A 1951–1993 warming relation overpredicted CalCOFI zooplankton volume by 25–32% in 1994–2021 at 66 southern California stations, because volume stayed low in cool 1999–2013 years

Roemmich and McGowan (1995) related a large 1951–1993 decrease of CalCOFI zooplankton displacement volume to surface warming and stronger stratification. We tested this relation with the data that came after it. We used public NOAA ERDDAP zooplankton volumes (16,464 tows) and Scripps bottle hydrography (16,755 casts) at 66 core southern California stations, and we corrected for the 1969 and 1978 gear changes with the published factors. We fitted annual volume anomalies (log10) against stratification (T10 − T150) or 10 m temperature anomalies in 1951–1993 (37 years), and predicted 1994–2021 (28 years). The analysis plan was written before outcomes were calculated. The relations overpredicted 1994–2021 volume: the mean residual was −0.148 log10 (95% block-bootstrap interval −0.295 to −0.041) for total volume with stratification, and −0.126 to −0.168 for the other predictor and size-class combinations. All intervals exclude zero, also with gear-factor uncertainty and in tow-matched and spring-only variants. In 1999–2013, temperature and stratification were at the 1951–1968 level, but volume was 39% lower (−0.217 log10, −0.365 to −0.098). Total volume agreed with the prediction only in the warm years 2015–2021. The 1994–2021 slope against stratification was −0.04 per °C for total volume, against −0.24 in 1951–1993. For small plankton the year-to-year response stayed negative (−0.10 to −0.13 per °C). The linear decrease did not continue (1994–2022 trend +15% per decade, −13% to +38%). A relation fitted only to the 1978–1993 bongo years gave a smaller residual whose interval includes zero. Thus most of the 1951–1993 relation comes from the contrast before and after about 1977, and not from a reversible temperature response. Limits: displacement volume is not carbon and is strongly affected by salps. The core stations cover only southern California. Sampling was every third year in 1969–1984, and fewer volumes exist after 2017. The gear factors come from few paired tows. The gear correction, the analog-period contrast, first differences and the step model were added after the plan and are exploratory.

CalCOFI · zooplankton displacement volume · California Current · stratification · out-of-sample test · ocean warming · regime shift · gear correction · salps · long-term ecological data