Dire climate warnings are now coming thick and fast. The world is not reducing emissions quickly enough to avoid potentially catastrophic climate change – as if what is happening already is not bad enough.
The ocean can provide at least part of the solution, by delivering clean energy. But can we also use it to help with the other part of the net zero equation, by increasing the marine removal of carbon dioxide? Here we consider an idea that has been around for some time, ocean fertilisation, contrasting its political unpopularity with more recent enthusiasm for coastal blue carbon. In both cases, the weakest scientific link is measuring whether or not the claimed climate benefits are being achieved.
The global market in carbon offsets is booming. The total value of traded carbon dioxide (CO2) permits in 2021 was 760 billion euros (US$ 851 billion), a fifteen-fold increase in just four years. This increase is a direct consequence of the UNFCCC1 Paris Agreement on climate change, with its goal of balancing human-driven emissions of greenhouse gases with removals by 2050. Around two-thirds of the global economy is now committed to this net zero target, with many countries, regional administrations and industries being even more ambitious. For example, Microsoft, Amazon, Maersk and nearly 400 other companies have signed the Climate Pledge, committing them to “neutralise any remaining emissions with additional, quantifiable, real, permanent and societally beneficial offsets” by 2040 ‒ or earlier.
The world is not reducing emissions quickly enough to avoid potentially catastrophic climate change
What’s not to like? Offsets are an inbuilt feature of the net zero concept, matching hard-to-eliminate greenhouse gas releases with the managed removal of an equivalent amount of such gases (in terms of their warming potential) from the atmosphere. But reliable measurements are crucial for the removal part of the equation, determining the plausibility of policy pledges. The necessary protocols for monitoring, reporting and verification (MRV) are relatively well-established for emissions and their direct reductions. The MRV framework, based on transparency and additionally, was strengthened at UNFCCC COP26 by Article 6 of the Paris Agreement, and relatively detailed guidelines have been developed by both the European Union and the UN Framework Convention on Climate Change (UNFCCC). However, equivalent international standards are currently lacking for less conventional CO2 removal approaches.

This deficiency is of particular concern for proposed climate solutions using natural marine biological processes (often referred to as ‘blue carbon’2) to increase carbon uptake and storage in either the open ocean or the coastal zone, potentially at the gigatonne scale.
Increasing carbon uptake in the deep ocean
The theoretical potential for increasing the natural ocean uptake of CO2 is immense: the global ocean contains around 50 times more carbon than the atmosphere, and it has already absorbed around a quarter of the extra CO2 that has been released by human activities. Enhancement of such carbon uptake was first suggested as a climate intervention more than thirty years ago, using iron as a fertiliser to increase plant biomass in the upper ocean. Subsequent field experiments in the Southern Ocean and elsewhere showed that the idea might work in deep water regions where iron levels are naturally very low: when a few tonnes of soluble iron were added, extensive blooms of phytoplankton(microscopic marine plants) developed, that could be tracked from space.
Phytoplankton are, however, short-lived. Only a small ‒ and highly variable ‒ proportion of their carbon is transferred to ocean depths or seafloor sediments, taking it out of circulation for climatically-significant timescales. The key knowledge gaps are therefore how much and for how long: what measurements are needed to track the fate of the extra carbon uptake over many thousand square kilometres, and over many decades? In addition, what other environmental monitoring will be required to quantify the impacts of major productivity changes on other ecosystem components, not only at the sea surface but throughout the water column and at the ocean floor? Will it actually be possible to separate out the effects of ocean fertilisation from the profound alterations in ocean physics, chemistry and biology that are already occurring as a consequence of global warming?
To address these issues, the US National Academies for Sciences, Engineering and Medicine have recently recommended a research budget of up to $440 million to determine whether ocean fertilisation would be a feasible CO2 removal technique to help achieve net zero.
But the biggest obstacles limiting the development of this approach relate to its public acceptability and governance, closely linked to the risk of damaging side effects. Out of 15 approaches considered, ocean fertilisation has been identified as the carbon removal method that presents the greatest threat to biodiversity. On the basis of the precautionary principle, it is strongly opposed by many non-governmental organisations, and has yet to be adopted by any government as a formal component of national climate policy.

These concerns resulted in a non-binding moratorium on ocean fertilisation by the Convention on Biological Diversity (CBD) in 2008. Subsequently, more general regulation on marine geoengineering in waters outside national jurisdiction has been developed under the Convention on the Prevention of Marine Pollution by Dumping of Wastes and Other Matter (also known as the London Convention and London Protocol, LC/LP). Whilst further research on ocean fertilisation is still possible (and there are many researchers keen to explore this approach), such studies are now subject to regulatory approval through an internationally-agreed assessment framework.
Increasing coastal carbon uptake
In contrast to ocean fertilisation, the use of vegetated coastal ecosystems for carbon removal and storage is widely well-regarded by governments, non-governmental organisations and the general public. In particular, the restoration of such ‘blue carbon’ habitats was included in initial Nationally Determined Contributions by 23 signatories to the Paris climate change agreement. In this context, the term blue carbon is restricted to just three coastal ecosystems with rooted vegetation: mangroves, salt marshes and seagrass meadows. Because their waterlogged sediments are anaerobic (lacking oxygen), the decomposition of buried organic material in such habitats is very slow, with the result that large carbon stores accumulate – similar to freshwater swamps and peatbogs.
Awareness of the climatic importance of carbon removal and storage by these ecosystems has developed rapidly in recent years. During the period 1990-2010, there were fewer than 20 research papers published per year on coastal blue carbon ecosystems (CBCEs); by 2020 that number had increased more than ten-fold, to 270 per year.
Effective conservation and protection of such CBCEs prevents the release of their sediment carbon stocks. That release can occur as a result of habitat degradation or land-use change, such as conversion to agriculture, aquaculture or coastal development. Appropriate legal protection for coastal wetlands now exists in most countries (even if not always well enforced), and the high rates of historical loss that had previously occurred have generally lessened, at least in Europe. In addition to carbon removal and storage, CBCEs provide indirect climate benefits through storm protection and preventing coastal erosion, as well as a wide range of non-climatic ecosystem services relating to biodiversity, coastal fisheries and water quality.

Carbon removal arising from the continuation natural uptake processes in CBCEs is, of course, highly desirable. That removal would, however, happen anyway, so it cannot be claimed as a ‘negative emission’ or carbon credit to offset emissions elsewhere. The focus for using CBCEs for climate mitigation is therefore restoration: management actions that provide quantifiable additionality in CO2 uptake and long-term storage. This can be achieved in three ways.
First, by resource management to improve local environmental conditions and promote natural functioning. For example, by restoring natural hydrodynamics to increase freshwater flows, tidal exchanges and sediment supply; by reducing pollution, particularly by nutrients; and by re-instating natural food webs (by increasing the number of predators). Second, by re-establishing such habitats where they had previously been lost as a result of land-use change or other coastal developments. This action involves planting seagrass or mangrove seedlings in subtidal or intertidal sediments; many such initiatives have already been carried out, although with varying (and generally low) success. Third, by the creation of entirely new habitats, such as purposeful coastal flooding to stimulate saltmarsh formation. Although this is not strictly a ‘restoration’ process, it is widely considered as such in terms of policy action.
The complexities of carbon accounting
Reliable carbon accounting is crucial for emission offsetting using coastal blue carbon. If a project involves the restoration of (say) 10 km2 of degraded seagrass, saltmarsh or mangrove habitat for climate mitigation purposes, governments (and the private sector) need to know in advance how many carbon credits can legitimately be claimed – with investment decisions based on the cost per tonne of CO2 removed from the atmosphere. They also need to know that that the long-term storage of the removed carbon is secure, without risk of future leakage.
These key issues have yet to be satisfactorily addressed. Many of the estimates that do exist seem over-simplistic. For example, the calculation that around 25% of degraded coastal wetlands can be reinstated at a cost of <US$ 100 per tonne of CO2 removed (apparently ignoring the nature of the land-use changes that are involved), or that planting mangroves will remove 45.1 kg CO2 per tree per year. These values may be valid for some sites, but it is much more likely that they won’t.

Other studies give much higher cost estimates, ranging from US$ 560 to 469,000 per tonne CO2 removed, depending on the ecosystem under consideration. A fundamental problem is natural variability in carbon removal rates: there is a 19-fold range between the lowest and highest reported values per unit area by mangroves, with a highly-skewed distribution (many more low results than high ones). For seagrasses, the range of reported values is greater, with a 76-fold range; for saltmarshes, greater still, with a 600-fold range.
As discussed in greater detail in a recent review (see Box for reference), this variability arises from a complex combination of environmental factors, that are not easy to allow for. There are also potential measurement errors in carbon accounting for CBCEs, with an overall risk that future climate change (the combination of warming and sea level rise) will threaten the long-term survival of the restored habitats.
One large uncertainty arises from the fact that CBCEs don’t just take up CO2 but can also release methane (CH4) and nitrous oxide (N2O). These greenhouse gases don’t last so long in the atmosphere, but they do they have a much stronger warming effect than CO2. Their emissions can be highly variable, changing with tidal conditions and time of year for saltmarshes. To find out the scale of such effects, it is therefore desirable – but rarely possible ‒ to carry out continuous measurements of the fluxes of CH4 and N2O throughout an annual cycle. For a CBCE restoration project, such measurements need to be made before any management actions are taken to restore the habitat – such as re-planting or re-flooding ‒ as well as every few years afterwards, to provide comparison with baseline conditions.
“coastal wetlands are inherently unsuitable for quantified carbon offsets”
Another factor than can cause large uncertainty in the carbon offset value for coastal wetland restoration is how much of the carbon accumulating in the CBCE sediments originates from the local, restored vegetation, and how much from elsewhere; for example, land-derived sources in estuaries. Such non-local carbon may include soot, micro-plastics and other highly recalcitrant3 forms that arguably would have been preserved anyway, regardless of whether or not the restoration action had occurred. On the basis that carbon credits should only be awarded for ‘additionality’ (i.e. the climate benefit is an unambiguous consequence of the management action), externally-derived carbon should be excluded. The ratio of local to non-local carbon therefore needs to be determined; this is possible using ‘fingerprinting’ techniques, but such analyses are not straightforward for operational monitoring.


Model-based relationships between environmental factors and CBCE carbon removal rates are currently being developed to facilitate assessments of the climate mitigation potential for ‘blue carbon farming’; for example, the Australian BlueCAM method, Nevertheless, it seems likely that many site specific measurements will still be needed, potentially involving >30 parameters, with implications for the cost-effectiveness of using coastal blue carbon primarily for climate mitigation purposes. Until such issues are better resolved, there is risk of optimism bias, or even greenwashing – with CBCE restoration serving as a “smokescreen for inaction”.
Indeed, a strong case can be made that, as for other nature-based solutions, coastal wetlands are inherently unsuitable for quantified carbon offsets. Yet the associated risks and uncertainties in that regard do not therefore mean CBCE restoration is a poor use of financial resources, nor do they provide any excuse for weak protection of such habitats. Instead, the socio-economic valuation of coastal habitats should mostly be based on the many other ecosystem services that they indisputably provide, such as coastal protection and biodiversity support.
Notes
- UNFCCC, United Nations Framework Convention on Climate Change. The annual meeting of UNFCCC signatories is held in November- early December and known as the Climate COP (Conference of the Parties). COP21 was hosted in Paris in 2015; COP26 in Glasgow, UK in 2021; and COP27 in Sharm el Sheikh, Egypt in 2022.
- The term ‘blue carbon’ can cover a range of meanings. A general definition is:
“All biologically-driven carbon fluxes and storage in marine systems that are amenable to management can be considered as blue carbon” (first part of glossary definition in IPCC Special Report on Ocean and Cryosphere in a Changing Climate). However, most usage in the scientific literature relates to three carbon-accumulating coastal ecosystems: saltmarshes, mangrove forests and seagrass meadows. The restoration of these habitats for the purpose of climate mitigation provides the main focus of this article. - A further complication is that the proportion of recalcitrant (very slowly decomposing) to labile (rapidly decomposing) organic carbon will increase with time since burial in the sediment. Estimates of the ratio of locally-derived to externally-derived carbon will therefore vary according to the depth of sediment that is analysed. For carbon-offsetting, it is the steady rate ratio at depth (that may not be achieved for 20-30 years) that is important, not the near-surface values.
References are given in the online version of this article, through hyperlinks.