Fish mean body mass did not decrease in warmer years in 13 trawl survey units, spatial slopes were weak, and steps at 1985 and 2009 gear changes made most of the US Northeast shelf decrease

Question: do wild demersal fish follow the temperature-size rule (TSR, about -2.5 to -5% body mass per degC in laboratory studies) in space, from year to year, and over decades? Data: FishGlob v2.1 for 12 survey units and NOAA AFSC tables for the eastern Bering Sea; 13 units with in-situ bottom temperature, 624 species-by-unit populations of 296 species, 1963-2026. Method: the response is the catch mean individual mass (summed weight / summed number) per 0.5-degree cell and year. A two-way (cell, year) model gives cell and year effects. We regressed these on cell temperature climatology (with depth) and on the annual temperature anomaly (with a spline trend), and calculated decadal trends. We combined units with random-effects means and used a year-permutation null. The plan was written before the population results; four amendments are listed. Results: the spatial slope was -1.4% per degC (95% CI -5.4 to +2.9) and depended on depth. The year-to-year slope was +0.4% per degC (95% CI -0.8 to +1.5). In 14 variants with lags of 0-2 years, no lower 95% limit was below -2.5% per degC, so the data exclude a TSR-sized year-to-year response. The year-to-year slope was less than 10% of the spatial slope (transfer 0.076). As published, mean mass decreased by about 20% per decade on the US Northeast shelf. We found two survey-process causes. (1) The weighing resolution changed (whole pounds, 0.1 kg, 0.1 lb, whole kg with floor values before 1995-2000); applying old protocols to modern data raised species mean mass by a median of 2-4% and by more than 100% for some small species. We harmonized each unit to its coarsest protocol. (2) Most of the Northeast shelf decrease occurred as steps shared by 73-91% of species at the documented 1985 door change and 2009 vessel and net change (71-76% of the total log decrease). Within gear regimes the trends were -5 to -6% per decade there, -14% per decade on the Scotian Shelf (gradual, 1980-2000), and -1.7% per decade on average (95% CI -5.6 to +2.4), not related to local warming. Limitation: catch mean mass mixes growth, recruitment, fishing and gear selectivity; no lengths or ages; break criteria were chosen after inspection.

temperature-size rule · body size · demersal fish · bottom trawl survey · FishGlob · climate change · space-for-time substitution · survey gear change · data quality · weight resolution · Northeast US shelf · Scotian Shelf · eastern Bering Sea