Biological response
Field experiments show site- and season-specific phytoplankton responses to iron addition in HNLC waters.
SOUTHERN OCEAN OIF RESEARCH
Ocean iron fertilisation research connects iron-limited marine ecosystems to atmospheric carbon, carbon durability and ecosystem response.
Can Southern Ocean OIF deliver durable, additional CO₂ removal at climate-relevant scale?
Liquid Trees researches four linked requirements: additional net atmospheric CO₂ removal, durability, ecosystem response, and credible measurement, reporting, verification and governance.
Research priorities: quantify air–sea CO₂ flux, trace carbon fate, measure ecological effects and define regulatory and governance thresholds.
- 1 Equatorial Pacific
- 2 Subarctic North Pacific
- 3 Southern Ocean
Ocean Iron Fertilization (OIF):
Remove carbon.
Restore ocean life.
Ocean Iron Fertilization (OIF) is a measured, evidence-based method to enhance marine productivity and durably sequester carbon in the deep ocean.
Ocean iron fertilisation mechanism
How iron availability changes phytoplankton productivity and carbon pathways.
Trace iron can stimulate phytoplankton growth in HNLC waters. The central questions are how much additional atmospheric CO₂ removal follows, how durable it is and which ecosystem changes accompany it.
What is OIF?
Ocean Iron Fertilization (OIF) involves adding trace amounts of bioavailable iron to the surface waters of high-nutrient, low-chlorophyll (HNLC) regions where iron limits phytoplankton growth. The process mimics natural iron inputs to enhance marine productivity and increase the transport of carbon into the deep ocean under rigorous scientific monitoring — helping to mitigate climate change.
Professor Victor Smetacek explains the basics of OIF in this video.
Climate-scale research question
2–4
GtCO₂/yr
GtCO₂/yr
Could Southern Ocean OIF remove 2–4 GtCO₂ per year?
Answering this requires full-system carbon accounting, durable storage, ecological safety, credible measurement and lawful governance at climate-relevant scale.
Scientific and regulatory context: Bach et al. (2023), DOI · IPCC AR6 WGIII, Chapter 12 · IMO assessment framework
Research scope
From biological response to net atmospheric removal.
Removal and durability
Measurement must connect air–sea CO₂ exchange, carbon export, remineralisation, circulation and re-emission over relevant timeframes.
Environmental and legal conditions
Ecological monitoring, uncertainty bounds, independent review and applicable international and national frameworks define the conditions for research.
Research pathway
Five evidence gates from mechanism to governed research.
Each gate defines the evidence required to proceed, redesign or stop. Together they connect biological response to net removal, durability, ecological safety and legitimate governance.
Scientific case & baseline
Define mechanism, local conditions, measurement limits, and ecological context before any intervention is considered.
Additional net atmospheric removal
Distinguish a bloom or carbon export signal from additional net atmospheric removal across the full system.
Durability & leakage
Test where carbon goes, how long it remains out of the atmosphere, and whether displacement or leakage erodes the result.
Ecological safety
Measure ecological indicators alongside carbon pathways; define precautionary thresholds before proceeding.
Legitimate governance, independent review & proceed-or-stop
Enable scrutiny, authority to pause, and a clear decision: proceed only if evidence and legitimacy justify it; otherwise stop.
Evidence and uncertainty
Separate field observations, model results and unresolved system effects.
The evidence base connects mechanism, attribution, durability, ecological effects and governance while preserving the confidence level and limits of each source.
Evidence requirements
Mechanism, measurement, ecology and governance form one assessment.
- Mechanism
What process is proposed, and what assumptions make it plausible? - Measurement
What can actually be observed, for how long, and at what confidence? - Ecology
What signals would indicate unacceptable harm or insufficient understanding? - Governance
Who can review, challenge, pause, or prevent the work?
Decision threshold
Conditions to redesign or stop
Unacceptable ecological risk, an outcome that cannot be measured, inadequate consent or oversight, or evidence that the carbon-removal premise does not hold.
A stop decision is a valid research result.
Primary and authoritative sources
Evidence by mechanism, measurement, risk and governance.
Field experiments, syntheses, models and regulatory frameworks are listed with study context, supported conclusions, limitations and relevance to research decisions.
1 Mechanism evidence→
2 Carbon export evidence→
3 Net-removal modelling→
4 Scale, ecology & governance
Iron can stimulate phytoplankton in HNLC waters
- Source & study context
- Boyd et al. (2007), Science, DOI 10.1126/science.1131669 — synthesis of 12 mesoscale iron-addition experiments conducted from 1993–2005 across several HNLC regions.
- What it supports
- Iron limitation and a bloom response are well supported at the locations, seasons, doses and observation periods studied.
- Limits / counterevidence
- Bloom formation is not equivalent to additional net atmospheric CO₂ removal, deep export, durability, ecosystem benefit or scalability. Responses varied and regional extrapolation was limited.
- Decision relevance
- A site-specific mechanism and baseline case is necessary, but it cannot pass the removal, durability or safety gates by itself.
Deep export occurred in a particular Southern Ocean experiment
- Source & study context
- Smetacek et al. (2012), Nature, DOI 10.1038/nature11229 — five-week EIFEX study inside a coherent Antarctic Circumpolar Current eddy.
- What it supports
- Multiple measurements indicated that at least half of the fertilised bloom biomass sank below 1,000 metres in this diatom-dominated setting.
- Limits / counterevidence
- One eddy, one bloom and a short observation window do not establish general export efficiency. Martin et al. (2013), LOHAFEX, DOI 10.1002/gbc.20077 found doubled productivity but no fertilisation-induced increase in downward particle flux during a 39-day, 300 km² study.
- Decision relevance
- Any proposed study must measure carbon fate rather than infer it from chlorophyll, surface drawdown or a bloom alone.
Export is not the same accounting boundary as removal
- Source & study context
- National Academies (2022), Chapter 3: Nutrient Fertilisation, DOI 10.17226/26278 — independent assessment of efficacy, durability, MRV, impacts and research needs.
- What it supports
- The biological response has a substantial knowledge base, while carbon fate, efficiency, durability, MRV and unintended consequences require further research.
- Limits / counterevidence
- Full-system carbon accounting requires attribution, leakage, life-cycle emissions, counterfactuals, uncertainty and reversal.
- Decision relevance
- Verification requires a public method for net atmospheric removal, durability, leakage, life-cycle emissions and uncertainty.
Latest modelling sharpens—not closes—the trade-off question
- Source & study context
- Yu et al. (2026), Nature, DOI 10.1038/s41586-026-10795-y — process-rich simulation of 60 years of fertilisation across ten ocean biomes, published 29 July 2026.
- What it supports
- The model found strong regional differences: the Southern Ocean combined comparatively higher efficiency and lower modelled ecological risk than the other assessed biomes.
- Limits / counterevidence
- This is a scenario model, not field validation. More than half of modelled CDR was re-emitted within decades after fertilisation stopped, and non-local outcomes complicated crediting, equity and governance.
- Decision relevance
- Geography and downstream nutrient effects are core MRV variables; model ranking alone does not establish permanence or ecological safety.
Universal ‘no harmful bloom’ claims are not supportable
- Source & study context
- Trick et al. (2010), PNAS, DOI 10.1073/pnas.0910579107 — iron-enrichment incubations and observations at Ocean Station PAPA in the eastern subarctic Pacific.
- What it supports
- Iron addition favoured toxin-producing Pseudo-nitzschia and increased cellular domoic acid in this setting, establishing a credible hazard signal.
- Limits / counterevidence
- The study did not demonstrate ecosystem damage from a Southern Ocean field programme, and its context cannot be transferred universally. It also prevents a universal claim that toxic responses are impossible.
- Decision relevance
- Species composition and toxins require pre-specified monitoring and stop thresholds; contrary evidence must remain visible rather than being dismissed as a misconception.
Nutrients, oxygen, acidity, trace gases and food webs can shift
- Source & study context
- IPCC AR6 WGIII, Chapter 12, ocean fertilisation assessment — synthesis of field, modelling and governance literature.
- What it supports
- The assessment identifies nutrient redistribution, ecosystem restructuring, enhanced oxygen consumption and subsurface acidification, altered food webs, and possible N₂O/CH₄ effects as risks or spillovers.
- Limits / counterevidence
- Direction and magnitude are site-, duration- and scale-dependent; the assessment neither proves harm in every bounded study nor supports universal safety or benefit.
- Decision relevance
- Measure downstream and subsurface effects across a defensible space/time boundary; uncertainty or unacceptable risk can require redesign or stop.
Drinking-water limits do not establish marine ecosystem safety
- Source & study context
- IMO / London Convention and Protocol Assessment Framework (LC-LP.2(2010)) — official case-by-case framework for scientific research involving ocean fertilisation.
- What it supports
- A proposal requires site description, exposure assessment, effects assessment, risk characterisation with uncertainty, risk management, monitoring and contingency planning.
- Limits / counterevidence
- A human drinking-water comparator does not resolve iron speciation, bioavailability, dose geometry, sensitive taxa, food-web exposure, cumulative effects or downstream transport.
- Decision relevance
- Replace generic safety-margin reasoning with a proposal-specific exposure/effects case reviewed under the complete framework; there is no experiment-size exemption.
Every proposal needs a complete legitimacy and assessment pathway
- Source & study context
- Resolution LC-LP.2(2010) and its Assessment Framework; IMO marine-geoengineering status and 2025 update; Antarctic Treaty environmental-impact assessment record.
- What it supports
- Scientific attributes, environmental assessment, precautionary risk management, consultation, notification, reporting and monitoring must be evaluated case by case; activities in the Antarctic Treaty area require prior environmental-impact assessment through a relevant national Party.
- Limits / counterevidence
- The 2013 London Protocol amendment is not yet in force. The 2010 Assessment Framework remains the principal international reference for proposal-level assessment.
- Decision relevance
- Proposal-level assessment, monitoring, contingency planning and lawful authorization are required under applicable frameworks.
Phytoplankton & the carbon cycle
Phytoplankton are the ocean’s invisible forests. Through photosynthesis, they absorb CO₂, fuel food webs, and help transfer carbon into the deep ocean — making them a critical part of Earth’s climate system.
Whales & the iron pump
See our conversation with Matthew SavocaGreat whales help recycle iron at the surface. Their feeding and nutrient-rich plumes keep iron bioavailable for phytoplankton, amplifying productivity and linking top predators to microscopic climate work.
“Iron Is the Switch” — Christine Klaas
See our conversation with Christine KlaasPlankton ecologist Dr. Christine Klaas (AWI) discusses diatoms, silica, and why iron availability flips Southern Ocean ecosystems into bloom — plus what good science and governance need next.
Ecosystems in motion
Professor Victor Smetacek traces the pathway: diatoms → krill → whales — and back to iron at the surface. He outlines nature-mimicking OIF to help restore biodiversity in the Southern Ocean by reviving the iron cycle that powers this web.
Risks and governance
Assess ecological effects, downstream transport, reversibility and decision authority together.
Evidence on harmful algal species, nutrient redistribution, oxygen, acidity, trace gases and food webs informs site selection, monitoring thresholds, contingency planning and stop criteria.
Misconceptions of OIF
Explore Common Misconceptions →Clear, evidence-based answers to the most common questions and headlines.
Misconceptions, cleared
Professor Victor Smetacek addresses domoic acid headlines and harmful-bloom fears — clarifying how toxin events arise, why iron additions aren’t blunt triggers, and how informed, evidence-based design can restore ocean function without fueling fear.
Research library
Science, safety, governance and expert perspectives.
Evidence & measurementMechanism, carbon partitioning, natural iron inputs, field observations, models and measurement questions.
Risks & governanceEcological risks, exposure assessment, regulatory frameworks, monitoring and stop criteria.
Questions & counterargumentsEvidence on carbon export, harmful blooms, nutrients, food webs and governance.
Expert perspective: Christine KlaasDiatoms, silica limitation, ecosystem response and scientific governance in the Southern Ocean.
Expert perspective: Matthew SavocaWhales, iron recycling, phytoplankton and marine nutrient cycles.
Whales & the iron pumpResearch on whale-mediated iron recycling and Southern Ocean productivity.
PeopleLiquid Trees team, scientific advisers and organisational experience.
Track recordFreshwater restoration, photobioreactor work and the development of Liquid Trees.
Open research questions
Measurements and decisions that determine research viability.
Which baseline distinguishes intervention signals from natural variability?
Physical, chemical and biological observations must define seasonal variability, spatial controls, sampling resolution and counterfactual conditions.
Discuss baseline design ↗How can MRV quantify additional net atmospheric CO₂ removal and uncertainty?
The accounting boundary must include air–sea exchange, carbon export and fate, leakage, lifecycle emissions, reversals and confidence intervals.
Discuss MRV methods ↗Which institutions hold review, permit, pause and stop authority?
A viable framework must connect London Convention and Protocol processes, Antarctic Treaty environmental assessment, relevant national authorities, independent review and affected interests.
Discuss governance ↗Research and collaboration
Evidence and decision pathways for each audience.
ScientistsExperimental evidence, carbon accounting, durability, ecosystem monitoring and open research questions.RegulatorsEnvironmental assessment, applicable international frameworks, monitoring, contingency planning and pause or stop authority.VCs & angelsResearch milestones, technical risks, regulatory dependencies and milestone-based capital allocation.Climate & geoengineering supportersMechanism, climate relevance, safety questions, counterevidence and routes to informed participation.