Warming and cooling contrasts in corrected community change were uncertain in four US trawl samples in 1982-2019

We tested whether warming and cooling gave different amounts of annual community change at equal temperature-change magnitude. We used fixed-stratum trawl records from the eastern Bering Sea and Northeast US fall surveys in 1982-2019. A published measure fixes haul effort and corrects finite-sample noise under independent sampling assumptions. Four region-period models control change magnitude and linear year. Warming-minus-cooling contrasts at one degree range from -4.71 to +0.25 squared detection-probability units. All four adjusted intervals include zero. Covariance, uncorrected-distance and survey-transition checks give the same broad result. This does not show equal responses. Short series, coarse surface temperature, gear and detection limit inference. We do not estimate causal climate effects, heatwaves or pollution. The package gives data, raw source hashes, code, a figure and independent numerical checks.

biodiversity · trawl survey · warming · cooling · community change · observation bias · null result