Marine plankton, encompassing phytoplankton and zooplankton, are fundamental to marine ecosystems. Phytoplankton, through photosynthesis, contribute significantly to global primary production, forming the base of the food web that supports global fisheries. This intricate interplay between phytoplankton and zooplankton drives the biological carbon pump, a crucial regulator of the ocean-atmosphere CO2 balance. The biodiversity of these plankton communities is a key factor influencing the efficiency of this pump and the recruitment of commercially important fish stocks. A large body of research indicates a strong correlation between plankton diversity and climate, with temperature emerging as a primary driver. Warmer temperatures generally enhance species diversity by increasing metabolic rates and promoting speciation. However, ocean warming also forces species to shift their distributions poleward to maintain their optimal thermal habitats. These poleward shifts can weaken the biological carbon pump and, more critically, future warming could lead to species extinctions by exceeding thermal limits and altering community composition. The potential impacts of future warming on plankton diversity across different clades and trophic levels remain poorly understood. Earth system models (ESMs) are still limited in their ability to resolve species-level traits and interactions, while historical field observations are too sparse for robust global diversity projections. This study addresses these limitations by employing species distribution models (SDMs) to project future plankton diversity under a high-emission scenario.

Previous research has highlighted the importance of marine plankton biodiversity in maintaining vital ecosystem services such as the biological carbon pump and supporting fisheries. Studies have shown that temperature is a major driver of plankton diversity, with warmer temperatures generally leading to increased diversity. However, the effects of climate change, specifically ocean warming, have also been shown to induce poleward range shifts in species, potentially weakening the biological carbon pump and impacting ecosystem functioning. Existing studies on future plankton diversity have been limited by the resolution of Earth System Models (ESMs) or by limited spatial and temporal coverage of field observations. These limitations motivated the need for a more comprehensive approach using species distribution models to project global plankton diversity under future climate change scenarios. The authors also discuss previous research efforts, specifically highlighting Ibarbalz et al. (2019), which reported SST-driven latitudinal diversity gradients for photosynthetic protists and copepods, while noting discrepancies in the amplitude and location of these gradients compared to their findings.

This study utilized an ensemble approach based on Species Distribution Models (SDMs) to project the monthly and annual diversity patterns of marine plankton. The study incorporated a comprehensive dataset of species occurrence records for 860 plankton species (336 phytoplankton and 524 zooplankton) spanning various phyla, orders, and genera, covering ten major plankton functional groups (PFGs). These occurrence records (n=934,696) were compiled from various sources, aggregated onto a monthly resolved 1° × 1° grid, and matched with observation-based climatologies of environmental predictors including temperature, dissolved oxygen, solar irradiance, macronutrient concentrations, and chlorophyll *a* concentration. Four types of SDMs (Generalized Linear Models, Generalized Additive Models, Artificial Neural Networks, and Random Forests) were used, each with four alternative pools of predictors, resulting in a total of 16 habitat suitability models for each species. These models were projected into the future using outputs from five ESMs under the RCP8.5 high greenhouse gas emission scenario. Monthly climatologies for the 2012-2031 and 2081-2100 periods were computed, and future monthly anomalies were derived and added to the observation-based climatologies to estimate future environmental conditions. The study estimated mean annual alpha diversity (species richness) and beta diversity (species turnover) patterns for both trophic levels, using the sum of species habitat suitability patterns averaged across all SDM, ESM, and predictor pool combinations. Uncertainties in projections were assessed using the interquartile range of the ensemble members. Species turnover was calculated using Jaccard's dissimilarity index, decomposed into true species turnover and nestedness components. Robustness tests were conducted to account for spatial and temporal sampling biases. Additional analyses investigated the relationship between temperature and species richness, identifying the main environmental drivers using predictor dominance analysis in the SDMs, and analyzing the overlap of diversity changes with marine ecosystem services. Specific details regarding data sources, variable selection, SDM algorithms, model evaluation, and uncertainty analyses are extensively documented in the methods section of the paper. To account for uncertainty in threshold selection, the authors used a range of thresholds (0.10 to 0.80) when calculating Jaccard's dissimilarity index to obtain a range of beta diversity estimates.

The study's key findings include: 1. **Contemporary Latitudinal Diversity Gradients:** The models predicted a strong latitudinal diversity gradient, with species richness decreasing from the equator to the poles for both phytoplankton and zooplankton. However, the richness maxima for these groups were not collocated, with phytoplankton peaking near the equator and zooplankton peaking in the subtropics. These patterns aligned with previous global and regional observations. 2. **Future Changes in Species Richness:** By the end of the century (2081–2100), under RCP8.5, the models projected a significant global increase in mean annual plankton species richness (5%), driven primarily by strong increases in temperate to subpolar latitudes (22%). Phytoplankton species richness was projected to increase globally by 16%, with even stronger increases in tropical regions. In contrast, zooplankton richness showed little global change, resulting from contrasting regional trends: a strong increase in temperate to subpolar latitudes (24%) and a slight decrease in the tropics (-4%). 3. **Community Restructuring:** The increase in species richness at higher latitudes was primarily driven by poleward range shifts of tropical and subtropical species. The overall median poleward shift velocity was 35 km/decade. This poleward shift leads to major restructuring of plankton communities, with significant species turnover (true turnover component of Jaccard's dissimilarity index averaging 18%). Turnover rates were particularly high in temperate and subpolar latitudes (reaching 40%) and even higher in the Arctic (>45%). About 40% of species associations were reshuffled due to the changes in community composition, revealing shifting species interactions. 4. **Temperature as a Primary Driver:** Temperature emerged as the primary driver of both contemporary species richness patterns and future changes. The non-linear temperature-diversity relationship explained the non-uniform response of phytoplankton and zooplankton richness to climate change. The steeper temperature-diversity slope for zooplankton resulted in larger increases in richness at higher latitudes while a reduction in richness in the tropics was driven by warming exceeding thermal tolerances. 5. **Ecosystem Service Implications:** The projected changes in plankton diversity and composition overlap significantly with areas of high ecosystem service provision. Regions with the most severe changes were also key contributors to carbon cycle services and small pelagic fisheries. The replacement of larger species with smaller ones at higher latitudes is expected to weaken carbon export efficiency. Conversely, regions such as the Southern Ocean and some upwelling areas exhibited weaker projected impacts, suggesting that some regions with high carbon export efficiency might remain relatively less affected.

The findings of this study provide a comprehensive assessment of the projected impacts of climate change on global marine plankton diversity. The significant increase in species richness at higher latitudes, driven by poleward range shifts, highlights a major restructuring of plankton communities. The identification of temperature as a key driver aligns with the metabolic theory of ecology. However, the non-linear temperature-diversity relationship reveals nuances in the response of different plankton groups and highlights the potential for regional variations. The significant overlap between regions projected to experience major changes in plankton diversity and those providing important ecosystem services raises significant concerns. The projected weakening of the biological carbon pump and potential impacts on fisheries emphasize the broader ecological consequences of climate change on marine ecosystems. While this study provides valuable insights, further research is needed to fully understand the complex interactions within plankton communities and the long-term consequences of these changes. The discrepancies between the findings of this study and those of Ibarbalz et al. (2019) highlight the importance of methodological choices in projecting future biodiversity.

This study provides a novel, comprehensive projection of future plankton diversity under a high-emission scenario. The findings reveal substantial increases in species richness in temperate and subpolar regions driven by poleward range shifts, alongside a restructuring of communities. These changes pose significant risks to ecosystem services, particularly the biological carbon pump and fisheries. Temperature is identified as the key driver, but the non-linear relationship between temperature and diversity, along with species-specific responses, underscores the complexity of these shifts. Future work could focus on higher-resolution modeling, integrating more detailed species interactions, and quantifying the functional consequences of shifts in diversity and size structure.

The study's conclusions are based on a species distribution modeling approach that relies on several assumptions, including niche conservatism and limited influence of dispersal and biological interactions. The reliance on presence-only data and the potential for sampling biases could also affect the accuracy of species distribution estimates. The choice of SDMs and ESMs introduces uncertainty into the projections, as indicated by the variance across ensemble members. Additionally, the projections are limited to the RCP8.5 scenario, and results may vary under different climate change scenarios. Finally, the study focuses on species richness and composition, without explicit quantification of the functional consequences of community shifts on ecosystem functioning.