Marine mesozooplankton play an important role for marine ecosystem functioning and global biogeochemical cycles. Their size structure, varying spatially and temporally, heavily impacts biogeochemical processes and ecosystem services. Mesozooplankton exhibit size changes throughout their life cycle, affecting metabolic rates and functional traits. Despite this variability, many models oversimplify mesozooplankton as a single, unchanging size class, potentially biasing carbon flux estimates. Here, we include mesozooplankton ontogenetic growth and reproduction into a 3-dimensional global ocean biogeochemical model, PISCES-MOG, and investigate the subsequent effects on simulated mesozooplankton phenology, plankton distribution, and organic carbon export. Utilizing an ensemble of statistical predictive models calibrated with a global set of observations, we generated monthly climatologies of mesozooplankton biomass to evaluate the simulations of PISCES-MOG. Our analyses reveal that the model and observation-based biomass distributions are comparable (r$_{pearson}$=0.40, total epipelagic biomass: 137TgC from observations vs. 232TgC in the model), with similar seasonality (r$_{pearson}$=0.25 for the months of maximal biomass). Including ontogenetic growth in the model induced cohort dynamics and variable seasonal dynamics across mesozooplankton size classes and altered the relative contribution of carbon cycling pathways. Younger and smaller mesozooplankton transitioned to microzooplankton in PISCES-MOG, resulting in a change in particle size distribution, characterized by a decrease in large particulate organic carbon (POC) and an increase in small POC generation. Consequently, carbon export from the surface was reduced by 10\%. This study underscores the importance of accounting for ontogenetic growth and reproduction in models, highlighting the interconnectedness between mesozooplankton size, phenology, and their effects on marine carbon cycling.

Inter-annual to decadal variability in the strength of the land and ocean carbon sinks impede accurate predictions of year-to-year atmospheric carbon dioxide (CO2) growth rate. Such information is crucial to verify the effectiveness of fossil fuel emissions reduction measures. Using a multi-model framework comprising prediction systems based on Earth system models, we find a predictive skill for the global ocean carbon sink of up to 6 years. Longer regional predictability horizons and robust spatial patterns are found across single models. On land, a predictive skill of up to 2 years is primarily maintained in the tropics and extra-tropics enabled by the initialization of the physical climate variables towards observations. We further show that anomalies of atmospheric CO2 growth rate inferred from natural variations of the land and ocean carbon sinks are predictable at lead time of 2 years and the skill is limited by the land carbon sink predictability horizon.

Filter-feeding gelatinous macrozooplankton (FFGM), namely salps, pyrosomes and doliolids, are increasingly recognized as an essential component of the marine ecosystem. Unlike crustacean zooplankton (e.g., copepods) that feed on preys that are an order of magnitude smaller, filter-feeding allows FFGM to have access to a wider range of organisms, with predator over prey ratios as high as 10$^5$:1. In addition, most FFGM produce carcasses and/or fecal pellets that sink 10 times faster than those of copepods. This implies a rapid and efficient export of organic matter to depth. Even if these organisms represent $<$5\% of the overall planktonic biomass, the induced organic matter flux could be substantial. Here we present a first estimate of the influence of FFGM organisms on the export of particulate organic matter to the deep ocean based on the marine biogeochemical model NEMO-PISCES. In this new version of PISCES, two processes characterize FFGM: the preference for small organisms due to filter feeding, and the rapid sinking of carcasses and fecal pellets. To evaluate our modeled FFGM distribution, we compiled FFGM abundance observations into a monthly biomass climatology using a taxon-specific conversion. A model-observation comparison supports the model ability to quantify the global and large-scale patterns of FFGM biomass distribution, but reveals an urgent need to better understand the factors triggering the boom-and-bust FFGM dynamics before we can reproduce the observed spatio-temporal variability of FFGM. FFGM contribute strongly to carbon export at depth (0.4 Pg C yr$^{-1}$ at 1000 m), particularly in low-productivity region (up to 40\% of organic carbon export at 1000 m) where they dominate macrozooplankton by a factor of 2. The FFGM-induced export increases in importance with depth, with a simulated transfer efficiency close to one.