the "Free Willy" films in the 1990s. We know that the animals aren't really whales; they're big dolphins. But when did everyone start saying "orca"?

The plan released last month calls for spending nearly $50 million over the next 28 years, largely for research, and for improving a variety of the orcas' living conditions, largely by doing what is already being done. No one seems very enthusiastic about it, but no one seems very upset, either. Presumably, people's expectations were low.

The fact is, no one really knows exactly why the southern resident killer whale population has declined to its current level of 87 animals — although it seems safe to say that when your favorite food — in this case, Puget Sound chinook salmon — makes the endangered species list, you yourself won't be far behind.

Beyond declining food stocks, pinpointing a cause for orcas' decline is difficult. Orcas have high levels of PCBs, PBDEs and other toxic chemicals in their bodies. In laboratory tests of rats, very high levels of toxic chemicals have suppressed the animals' immune systems. Chronic exposure to toxic chemicals at lower levels may suppress the killer whales' immune systems, making them vulnerable to disease. However, transient killer whales, which eat seals rather than salmon, show higher levels of chemicals but no population decline. Toxic chemicals in body fat can't be good for the southern resident killer whales, but no one knows what the actual effect has been. Ditto underwater noises from sonar, boat engines, etc., which may interfere with communication and echolocation, and increase the level of stress. The same goes for harassment by whale watching boats.

The 2005 endangered species listing suggested — plausibly — that the population might still be feeling the effects of the pursuit, capture and incidental killing of orcas in the 1960s and 70s — before it was illegal — by people acquiring them for aquariums. The whale catchers took half the population. And it's not just a question of raw numbers. Long-lived animals with lives shaped by what many people consider a culture might in fact have felt the effects of the captures and incidental killings for a long time. But the whales have probably gotten over it by now, suggests Ken Balcomb of the Center for Whale Research, much as Japan's society and our own have largely recovered from World War II.

The numbers certainly suggest that whatever the role of chemicals and noise, the food supply is crucial. Puget Sound's orca population was very low when people started counting in the mid-1970s, rebounded considerably over the next 20 years, but then plummeted 20 percent between 1996 and 2001, which prompted Canada, the state of Washington, and — reluctantly — the United States to list it as endangered. The population drop in the late 1990s coincided with a bad time for chinook salmon.

In Puget Sound, samples taken of fish scales that stick to the orcas after feeding and scat they leave behind suggest that 80 percent of their diet is chinook. Some report seeing them go after chinook further south, but there's no real scientific evidence. No one really knows what else they eat, or whether their diet varies seasonally. Balcomb notes that if you capture a killer whale and put it in SeaWorld, it will learn to eat ground-up squid, mackerel, and whatever else it's given. So they are adaptable. On the other hand, whale survival rates do seem to track coastwise rises and falls of chinook populations. In bad salmon years, young killer whales are the members of the population that don't survive. Balcomb explains that it takes a lot more energy to raise a young whale than to give birth, and if the food supply isn't there, parents just won't have that much energy to spare.

Days after the NMFS recovery plan came out, southern resident orcas got some good Seattle press when L pod was photographed in California, off Monterrey Bay. Given that their known range extends from the Queen Charlottes to Monterrey Bay, their presence off the Golden State shouldn't have surprised anyone. Balcomb says that pods from the southern resident population have turned up off Monterrey Bay for the past eight years. But taking California vacations may be a new habit. In the 1970s and 80s, researchers found at least some of the whales around the San Juans in all twelve months of the year. The whales may have been taking winter trips to California all that time. Moving at 75 miles a day, they could make the round trip in a little over two weeks, so they could conceivably spend part of each month in California and still get back in time to be counted in the San Juans, however unlikely that seems.

At any rate, the appearance of L pod off California should have made clear that saving Puget Sound's "resident" orcas is not just a Puget Sound problem. It's all well and good to talk about preserving the entire ecosystem, but what, exactly, is that? How do you define it? For the orcas, it includes roughly 1,000 miles of coast. Fish that spawn in and pollution that flows from not only the Sound and its tributaries, but the rivers of British Columbia's Inside Passage, the Columbia, the Rogue, the Klamath, and the Sacramento may all have impact on orcas' survival.

The wide scope of the problem is not encouraging. Right around the time L pod arrived off California, Pacific Fishery Management Council Director Donald Mc Isaac announced that California's Central Valley fall chinook, which spawn primarily in the Sacramento River system, had hit a new population low. The number of returning fish plunged to 90,000 last year, only about one-ninth of the figure five years ago. The federal government already considers them a species of concern — which at least means they've been doing better than the Central Valley spring run chinook (threatened) or the Sacramento River winter run chinook (endangered), which were the very first salmon populations on the federal endangered species list.

Chinook from the Klamath River haven't been doing so well, either. Coastal chinook fishing was curtailed two years ago because of low Klamath River returns. Six years ago, in the fall of 2002, as salmon waited until water in the Klamath River got high enough for them to swim upstream, more than 30,000 chinook salmon died — an unprecedented event on the Klamath — along with more than 300 federally protected coho. The state of California blamed the fish kill on crowding and high temperatures that would not have occurred if the feds had not withdrawn so much water from the river for irrigation.