Read our COVID-19 research and news.

A 1919 notebook records measurements for the salmon sampled on one day, complete with slimy scales that preserved fish DNA.

Lesley Evans Ogden

Century-old salmon-smeared notebooks reveal past bounty of fisheries

One day in June 1919, workers in a busy Canadian cannery in Port Essington rushed to clean, cook, and can the bright red flesh of a huge number of sockeye salmon hauled from the nearby Skeena River. Watching the frenzy was a government “fisheries overseer” named Robert Gibson. Periodically, Gibson selected a fish, scraped off a few scales, and affixed them to the pages of a small notebook using the salmon’s own slime. Next to each sample—he collected a total of 125 on this day—Gibson wrote the weight, length, sex, and catch date. A U.S. fish biologist hired by British Columbia would follow up by calculating each fish’s age with the then-new technique of using a microscope to count the growth rings visible on the scales, much as botanists age a tree.

Over more than 3 decades, from 1912 to 1948, Gibson and colleagues filled dozens of notebooks with fish scales from the Skeena, Canada’s second-largest salmon river. Ultimately, however, the records were dumped in a box and largely forgotten.

Until now. This week, scientists unveiled a study that makes clever use of the fish DNA preserved in the battered, smeared books to reconstruct how wild Skeena salmon have fared over the past century. The conclusion is sobering: Declines have been more precipitous and widespread than previously understood, with the river’s 13 major wild sockeye salmon populations plummeting by 56% to 99% over the period from 1913 to 2014, largely because of overfishing.

The notebooks were rediscovered 23 years ago in a Vancouver, Canada, office—but only recently analyzed. They have provided a rare “window into the past,” says biologist Michael Price, a doctoral student at Simon Fraser University in Burnaby, Canada, and the lead author of the study, published 20 August in Conservation Letters. And the notebook samples could be “a gold mine for future studies,” says ecologist Daniel Schindler of the University of Washington in Seattle, including efforts to understand why many wild Pacific salmon stocks, including the Skeena’s, have not rebounded to past levels despite new limits on fishing.

Gibson’s notebooks resurfaced in 1996 thanks to fisheries consultant Skip McKinnell, then with Canada’s Department of Fisheries and Oceans (DFO) and now based in Victoria. At the time, he was studying the growth and distribution of sockeye in the Gulf of Alaska. As part of the work, McKinnell was using average numbers derived from the Skeena logs. But he really wanted the original, more detailed data.

McKinnell recalls mentioning his quest “to any salmon biologist who would listen.” But no promising leads emerged. Then, while attending an unrelated meeting at the Vancouver offices of the Pacific Salmon Commission, a joint U.S.-Canadian body, he learned the records were sitting in a closet down the hall. Their yellowing pages, McKinnell quickly realized, had potentially preserved salmon DNA, which could enable modern molecular biologists to link the long-dead fish to current wild Skeena populations, each genetically distinct because adults breed in a complex of nursery lakes where their offspring grow for at least 1 year before migrating downstream to the sea. “In my wildest dreams,” he says, “I didn’t imagine that all of the original scales would be there, too.”

Technician Darlene Gillespie conducts an analysis of the growth rings on the scales of a fish to determine its age.

Lesley Evans Ogden

Price and his colleagues, including DFO researchers based in Nanaimo and Sidney, have now sequenced DNA from the scales of 3400 fish caught between 1913 and 1923. They were able to assign about 85% of the fish to one of the 13 major Skeena populations by comparing the old sequences to DNA from modern salmon. Then, they used the measurement and aging data derived from the notebooks, together with numbers from more recent salmon surveys and other sources, to model historical abundances.

Overall, the number of wild sockeye salmon returning to spawn in the Skeena has dropped by about 75% since the early 1900s, falling from some 1.8 million annually to 470,000, the researchers conclude. And they find that all 13 of the river’s stocks have shrunk—in contrast to previous studies, which used sockeye numbers from the 1960s as a baseline and found just seven of the runs had declined in recent decades. Several wild stocks are nearly extinct. A population that once thrived in the Motase Lake watershed has declined from about 40,000 spawning fish annually to just 600.

Price and his colleagues say the wild stocks that have fared the worst are those that once had the biggest fish—likely making them popular targets for fishing fleets, which often used gill nets that preferentially catch larger fish and let smaller ones escape. Surprisingly, the team’s analysis suggests habitat destruction, such as land clearing by logging and agriculture, did not play a major role in the declines.

Changing climate or other less obvious factors may also be influencing salmon success. Schindler notes that sockeye farther north in Alaska are thriving despite heavy fishing pressure. That suggests “the biogeographic hot spot for sockeye salmon has been moving northward,” he says, with the species struggling at the southern end of its range. 

Looking ahead, Price’s group writes that the “deep historical perspective” provided by the notebooks could help researchers identify the factors depressing the Skeena’s wild sockeye populations and better design plans for their recovery. If so, Gibson’s slimy records likely won’t be forgotten a second time.