The snow crab limit is going up. There aren't more crab.

Federal regulators are being asked to raise the Bering Sea snow crab catch limit 13 percent this winter. Not because there are more crab. The increase comes entirely from cutting the safety margin that scientists subtract to account for how little they know.

The assessment's authors say so plainly: the rise "is therefore entirely a consequence of the buffer, not of a change in the stock."

That buffer is the cushion between the most crab that could theoretically be taken and the number the fleet is actually allowed to catch. The bigger the scientific uncertainty, the bigger the cushion. Regulators set it at 65 percent in 2024 and 40 percent last year. This year's assessment proposes 20 percent — the smallest in five years, and roughly half the five-year average.

Meanwhile, the crab the fleet actually wants has gotten scarcer. Biomass of preferred males — the large ones that go to market — fell 29 percent in this year's survey. Their numbers dropped 28 percent, to the fourth-lowest count in a 44-year series.

Put those together and you get the assessment's most striking line: fishing at the proposed ceiling "could remove a considerable proportion of industry-preferred males," because that ceiling amounts to 107 percent of all the preferred-male crab the survey actually found. That comparison doesn't yet account for crab that die naturally between the survey and the season, or those killed and discarded during fishing.

The fleet out of St. Paul, St. George and Dutch Harbor doesn't fish to the federal number. The state sets a lower quota beneath it — 9.3 million pounds last season, of which fishermen kept 9.11 million. So a higher federal ceiling doesn't automatically mean a bigger harvest. It raises the room the state has to work in.

Underneath all of this is a model that isn't working. The full population model regulators used until recently "does not converge reliably," and diagnostics show how shaky the fit is: when scientists restarted it from randomly varied starting points to see whether it landed in the same place, it reproduced its own result once in 78 tries. Regulators moved the stock off that model last October and onto a simpler smoothing of survey data, which can't produce a forecast at all. Asked directly whether the stock is approaching an overfished condition, the authors answer: "Unknown."

Snow crab has been under a federal rebuilding plan since it was declared overfished in 2021, after a population collapse that closed the fishery entirely for two seasons. It is not rebuilt. Current mature male biomass sits well below the rebuilding target, though above the threshold that would trigger another overfished declaration. Projections put recovery around 2029 or 2030 — unless the kind of die-offs seen in 2018 and 2019 happen again, in which case the stock doesn't rebuild within the projection window at all.

The public had little chance to weigh in on any of it. Written comment on this week's meeting closed Sept. 13, and the meeting page records no comments received. The assessment itself was posted at 9:34 the morning the meeting opened.

There is still time before anything is final. The science committee that sets the buffer meets in Anchorage Oct. 5 to 7 and takes public testimony. The Council sets final catch limits Oct. 8 to 13. Written comment closes at noon Oct. 2.