Sea warmth gave no clear link to expected shared-taxa fractions in four US trawl samples in 1982-2019

We tested seasonal sea temperature against the fraction of expected taxa shared by two disjoint samples with fixed effort. We used eastern Bering Sea and Northeast US fall catch records in 1982-2000 and 2001-2019. An exact finite-table calculation drew one haul per fixed stratum for each sample, without haul reuse. All four adjusted slope intervals included zero. Slopes were 0.08, -0.32, 0.02 and -1.26 percentage points per C. Lag, difference and survey-step checks also gave intervals that included zero. The prior late Bering and early Northeast richness links remained in expected shared richness as a descriptive result. Taxa expected in only one sample cannot alone explain those point links. The main null result does not prove stable detection or no climate effect. This is an empirical sampling diagnostic, not new repeat surveys or a true occupancy estimate. Short series, surface temperature, catchability and survey changes limit inference. The package includes code, exact draw formulas, data provenance, sufficient evidence, a raw rebuild route, independent numerical checks and a figure.

biodiversity · temperature · trawl survey · observation bias · rarefaction · shared richness · null result