Ocean Fertilization: Science, Policy, and Commerce

. Over the past 20 years there has been growing interest in the concept of fertilizing the ocean with iron to abate global warming. This interest was catalyzed by basic scientific experiments showing that iron limits primary production in certain regions of the ocean. The approach—considered a form of “geoengineering”—is to induce phytoplankton blooms through iron addition, with the goal of producing organic particles that sink to the deep ocean, sequestering carbon from the atmosphere. With the controversy surrounding the most recent scientific iron fertilization experiment in the Southern Ocean (LOHAFEX) and the ongoing discussion about restrictions on large-scale iron fertilization activities by the London Convention, the debate about the potential use of iron fertilization for geoengineering has never been more public or more pronounced. To help inform this debate, we present a synoptic view of the two-decade history of iron fertilization, from scientific experiments to commercial enterprises designed to trade credits for ocean fertilization on a developing carbon market. Throughout these two decades there has been a repeated cycle: Scientific experiments are followed by media and commercial interest and this triggers calls for caution and the need for more experiments. Over the years, some scientists have repeatedly pointed out that the idea is both unproven and potentially ecologically disruptive, and models have consistently shown that at the limit, the approach could not substantially change the trajectory of global warming. Yet, interest and investment in ocean fertilization as a climate mitigation strategy have only grown and intensified, fueling media reports that have misconstrued scientific results, and conflated scientific experimentation with geoengineering. We suggest that it is time to break this two-decade cycle, and argue that we know enough about ocean fertilization to say that it should not be considered further as a means to mitigate climate change. But, ocean


Ocean Fertilization
Science, Policy, and Commerce the ongoing discussion about restrictions on large-scale iron fertilization activities by the London Convention, the debate about the potential use of iron fertilization for geoengineering has never been more public or more pronounced. To help inform this debate, we present a synoptic view of the two-decade history of iron fertilization, from scientific experiments to commercial enterprises designed to trade credits for ocean fertilization on a developing carbon market. Throughout these two decades there has been a repeated cycle: Scientific experiments are followed by media and commercial interest and this triggers calls for caution and the need for more experiments. Over the years, some scientists have repeatedly pointed out that the idea is both unproven and potentially ecologically disruptive, and models have consistently shown that at the limit, the approach could not substantially change the trajectory of global warming. Yet, interest and investment in ocean fertilization as a climate mitigation strategy have only grown and intensified, fueling media reports that have misconstrued scientific results, and conflated scientific experimentation with geoengineering. We suggest that it is time to break this two-decade cycle, and argue that we know enough about ocean fertilization to say that it should not be considered further as a means to mitigate climate change. But, ocean fertilization research should not be halted: if used appropriately and applied to testable hypotheses, it is a powerful research tool for understanding the responses of ocean ecosystems in the context of climate change.
Oceanography Vol.22,No.3 236 This article has been published in Oceanography, Volume 22, Number 3, a quarterly journal of The Oceanography Society. © 2009 by The Oceanography Society. All rights reserved. Permission is granted to copy this article for use in teaching and research. republication, systemmatic reproduction, or collective redistirbution of any portion of this article by photocopy machine, reposting, or other means is permitted only with the approval of The Oceanography Society. Send all correspondence to: info@tos.org or Th e Oceanography Society, PO Box 1931, rockville, md 20849-1931 1 "LOhA" is hindi for "iron," FeX stands for "Fertilization eXperiment." seeking to carry out large-scale ocean fertilization activities to sell carbon-offset credits in a carbon trading market. These ventures have been a cause of concern for some scientists and several environmental NGOs, who have argued that claims of significant carbon sequestration are unsupported by scientific evidence, and that large-scale iron fertilization will, by design, profoundly alter marine ecosystems (Chisholm et al., 2001;Gnanadesikan et al., 2003;Cullen and Boyd, 2008;Denman, 2008;ETC Group News Release, 2009;World Wildlife Fund International, 2009). These concerns have helped spur recent UN resolutions, which were intended to restrict iron fertilization activities to small-scale scientific research (UN CBD, 2008;London Convention Meeting Report, 2008). The story was featured in Wired magazine (Keim, 2009) and on Reuters (Szabo, 2009) and the BBC (Morgan, 2009). A blog headline covering the controversy posed the question: "LOHAFEX-If you mean well, are you allowed to screw up the oceans?" (Campbell, 2009) press, including news articles in Science, consistently referred to LOHAFEX as a "geoengineering project, " an experiment designed to test the potential of ocean iron fertilization to change global climate (Kintisch, 2009).
The LOHAFEX scientists, however, defended their experiment as purely scientific and consistent with relevant UN regulations. And, the Director of the Alfred Wegener Institute (AWI), which sponsored LOHAFEX, defended the scientific validity of the research, stating that they "neither plan to nor want to smooth the way for a commercial use of iron fertilization with our expedition, " and that they "oppose iron fertilization with the aim to reduce CO 2 to regulate the climate" (AWI News, 2009) gineering are beginning to emerge (Latham et al., 2008;Rasch et al., 2008).

hiStOry OF PuBLiC-SeCtOr SCieNtiFiC OCeAN
FertiLizAtiON eXPerimeNtS the "iron hypothesis" Figure 2. relationship between Fe and CO 2 concentrations from the Vostok ice cores. After Martin (1990b) Martin famously quipped, "Give me half a tanker of iron, and I'll give you an ice age" (Martin, 1990a (Figure 3), it could explain the observed relationship between atmospheric iron dust deposition and atmospheric CO 2 during the last glacial maximum (Martin, 1990b).
He also commented that this "paleoiron" hypothesis could be important, because intentional oceanic iron fertilization could prove an effective method of drawing down atmospheric CO 2 "should the need arise" (Martin et al., 1990; see Box 1).  (Martin et al., 1989) and the HNLC Southern Ocean (Martin et al., 1990) Figure 4). The results demonstrated a phytoplankton bloom in response to iron addition, but they were confounded when the fertilized patch was subducted under low-density water (Martin et al., 1994).
Thus, several hypotheses remained untested (Cullen, 1995 (Coale et al., 1996). The authors suggested that the logical next step was to conduct an experiment in the Southern Ocean as this is "where most of the HNLC waters are found and where paleoclimate coherence between iron flux and carbon export has been observed" (Coale et al., 1996).  Boyd et al., 2000;Smetacek, 2001), and placed more emphasis on longerduration experiments tracking particle export, remineralization, and changes in zooplankton communities. Both experiments confirmed the hypothesis of iron limitation of primary production in the HNLC Southern Ocean. Although diatom production increased in response to iron addition in the SOIREE patch, carbon export did not (Boyd et al., 2000).
EisenEx also demonstrated a diatom bloom, and measured a larger net atmospheric CO 2 drawdown than SOIREE, but storms interrupted the experiment and the fate of fixed carbon could not be tracked (Assmy et al., 2007). A year later, SEEDS-I (Subarctic Pacific Iron Experiment for Ecosystem Dynamics Study) confirmed that productivity was limited by iron in the HNLC region of the western subarctic Pacific, Figure 3. Carbon dioxide that would otherwise be in the atmosphere is stored in the deep sea because the biological pump puts and keeps it there. Phytoplankton in the lighted surface layer take up nutrients (e.g., nitrate and phosphate) and grow, converting CO 2 to organic matter that fuels marine food webs. Some of the organic matter-for example, senescent phytoplankton, fecal pellets, and aggregated debrissinks to the deep ocean where it decomposes, releasing CO 2 and nutrients while consuming oxygen. When the ocean carbon cycle is roughly in balance, this carbon-and nutrient-rich deep water does not reach the surface for decades to hundreds of years, and when it does, biological productivity consumes the CO 2 and nutrients and sends C, N, and P back to deep waters as sinking organic matter. The amount of CO 2 thus stored in the deep sea largely corresponds to the amount of major nutrients (N and P) consumed in the lighted surface layer of the ocean. When iron limits productivity, N and P persist where they wouldn't otherwise; if iron limitation is alleviated, major nutrients are consumed, more organic matter is produced, and more carbon sinks to the deep sea. This extra carbon associated with added iron (either natural or intentional) could be considered sequestration. But the amount and duration of carbon sequestration depends on how deep the organic matter sinks before it is decomposed and whether or not iron is still available in excess when carbon-and nutrient-enriched waters reach the surface again. Figure modified from Chisholm (2000) documenting a floristic shift toward diatom production. Carbon export was not measured .  (Coale et al., 2004).
SOFeX produced the first conclusive measurement of enhanced particulate organic carbon (POC) export resulting from an intentional iron-fertilizationinduced bloom . Although POC export from the iron-enriched patch was elevated, the incremental flux was small compared to those observed during a natural iron fertilization event in the same location in 1998 (Buesseler et al., 2001). The authors concluded that the observed carbon flux was small relative to what would be required by proposed geoengineering plans to sequester carbon using iron fertilization, but they stressed that they had not been able to observe the termination of the iron-induced bloom .
The Subarctic Ecosystem Response to Iron Enrichment Study (SERIES) expedition followed. Conducted over a month in the subarctic Pacific, a 77-km 2 iron-enriched patch was created and POC export flux monitored (Boyd et al., 2004). The decline and termination of the iron-induced bloom (caused by silicate limitation) was observed.
The majority of the carbon fixed in SERIES was remineralized by bacteria and zooplankton grazing in the surface waters. Only a small fraction (8%) of the fixed carbon sank below the 120-m permanent pycnocline, significantly lower than the deep-export rate observed in natural blooms (Buesseler, 1998).
Further, the iron content (Fe:C) of the exported material was a thousandfold higher than that assumed in assessments of the efficiency and cost of OIF for climate mitigation. Noting increased attention to the idea of iron fertilization for geoengineering, Boyd et al. (2004) argued that "inefficient vertical transfer of carbon may limit the effectiveness of iron fertilization as a mitigation strategy. " The European Iron Fertilization Experiment (EIFEX), conducted in 2004 and the longest iron fertilization experiment to date, was also designed to evaluate the carbon export response and community shifts in a Southern Ocean iron-induced bloom (Hoffmann et al., 2006). Biomass export resulting from the sinking of an iron-induced bloom represented the highest ratio of carbon exported to added iron to date (Jacquet et al., 2008). At the same time, SEEDS-II, a second subarctic Pacific experiment, detected no significant bloom response to iron enrichment (Tsuda et al., 2007).
Box 1. The Paleo-Climate Portion of the iron hypothesis: Still an Open Question recently, some paleoceanographers have called into question the causal link between atmospheric dust deposition and lower atmospheric CO 2 (and thus a cooler climate) during the last glacial maximum. Kohfeld et al. (2005) analyzed the role of the biological pump in glacial CO 2 drawdown using sediment records, for example. Their data indicate that in large portions of the Southern Ocean, export productivity was actually lower during the last glacial maximum-exactly during a period of increased dust flux. They thus argued that iron fertilization from dust could not have been solely responsible for CO 2 drawdown (Kohfeld et al., 2005).
using these productivity data and those of others (Paytan et al., 1996;Anderson et al., 2008) and comparing them to dust flux records from Winckler et al. (2008), Anderson et al. (2007) presented an argument that there is no correlation in the paleoceanographic data between dust flux and increased export productivity in the equatorial Pacific or the Southern Ocean. While their data do show a strong anti-correlation between dust flux and CO 2 , they argue that there is no evidence that the dust caused the CO 2 drawdown. The causality inferred from the original ice core data ( Figure 2) has been a central thread in the argument for OiF for geoengineering, and, at the very least, these recent arguments questioning that causality deserve more attention and research (Anderson et al., 2007). Alternate hypotheses include strong influences of changing wind patterns on the overturning of carbon-rich southern deep water (toggweiler et al., 2006). Earth's albedo (Charlson et al., 1987).
In this experiment, scientists fertilized serially with over 5 tonnes of iron sulfate, but found only a modest chlorophyll increase and no increase in either CO 2 drawdown or DMS production, which they believed may have been due to light limitation (Law et al., 2006). Another iron-enrichment experiment, FeeP, performed a combined Fe and phosphate addition to low nutrient, low chloro-

phyll (LNLC) waters in the Northeast
Atlantic to test the hypothesis that iron enrichment of LNLC waters can lead to a net N import (ultimately supporting increased productivity) by stimulating nitrogen fixation. Although rates of nitrogen fixation increased in response to enrichment, productivity did not.
Carbon export in response to enrichment was not measured (Rees et al., 2007;Karl and Letelier, 2008     high potential for methane production as a result of this anoxia (Sarmiento and Orr, 1991; see also Fuhrman and Capone, 1991). Of essential importance is that these results, by design, represent the extreme (unrealistic) case scenariofertilizing the entire Southern Ocean with iron for 100 years, and assuming that all of the macronutrients available were completely used up. Thus, these results represent an unachievable (both logistically and ecologically) upper limit.
Since these initial modeling efforts, there have been many published variations on the theme (Table 2) including a long-term reduction in ocean productivity, alteration of the structure of marine food webs, and a more rapid increase in ocean acidity (Denman, 2008 Boyd et al., 2007;Powell, 2007Powell, -2008 and began a more public discussion about where the field was, and should be, going (Boyd, 2008;Buesseler et al., 2008;Lampitt et al., 2008;Smetacek and Naqvi, 2008 disruption and other potential downstream negative side effects were also recognized as a cause for concern. In an outgrowth of the WHOI workshop and in an effort to move the field forward, Buesseler et al. (2008) focused on the need to reduce uncertainties about OIF as a climate mitigation strategy, arguing that there is "as yet, no scientific evidence for issuing carbon credits from OIF. " They urged a move to larger and longer experiments because "ecological impacts and CO 2 mitigation are scale-dependent" (Buesseler et al., 2008 warming" (Monastersky, 1994), and another account reported that "the idea of fertilizing the entire Southern Ocean should probably be considered dead" (Kunzig, 1994 (Markels, 1995). The method referred to fertilizing with "all nutrients that are found to limit production in the surface ocean. " In this patent, Markels cited the results of the IronEx I experiment and suggested that he could constantly fertilize 140,000 km 2 of the Gulf Stream (which is not an HNLC region) with enough iron, phosphate, and micronutrients to remove 1.3 Gt of CO 2 and produce 50 Mt of additional seafood production annually (Markels, 1995 "really more of a business experiment" (Williams, 2003). Although the results of the experiment were never made public, a description of the activity was published in the news feature section of Nature the following year (Schiermeier, 2003), including a photo of George on Ragland. The Planktos Foundation, and the entire idea of carbon credit sale from ocean fertilization, was gaining attention and momentum.

Legal issues Arise
As it became evident that commercial operations were moving forward, concerned scientists and environmental groups continued to raise questions about the efficacy and legality of commercial ocean fertilization. There appeared to be no law preventing ocean fertilization beyond the 200-mile exclusive economic zone of any country (Markels and Barber, 2001;McKie, 2003 hoW It ReLAteS to oIF: Several general provisions of uNCLOS are relevant to OiF. Article 145 stipulates that use of the high seas for marine scientific research must be for peaceful purposes geared toward the increase of knowledge and understanding for all humankind. Various other articles outline requirements of states to protect life of the marine environment. uNCLOS is also relevant to OiF for fish production, as it governs both the delimitation of fishing rights within exclusive economic zones (eezs) and fishing rights on the high seas.

Key StAtementS And deCISIonS on oIF:
Dec 2007: General Assembly resolution calls for more research into the effects of ocean iron fertilization. dec 2008: general Assembly resolution welcomes London Convention and Convention on Biological diversity decisions against large-scale OiF.
FutuRe ReguLAtIonS: Although the general Assembly has not yet passed specific regulatory resolutions regarding OiF, the centrality of uNCLOS to maritime regulation, and its relevancy to environmental protection, scientific research regulation, and international waters legal issues, could make it a good forum for coordinated regulation of iron fertilization in the future. hoW It ReLAteS to oIF: Because iron fertilization activities require the addition of inorganic material in the hNLC regions of the world ocean, and generally take place in international waters, the London Convention is deemed relevant. This relevancy has been the subject of some discussion, however, as the London Convention regulates dumping for disposal of waste. Because OiF for carbon sequestration would sequester carbon in the deep ocean, the discussion of whether OiF is more appropriately the dumping of carbon has also been raised.

Key StAtementS And deCISIonS on oIF:
Jun 2007

FutuRe ReguLAtIon:
The London Convention is clearly grappling with the difficult issue of regulating scientific research at the supranational level. At the same time, there is a clear push toward further restriction of nonscience OiF activities. Specific proposals from the February 2009 working groups range from outright bans of all nonscience OiF activities to "suspensions" of activity until more data can be gathered. The questions of commercialization, ecological assessment, and degree of enforcement/intensity of regulation have all been raised but none of them is resolved. The Scientific group have started to discuss the questions about how much of an environmental impact from scientific research is too much and will present a report to the governing bodies of the convention when they meet in London in the fall of 2009. biogeochemical modeling (Chu et al., 2003;Rice, 2003). Meanwhile, Dan Whaley, a former information technology entrepreneur, founded Climos Inc., a for-profit company setting out to pursue the same goals as Planktos Inc.: selling carbon credits from ocean iron fertilization (Riddell, 2008 • Dual purposes of biodiversity conservation and sustainable use and equitable distribution of resources.
• All UN nations are party, except the United States, which has signed but not ratified the treaty.
• The Convention is administered by the UN Environment Programme (UNEP).
• Enforcement by individual ministries of member states.
hoW It ReLAteS to oIF: The Convention regulates actions that threaten biodiversity, including marine biodiversity. in this way, the unknown ecological effects of large-scale OIF implementation fall under the purview of the Convention.

Key StAtementS And deCISIonS on oIF:
may 2008 CBd decision: "All large-scale OiF activities should not be allowed." • Makes exception for "small scale studies in coastal waters." Coastal waters may be an aberration.
• Decision explicitly mentions commercial interests as a reason not to allow OIF.
• Decision passed because of ecological risks of OIF and uproar over Planktos' experimental plans.
• Widely viewed as a "UN moratorium" on commercial OIF activities.

FutuRe ReguLAtIonS:
The CBd suggested that OiF regulation be done in coordination with the international maritime Organisation (imO) and London Convention. it calls for a global transparent control and regulatory mechanism to be established, and urges consultation with all parties involved to establish a knowledge base about the associated ecological risks.
note on CBd enFoRCement: in early 2009, the german research ministry cited the uN CBd COP 9 moratorium on ocean fertilization in its decision to suspend the indo-german LOhAFeX OiF experiment. it demanded more environmental risk assessments and independent scientific assessments of the project, specifically mentioning the coastal water stipulation and citing the ecological concerns raised by CBd. Although the experiment was eventually given the green light, it was a first test of country-specific enforcement of international treaties on this issue and has informed the London Convention working groups' discussions on how best to regulate the science of OiF. experiment was to be on a previously unachieved scale of 10,000 km 2 ; it would be the first pilot project in a planned "Voyage of Recovery" (see Figure 5 for a sense of the relative size of this experiment  Assessment, 2007), which seemed to contradict that Weatherbird II was a US-registered vessel and would set out from a US port for the experiment. Planktos' experiment did not take place (Courtland, 2008). Citing concerns over disruption of their activities, Weatherbird II left port headed to an undisclosed location in the Atlantic   Wire, 2008;Kerry, 2008), citing both a lack of funds and a "highly effective disinformation campaign waged by anti-offset crusaders" (Brown, 2008).

Several months later in June 2008, Russ
George started a new iron fertilization "ecorestoration" company, named Planktos-Science, though there has been little activity other than Web posts from this company thus far (see http://www. planktos-science.com/).

Climos' Path Forward
These events left Climos as the leading company involved in the nascent commercial iron fertilization business.  (Climos FAQ, 2008). In a statement to the press, Climos officials indicated that they hope to conduct their first trial by the end of 2009 and to be able to start selling carbon credits shortly thereafter (Murray, 2008

uncertainty and risk
While the uncertainties about the efficacy of carbon sequestration from OIF as a geoengineering proposal are high (Buesseler et al., 2008), the certainty of ecological disruption is also high. OIF for carbon sequestration is  and nitrogen (Denman, 2008). This would result in longer-term and far-field changes in nitrous oxide production that could potentially offset significant amounts of predicted green-house gas benefits of OIF (Law, 2008 (Cullen and Boyd, 2008).

Long-term Ocean Carbon Sequestration from iron Fertilization is Not Verifiable
These complex downstream responses to ocean fertilization make verification of net greenhouse gas reduction through fertilization next to impossible (Chisholm et al., 2001;Cullen and Boyd, 2008;Gnanadesikan and Marinov, 2008).
Furthermore, carbon export measured as a result of a fertilization-induced bloom would have to be referenced to a baseline rate of carbon export. As ocean fertilization and carbon flux research has shown, this natural rate of carbon export is highly variable in space and time; establishing an appropriate baseline to grant carbon credits for individual applications would be exceedingly difficult, and rife with uncertainty ( Figure 6).
At present, there is no system under the Kyoto Protocol's Clean Development Mechanism to provide for carbon credits from offsets by marine carbon sequestration (Powell, 2007(Powell, -2008.
Thus, under the current international mechanism, any credits granted would therefore have to be sold on the currently unregulated "voluntary carbon credit market. " Although Climos Inc. submitted a carbon sequestration methodology to Det Norske Veritas, an international verification company, in late 2007 (Climos Press, 2007), official approval or verification has not been given to OIF as a carbon sequestration methodology. Nonetheless, in the future, international carbon credit regulatory systems may well include provisions for marine "sequestration" offsets (Powell, 2007(Powell, -2008 Ocean fertilization will not solve the CO 2 problem, and if implemented for profit, regardless of scale, has the potential to change the nature of the ocean through the "tragedy of the commons" (Hardin, 1968

moving On
Climate change is already upon us.
Society needs to know what the ocean will be like in a high CO 2 world-which ecosystems will be at risk as the ocean warms and acidifies, and how the altered ocean will in turn influence climate.
Understanding how ocean biogeochemical cycles are linked, and the processes that drive these cycles, is essential for climate prediction. Transformational developments in genomics along with rapid advances in ocean observations and modeling (Doney et al., 2004) allow us to study the ocean as a system on scales from molecules to ocean basins, and we are poised to fully integrate studies of evolution and biogeochemistry (Woese and Goldenfeld, 2009