The happy, healthy dog suddenly stopped eating and drinking, and she became lethargic. A trip to see Dr. Taylor Douglas of Bend and some blood tests revealed nothing, Brooks said.

Back home, Abby's condition continued to decline, and a second trip to the vet's office resulted in exploratory surgery, where doctors found swelling, but little else that would lead them to a diagnosis, she said.

"It was kind of funny, because I was like, 'Well, she needs to get better because I'm leaving for Alaska for a week on this vacation,'" said Brooks, a physical therapist. "And (Dr. Douglas) asked me, 'What are you going to Alaska for?' And I said, 'I'm going fishing.'"

It was that information that served as a turning point in Abby's case, Douglas said.

"He looked at me and he said, 'Do you fish for salmon?'" Brooks said. "And I told him that, yeah, I steelhead fish all the time. And he said, 'Has Abby ever gotten into any salmon?'"

She had. Two weeks earlier, Brooks and Abby were on a fishing trip to the Umpqua River system in southern Oregon when Brooks noticed her pup gnawing on a fish head.

"I took it away from her, threw it in the river and didn't think anything of it," she said.

After hearing that story, though, Douglas had a pretty good idea of the dog's affliction: salmon poisoning, a condition that kills about 90 percent of affected dogs if left untreated.

At the time, Abby was 12 days removed from eating the fish, and her life was in danger. She stayed at the hospital for a week, receiving fluids, antibiotics to fight the bacteria, deworming treatment to battle the parasite, and plasma because she had lost so much blood.

She couldn't walk or eat, so Brooks and vet staff force-fed her. Eventually, Brooks took her home.

"We just didn't think she was going to make it," Brooks said. "I was like, 'Well, if she's going to die, I'm going to take her home.'"

Here's the good news: Abby did live. She lost a lot of weight, but eventually she began eating again and recovered.

But here's the bad news: Salmon poisoning is specific to one region, and that region happens to be west of the Cascade Range in Oregon, Washington, California and British Columbia, said Colin Gillin, veterinarian for the Oregon Department of Fish and Wildlife.

Gillin, based in Corvallis, said reports of salmon poisoning are infrequent. Its origin is in a bacterial organism called neorickettsia, which is carried by a specific parasite called a fluke that thrives in the waterways of the western Pacific Northwest.

The story of how the bacteria infects a dog plays out like Russian nesting dolls.

"There's four players in this besides the dog," Gillin said. "There's the fish, and then there's a snail, and a fluke. And the fluke infects the snail, and the fish eats the snail and gets the fluke. And inside the fluke's some bacteria, and the bacteria is what messes up the dog."

Only canids - dogs, wolves, foxes, coyotes - are affected by the bacteria, Gillin said. And only salmon-type fish such as steelhead and trout - anadromous fish that swim upstream to breed - seem to provide a suitable home for the fluke.

Only certain species of aquatic snail can carry the fluke to the fish, he said. And the fluke is the most specific of all the hosts; only one genus and species of the fluke carries the bacteria.

If a dog eats a fish that contains the parasite, the fluke will release eggs that enter the animal's intestinal tract, where they release the neorickettsia. Dogs that are infected will usually begin to show symptoms within one to three weeks, and symptoms typically include vomiting, diarrhea, swollen lymph nodes and fever.

Owners whose dogs are showing those kinds of symptoms should get their animal to a vet immediately, regardless of the reason for the illness, Gillin said. An owner with an inkling that a dog has eaten uncooked salmon, either from a garbage can, a kitchen counter or a campsite, should inform the vet of that possibility, he said.

"If it's got any kind of gastrointestinal symptoms," Gillin said, "it's in a losing battle with a bacteria."

What made Abby the golden retriever a tricky case was that she wasn't showing symptoms typical of salmon poisoning, Douglas said.

"Abby wasn't running a fever, wasn't having diarrhea and didn't have enlarged lymph nodes, peripherally. So the signs weren't there. We didn't have the classic presentation," he said. "The diagnostics didn't really show the typical presentation. We didn't see the things that we would expect to see. And then once we saw something else, we said, 'Maybe this is it.'"

"What makes it more confusing is when they don't show those clinical signs or you don't pick them up on the test," he said. "Some of these diseases don't always read the book."

It's helpful for veterinarians to know if a dog has eaten or been around fish to help narrow possible causes of illness, Douglas said. People cleaning fish should take care to ensure their dogs can't get to them.

"In the Northwest, it is an issue," he said. "Even smoked fish - not cooked, but raw, smoked fish - could potentially transmit this, too."

ODFW used to include literature about salmon poisoning with its fishing regulations, said Gillin, although that material has been removed this year. Department spokesman Jessica Sall said she doesn't know of any educational efforts to warn anglers about salmon poisoning, especially in Central Oregon.

Brooks said many fishermen know about the issue, but she didn't when she discovered Abby gnawing on that fish head. And she wants to try to make sure others don't run into the same problem she had with Abby.

"If I would've known what it was, I would've been like, 'Oh, she got into this salmon,' and they would've put her on an antibiotic and that would've been the end of it," Brooks said. "Instead, she almost died, and I spent all this money and she had to have surgery.

"It was an awful thing to go through," she said. "I don't want anyone else to go through it."