Northward trend signs survived one catch loss per year in two Northeast US trawl periods within 1982-2019

We tested the effect of lost catch from one shared haul per survey year on mean latitude trends. We used fixed strata and taxon cohorts in two US trawl surveys in 1982-2000 and 2001-2019. Effort counts stayed fixed, and each taxon had to retain positive catch. Both Northeast US fall trends stayed northward even when a loss was allowed in every year. Their exact bounds were +6.02 to +16.29 and +11.48 to +22.73 km per decade. Bering Sea trend signs could reverse with losses in at least 5 early-period years or 14 later-period years. A single affected year could not reverse any of the four signs. We give sharp budget bounds, an independent calculation check and source retrieval code. These are fixed-table sensitivity bounds, not confidence intervals. Survey bias, catchability, selected taxa and other error types remain outside the test. No climate or pollution cause was measured.

species distributions · observation bias · marine biodiversity · trawl survey · sensitivity bounds · climate indicators