Warm years and warm stations had less zooplankton off California and the US Northeast in both periods, but temperature slopes did not forecast later decadal volume (1951-2023)

Roemmich and McGowan (1995) related an 80% decrease of zooplankton volume off southern California in 1951-1993 to surface warming. We tested two predictions of this explanation. We used 15,705 CalCOFI tows at 66 fixed stations (1951-2023) and 29,781 EcoMon tows on the US Northeast shelf (1977-2023). We corrected the CalCOFI net and depth changes with the Ohman and Smith (1995) factors and controlled for station, season and day or night. Within cruises, warmer stations had less volume in both periods (CalCOFI -0.21 and -0.19 log units per C; EcoMon -0.03 to -0.06). The annual CalCOFI slope weakened from -0.58 (1951-1993) to -0.23 per C (1994-2023). The 1951-1993 SST relation did not forecast the 1994-2023 level (skill 0.17, 95% interval -0.36 to 0.48; forecast 0.39 log units too high). In 1999-2013, SST was back at the 1951-1976 level but volume stayed 0.70 log units lower (ratio 0.50; gear-factor limits -0.47 to -0.90); the fraction of small organisms gave the same result. On the Northeast shelf, about 1.3 C of warming after 2000 came with no decrease, and the 1977-1999 SST slope forecast was worse than the old mean (skill -1.8). Thus local and year-to-year temperature relations repeated, but they did not predict the multidecadal level in either system. Limitations: displacement volume mixes taxa, gear factors come from few paired tows, years after 2017 have few CalCOFI tows, and the series are short.