Mean community direction scores remained negative after a sampling correction in four US trawl samples in 1982-2019

We tested whether annual changes in recorded marine communities tend to reverse direction. Adjacent changes share a survey year, which can create a negative direction score through sampling error. We derived a finite-sample correction under independent haul sampling assumptions. Mean corrected scores were -1.15, -1.30, -1.56 and -1.85 in eastern Bering Sea and Northeast US fall samples in two periods from 1982 to 2019. All four adjusted mean intervals were below zero. The correction removed 34-78 percent of the magnitude of direct negative means. A positive link to the product of successive seasonal temperature changes was clear only in early Northeast records. Increasing the modeled correction by 50 percent changed the sign of both Northeast means. Thus the sampling assumptions are material. This result does not prove ecological cycles or a causal climate effect. Spatial dependence, detection, gear, short series and coarse surface temperature limit inference. The package gives a proof, exact enumeration checks, raw source hashes, a full raw rebuild check, code, derived data and a figure.

biodiversity · community change · sampling error · trawl survey · temperature · observation bias · adjacent changes