Trophic amplification: A model intercomparison of climate driven changes in marine food webs

As shown in S9 Fig, trophic amplification based on low trophic level driver of each MEMis reduced but consistent with to trophic amplification based on NPP under GFDL.On the con- trary, the signal inversion between NPP and other MEMlow trophic level driver, under IPSL, lead to a different perception of trophic amplification (compared to trophic amplification rela- tive to NPP) for APECOSM; EcoOcean, FEISTY and ZooMSS. The inversion in trend under IPSL (Fig 3, changes in consumer biomass compared to changes in NPP) becomes an amplifi- cation for the four aforementioned models when considering the changes in the individual model’s low trophic level drivers (S9 Fig). 5. Conclusion The main response to the global decrease in NPP is a negative amplification, resulting in a more pronounced decrease in consumer biomass. Our study suggests that this response may be due to the way temperature influences different physiological and ecological processes in each ecosystem model. When an increase in NPP is projected, the temperature can offset the positive impact by attenuating or reversing the changes in biomass, leading to smaller increases or even decreases in biomass corresponding to positive attenuation or negative inversion respectively. Disaggregation of trophic amplification across different sizes within the food web highlighted a greater decrease in biomass of larger, high trophic levels on average. As a result, there are believable parts of the ocean for which we may see a large climate-induced decline in fishable biomass despite no change or increasing NPP.

Supporting information S1 Table. A taxonomy of marine ecosystem models taking part in the Fish-MIP project (modified from (Heneghan et al., 2021; Lotze et al., 2019; Tittensor et al., 2021)). (TIF) S2 Table. Conversion between trophic levels-based and size-based bins. (TIF) S1 Fig. Spatial patterns of projected total consumer biomass relative changes in percent. Shown are global ensemble projections at a 1 × 1 degree resolution. mean 2090s individual model projections of total consumer biomass relative change over the reference period (1995_2014) under: (a) GFDL-SSP5-8.5; and (b) IPSL-SSP5-8.5. (TIF) S2 Fig. Main outputs of the earth system models considered in the current study. mean changes in Sea surface temperature (a) and net primary production (b) in the 2090s relative to the reference period 1995_2014, under SSP5-8.5 for IPSL. (TIF) S3 Fig. Percentage of surface covered by each total consumer biomass response type, for the seven considered MEMS, under GFDL-SSP5-8.5 modelling. (TIF) S4 Fig. (a) Spatial distribution of the total consumer biomass response types, for the nine con- sidered MEMs, forced with IPSL-SSP5-8.5 modelling and (b) magnitude of the different types of change for the seven considered MEMS forced with IPSL-SSP5-8.5 Each response type is estimated at a 1 × 1 degree resolution, from the change in consumer biomass and NPP expected at the end of the century, relatively to the reference period. (TIF) PLOS ONE Trophic amplification revealed PLOS ONE | https://doi.org/10.1371/journal.pone.0287570 August 23, 2023 18 / 23