Dissertation Defense: Jun Yu

Title: Anthropogenic Perturbations to Marine Nutrient Cycling Reshape Ocean Biogeochemistry and Ecosystems

Abstract: Human activities are increasingly disturbing ocean conditions, altering ocean biogeochemistry, ecosystem structure, animal habitats, and the fishery yields that support human societies. However, the impacts of these perturbations remain poorly constrained because most climate models have limited representations of marine ecosystem structure and biogeochemical processes. In this thesis, I quantify how anthropogenic perturbations to marine nutrient cycling reshape ocean biogeochemistry and ecosystems across three related contexts: the development of a more ecologically complex ocean biogeochemical model, ocean iron fertilization as a proposed carbon dioxide removal strategy, and future climate change driven by warming, stratification, and changing atmospheric iron deposition.

First, I develop a next-generation ocean biogeochemical model, MARBL-8P4Z, with expanded ecosystem complexity and an improved ability to capture subtle ecological shifts that may shape ocean biogeochemistry under climate change (Chapter 2). The model broadly reproduces observed plankton community composition and group-specific biomass under present-day climate forcings, while maintaining skill in simulating key biogeochemical fields. These results suggest that MARBL-8P4Z can represent important climate-driven plankton community shifts that are likely to influence future ocean biogeochemistry.

Second, I use MARBL-8P4Z within CESM2 to assess the ecological consequences alongside the climate gains of ocean iron fertilization, a proposed marine CO2 removal (mCDR) strategy (Chapter 3). The results show strong regional disparities in climate–ecosystem trade-offs: biomes with comparable fertilization efficiency have sharply contrasting ecological outcomes through distinct pathways. Three tiers emerge: Southern Ocean (higher efficiency / lower risk), Equatorial Pacific and global fertilization (higher efficiency / higher risk), and subtropics (lower efficiency / medium risk). Decade-scale fertilization yields a net atmospheric CO2 removal of 1.1–5.3 ppm, with over half re-emitted within 60 years post-fertilization. However, the disconnect between local interventions and global outcomes complicates robust carbon crediting.

In Chapter 4, I investigate how anthropogenic iron deposition and warming-driven stratification jointly affect marine nutrient limitation, productivity, carbon export, and food-web structure from the historical period to 2100. Fully-coupled historical simulations show that increased soluble iron deposition since the Industrial Revolution has enhanced N₂ fixation and produced detectable downstream effects on primary production, carbon export, macrozooplankton biomass, and subsurface oxygen. However, future simulations show that warming-driven stratification and phosphorus stress dominate over differences in future iron deposition, driving a strong expansion of phosphorus limitation, a shift toward carbon-rich but nutrient-poor pico-sized phytoplankton, reduced diatom niches, and weakened particulate carbon export after mid-century. Using FEISTY fish-model projections forced by CESM2-MARBL-8P4Z outputs, I further show that these lower-trophic-level changes propagate through the food web and reduce global fish biomass, leading to projected declines in global fish biomass of 22% under SSP585 and 16% under SSP370 by the 2090s. Together, this thesis demonstrates that anthropogenic perturbations to marine nutrient cycling can produce persistent changes in ocean biogeochemistry and ecosystems, with important implications for carbon storage, biodiversity, ecosystem resilience, and food security.