The Global Distribution of Grazing Dynamics Estimated From Inverse Modeling

Abstract

Grazing dynamics are one of the most poorly constrained components of the marine carbon cycle. We use inverse modeling to infer the distribution of community‐integrated zooplankton grazing dynamics based on the ability of different grazing formulations to recreate the satellite‐observed seasonal cycle in phytoplankton biomass after controlling for physical and bottom‐up controls. We find large spatial variability in the optimal community‐integrated half saturation concentration for grazing (K1/2), with lower (higher) values required in more oligotrophic (eutrophic) biomes. This leads to a strong sigmoidal relationship between observed mean‐annual phytoplankton biomass and the optimally inferred grazing parameterization. This relationship can be used to help constrain, validate and/or parameterize next‐generation biogeochemical models.

Impact of Spatial Variability in Zooplankton Grazing Rates on Carbon Export Flux

“…Generally low k values were estimated in the subtropical oligotrophic gyres, where communities are dominated by faster grazing microzooplankton, in particular pico-and nano-sized flagellates (Calbet & Calbet, 2008). Anomalous high k 0 values inferred in the hyper-oligotrophic South Pacific gyre (Ras et al, 2007) are in disagreement to the study by Rohr et al (2024) and coincide with an underestimation of NPP, small phytoplankton biomass and near-zero growth rates. This suggests that in this region, small phytoplankton growth rates used to force the model may be too low.…”

“…In addition, the lack of explicit representation of temperature within the model may affect growth and grazing rates in these extreme environments. The 1P1Z 3D model in Rohr et al (2024) includes temperature which may negate these issues.…”

“…Laboratory measurements of k values (Hansen et al, 1997;Hirst & Bunker, 2003) show a large range in k values across all species studied, spanning 0.96-6,000 mgC m 3 . Laboratory (Hansen et al, 1997), ecological (Barton et al, 2013) and modeling (Rohr et al, 2024) studies also point toward strong spatial variability in k values. However, even the most complex BGC models have fixed, globally homogenous, k values.…”

“…The rate that these forms of particulate carbon are exported via the biological carbon pump is directly influenced by zooplankton grazing. Grazing rates impact the biomass of both predator and prey (Rohr et al, 2024) and consequently the production of algal aggregates and fecal pellets. The production of aggregates and fecal pellets are influenced in opposite directions by grazing pressure.…”

“…These models can simulate spatial variability in grazing dynamics through the competition of multiple plankton functional types, but this likely does not capture the full physiological variability. Some models use mechanisms such as multi-prey responses (e.g., T. R. Anderson et al, 2010Anderson et al, , 2015 and prey preferences (e.g., Aumont et al, 2015) to further emulate this variability, but these mechanisms are limited and there is little observational data to confirm what emergent grazing dynamics should be (Rohr et al, 2024). Furthermore, there is uncertainty around the impact of zooplankton grazing on carbon flux, which contributes to the large variability in global estimates of carbon export (Boyd, 2015;Rohr et al, 2024;Siegel et al, 2014).…”