Future phytoplankton diversity in a changing climate

Marine ecosystems provide crucial socio-economic services, including fisheries and climate regulation through CO2 absorption. Biodiversity loss, driven by human activities and climate change, threatens these services. Ocean warming, altered nutrient supply, ocean acidification, and deoxygenation are expected to cause significant community reorganization. Predicting these changes is challenging due to limited consistent sampling, inherent variability in community composition, and the complex interplay of stressors. Phytoplankton, forming the base of the marine food web, are particularly important due to their role in primary productivity and biogeochemical cycles. This study uses a complex ecosystem model, surpassing the simplicity of Earth System Models (ESMs) typically used in IPCC projections, to analyze the response of phytoplankton diversity to a high-emissions climate change scenario (similar to RCP8.5). Unlike simpler correlative or niche models, this model mechanistically represents phytoplankton community structure, considering multiple functional groups and size classes that interact with changing oceanic conditions, allowing for a more realistic assessment of diversity changes.

Existing studies on future phytoplankton changes have used various approaches, each with limitations. Niche models and correlative approaches assume contemporary relationships between environmental factors and phytoplankton abundance will remain constant in the future. Earth System Models (ESMs), commonly used for IPCC projections, typically incorporate only 2-3 phytoplankton types, which severely limits their representation of community complexity. These simpler models primarily focus on the effects of changing nutrient supply on phytoplankton, which favors small, nutrient-affine species. However, they fail to account for other factors influencing phytoplankton competitiveness and niche loss. This study addresses these limitations by employing a more comprehensive ecosystem model.

This research utilizes a previously described marine ecosystem model with 35 phytoplankton types and 16 zooplankton size classes, organized across seven biogeochemical functional groups (prokaryotes, pico-prokaryotes, coccolithophores, diazotrophs, diatoms, mixotrophic dinoflagellates, and zooplankton). The model incorporates the cycling of carbon, phosphorus, nitrogen, silica, iron, and oxygen, using Monod kinetics and size-related parameters for growth, grazing, and sinking. Zooplankton grazing is modeled using a Holling III function. The model is coupled with the MIT Integrated Global System Model (IGSM) for a physical framework with 2º × 2.5º horizontal and 22 vertical layers. The simulation runs from 1860 to 2100, using observed emissions until 1990 and a high emissions scenario (similar to RCP8.5) thereafter. Analysis focuses on biomass integrated over the full ocean depth and the period 2006–2100. Phytoplankton types contributing less than 0.1% to total biomass are excluded. Functional richness is defined as the number of coexisting phytoplankton types. Shannon diversity, incorporating both richness and evenness, is calculated. Turnover, representing the proportion of types changing between timepoints, is also quantified. Finally, the slope of the phytoplankton size spectrum, reflecting the dominance of different size classes, is determined.

The model projects a decrease in phytoplankton biomass across much of the tropical and subtropical ocean due to reduced nutrient supply, consistent with previous studies. Conversely, increased biomass is predicted in high-latitude regions due to sea ice retreat and longer growing seasons. Changes in biomass generally mirror changes in richness, with declines in the northern hemisphere subtropical and temperate regions and increases in polar and some equatorial regions. Up to 30% of modeled phytoplankton types may go locally extinct in some tropical areas, while colonization surpasses extinction in polar regions. The disappearance of less-competitive, larger phytoplankton due to declining nutrient supply rates leads to decreased richness in many northern hemisphere (sub)tropical regions. Changes in nutrient ratios impact functional group distributions, with diazotroph range expansion and diatom richness decline. Shannon diversity shows an almost global decline, primarily due to decreased evenness, indicating biomass concentration in fewer types. Community turnover is highest in temperate northern hemisphere regions and the South Pacific subtropical gyre, but the turnover *rate* increases across most of the ocean by the end of the century, indicating increasing community instability. The phytoplankton size spectrum's slope decreases in most subtropical regions and the Southern Ocean, indicating a shift towards smaller phytoplankton. In contrast, the increase in size in the North Atlantic is due to an influx of larger dinoflagellates and a diatom loss. These findings should be interpreted considering the model's limitations, primarily the lack of explicit representation of thermal norms or adaptation, coastal regions, sea-ice communities, and anthropogenic impacts beyond climate change.

The mechanistic nature of this model, more complex than typical ESMs, allows for a more nuanced assessment of niche loss and changing phytoplankton competitiveness. The predicted biomass changes align with previous studies, but the increased complexity reveals underlying community structure alterations. Reduced macronutrient supply leads to grazer abundance decline and trophic interaction changes, impacting size class and functional group coexistence. This study's findings broadly agree with previous research using correlative approaches, showing decreased diversity in the tropics and increased diversity in high latitudes, and a shift towards smaller phytoplankton, but discrepancies exist due to the differing assumptions regarding future environmental relationships. The decline in Shannon diversity is driven by both decreased richness and evenness depending on the region. High turnover and decreased richness in some regions suggest reduced niche diversity, while increased richness and high turnover in others imply niche expansion. The projected shift toward smaller phytoplankton has implications for food web productivity and carbon sequestration. The striking increase in the turnover *rate* indicates a loss of ecological resilience and community instability.

This study demonstrates that climate change will significantly disrupt marine phytoplankton communities, especially under high greenhouse gas emissions. The projections reveal the community's vulnerability, integrating exposure to stressors and sensitivity. The model's lack of adaptation implies the results may represent a worst-case scenario. Future research should incorporate phytoplankton adaptation and migration, along with more detailed trophic interactions, to enhance model realism. The predicted changes pose substantial challenges to the marine food web's productivity, particularly affecting nations relying on fisheries.

The model does not explicitly account for phytoplankton adaptation or evolution, which could increase ecological resilience. Coastal regions, sea-ice communities, and additional anthropogenic impacts (beyond climate change) are not explicitly represented. The model's resolution may also limit its ability to fully capture regional variations in community dynamics. The model doesn't explicitly incorporate the effects of many traits such as thermal niches, morphology, or colony formation, reducing the complexity of the diversity in the model compared to the real ocean.