Fish community temperature indices increased by 0.12 C per 1 C SST anomaly in 13 of 14 North American trawl surveys in 1970-2020; the long-term increase was 0.04-0.07 C per decade

Question: do demersal fish communities on North American shelves change toward warm-water species when the sea becomes warmer, from year to year and over decades? Data: 14 OceanAdapt fixed-strata bottom trawl surveys (121,135 hauls, 1970-2020, nine geographic groups), 667 fish species, ERSSTv5 and HadISST1 sea surface temperature (SST), and the WOA23 climatology. Method: the species temperature index (STI) is the mean climatological bottom temperature at the hauls that caught the species, fixed in time. The community temperature index (CTI) is the incidence-weighted mean STI of a survey year. We compared it with the footprint SST of the 12 months before the survey. Results: after linear detrending, the CTI was higher in warmer years in 13 of 14 surveys (10 with 95% CI above zero, none below). The pooled slope was 0.12 C of CTI per 1 C of SST (95% CI 0.08 to 0.16; group level 0.11). A species-shuffle null gave a combined z of 5.0 (3.8 over groups). The slope stayed positive with vessel-change steps, footprint and richness controls (0.08), core species, STI from other regions, HadISST1, and in both periods before and from 2000. It was smaller with biomass weights (0.036) and survey-season SST (0.05). The pooled long-term CTI trend was +0.041 C per decade (-0.005 to +0.088), against +0.26 C per decade of ERSSTv5 warming (ratio 0.19, CI -0.13 to 0.56). With step terms at reported vessel changes the trend was +0.059 (0.033 to 0.085; ratio 0.29). Only the two Northeast US surveys had long-term trends beyond the species-shuffle null. The two negative trends (Gulf of St. Lawrence) became near zero with vessel-change steps, but part of that decrease agrees with a real cold period. Limitations: SST is not bottom temperature; the two SST products disagree on trends; the STI comes from the same surveys; incidence depends on detection, and species per haul increased in 11 of 14 surveys; nominal survey months; association only. Conclusion: the year-to-year warm-species response is consistent but small, and the long-term change is about one fifth to one third of the SST warming rate.

community temperature index · species temperature index · climate debt · bottom trawl survey · demersal fish · OceanAdapt · sea surface temperature · ERSSTv5 · HadISST1 · World Ocean Atlas · thermophilization · observation bias · vessel change · species shuffle null · North America · continental shelf · range shift · regional replication