Introduction

Seagrass meadows are key coastal ecosystems that maintain the well-being of people and the planet, providing a range of ecological, social, and economic services1. These ecosystems provide critical habitat for marine biodiversity, stabilize sediments, cycle nutrients, and contribute substantially to climate regulation through their ability to sequester and store organic carbon2. The capacity of seagrass meadows to act as carbon sinks has gathered increasing attention, particularly within the framework of blue carbon, the carbon removed and stored as organic carbon by coastal and marine ecosystems3. In recent years, much of the focus around seagrass blue carbon has been placed on soil organic carbon stocks and accumulation rates, whereas carbon stores in seagrass above- and below-ground biomass, and seagrass net primary production (NPP) have often not been included, potentially leading to underestimates4. This oversight persists despite the critical importance of seagrass biomass and production in maintaining carbon fluxes and stabilizing sedimentary carbon stocks5.

The capacity of seagrass to act as carbon sinks is partially explained by their high NPP rates, which rank amongst the highest within marine ecosystems and are comparable to those of terrestrial forests and other highly efficient photosynthetic systems like seaweed forests, mangroves, and saltmarshes6. Environmental factors such as temperature, rainfall patterns, irradiance, water depth, and sediment characteristics play key roles in shaping the distribution of seagrass meadows globally together with their biomass and NPP rates7. These drivers vary across seagrass bioregions8 and interact with seagrass life-history strategies to shape the carbon storage potential and ecological functioning of seagrass meadows9. Understanding how these factors influence biomass carbon stocks and seagrass NPP is essential for predicting the responses of seagrass ecosystems to global environmental change.

Life-history strategies in seagrasses—defined as the set of traits that species exhibit to thrive in their environment—can be categorized into persistent, opportunistic, and colonizing10. Persistent meadows, often dominated by genera such as Posidonia and Thalassia, are characterized by large, slow-growing plants with extensive rhizome systems that promote meadow stability and longevity, making them highly resilient carbon sinks11,12. In contrast, opportunistic and colonizing genera like Zostera and Halophila, respectively, are smaller in size, exhibit rapid growth, and have higher turnover rates, allowing them to quickly establish in bare or disturbed areas. However, colonizing meadows are even more ephemeral than opportunistic meadows, with faster turnover, lower persistence, and reduced resistance to disturbances, but high recovery capacity10.

Bioregions and specific climate conditions further modulate the nature and dynamics of seagrass meadows, influencing their productivity and temporal patterns. In temperate regions, species often experience marked seasonal variability in biomass and NPP, with slower metabolism and growth during autumn and winter due to lower temperatures and light availability7,13. Conversely, tropical seagrass meadows can maintain relatively high biomass and NPP rates year-round but can exhibit periods of complete dieback in response to extreme climatic events, including prolonged heatwaves14,15. The interaction between life-history strategies and environmental dynamics underscores potential differences in the relative contribution of different seagrass ecosystems to the carbon cycle.

Seagrass biomass plays a dual role in carbon cycling: it serves as a transient but dynamic carbon reservoir, but also contributes organic matter to sediments, bolstering the long-term carbon storage in the soil compartment16. The turnover time of seagrass biomass ranges from weeks to years, depending on species and environmental conditions. Seagrass biomass constitutes a dynamic carbon sink that can persist in a near-steady state over annual, decadal, or millennial time scales17. However, the loss of seagrass canopy due to either natural or anthropogenic disturbances can disrupt this cycle as well as the role of seagrass ecosystems as carbon sinks and reservoirs by reducing organic matter inputs to the soil carbon pool and exposing the soil carbon reservoir to erosion, likely resulting in carbon emissions5. These disruptions are particularly concerning given the global decline of seagrass meadows over the past century, driven by threats, such as coastal development, pollution, eutrophication, and climate change18.

The rapid advancement of blue carbon science highlights the urgent need for equitable and effective approaches to implement blue carbon strategies that can yield multiple benefits for people and the planet. Although debate around elements of the blue carbon framework remain19,20,21,22, most peer-reviewed studies endorse the role of these dynamic carbon pools in climate mitigation16,23,24. Indeed, the Intergovernmental Panel on Climate Change (IPCC) provided conversion factors that include seagrass biomass, and current carbon crediting methodologies, such as VERRA VM0033 and the Australian carbon methodology for Wetlands, include seagrass biomass in carbon accounting25,26,27. Their inclusion is contingent upon conservation and restoration actions that enhance carbon removals or avoided emissions by increasing or maintaining these pools in a steady state condition. However, these initiatives lack global datasets of biomass carbon stocks and NPP that could contribute baseline data to aid and enhance current estimates and methodologies, ultimately promoting the large-scale uptake of seagrass conservation and restoration actions under crediting mechanisms for climate change mitigation and adaptation.

This study aims to address these knowledge gaps by providing a global-scale assessment of carbon stocks in seagrass biomass, NPP, and emissions from seagrass biomass linked to habitat loss according to different life-history strategies and countries within the world’s seagrass bioregions. This research offers insights into the ecological contributions of seagrass to the global carbon cycle and lays the groundwork for their inclusion in carbon crediting schemes linked to conservation and restoration actions. The results compiled not only enhance the recognition of seagrass meadows as climate mitigation tools but also highlight their broader ecological significance as foundational coastal ecosystems.

Results

Carbon stocks in seagrass biomass and NPP across genera and life-history strategies

Seagrass biomass carbon stocks per unit area and NPP varied by up to six orders of magnitude, ranging from 0.03 to 25,340 kg C ha−1 (mean ± SE; 1551 ± 40 kg C ha−1), and from 2.82 to 51,456 kg C ha−1 yr−1 (6696 ± 557 kg C ha−1 yr−1), respectively. Biomass and NPP refer to total biomass and total NPP (i.e., above plus below-ground) unless stated otherwise. Clear patterns in seagrass biomass carbon stocks and NPP were observed among seagrass genera, life-history strategies, and bioregions. Relatively large, long-living seagrass genera that typify persistent meadows (Posidonia, Enhalus, Amphibolis, Thalassia, Thalassodendron) held up to 4–fold higher carbon stocks in biomass (3699 ± 179 kg C ha−1) than opportunistic genera (Syringodium, Zostera, Cymodocea, and Phyllospadix), and up to 11–fold higher carbon stocks than colonizing (Halodule, Ruppia, and Halophila) genera (944 ± 26 and 327 ± 17 kg C ha−1; p-value < 0.001; Fig. 1 and Supplementary Table 1). Carbon stocks in above- and below-ground biomass followed a similar pattern among life-history strategies, with higher stores in persistent meadows than in opportunistic and colonizing meadows (p-value < 0.001; Supplementary Fig. 1 and Supplementary Table 1). Notably, Amphibolis, Phyllospadix, and Posidonia had the highest above-ground biomass, whereas Posidonia had the largest below-ground biomass. Both above- and below-ground biomass were lowest in Halophila, Halodule, Ruppia, and Syringodium (p-value < 0.001; Supplementary Fig. 2 and Supplementary Table 2). Among the 12 seagrass genera explored, Posidonia held the highest total biomass carbon stocks (9780 ± 877 kg C ha−1), being 50–fold higher than the genus Halophila, supporting the lowest carbon stocks in biomass (190 ± 10 kg C ha−1) (p-value < 0.001; Supplementary Fig. 2 and Supplementary Table 2).

Fig. 1: Total biomass carbon stocks per unit area and net primary production (NPP) among seagrass life-history strategies, genera and bioregions.
Fig. 1: Total biomass carbon stocks per unit area and net primary production (NPP) among seagrass life-history strategies, genera and bioregions.
Full size image

a Total carbon stocks in biomass (kg C ha−1) among seagrass life-history strategies. b Total NPP (kg C ha−1 yr−1) among seagrass life-history strategies. c Total carbon stocks in biomass among seagrass genera. d Total NPP among seagrass genera. e Seagrass total carbon stocks in biomass across bioregions. f Seagrass total NPP across bioregions. The horizontal black lines indicate medians, the box edges represent the interquartile range (25–75th percentiles), and whiskers show the range within 1.5 times the interquartile range. Black dots are outliers. Significant differences are indicated by different letters. Red dots represent the mean values, while crosses denote the absence of data. The number of observations (n) is provided below the box-and-whisker plots. The y-axes were 4th-root transformed for improved visualization of differences among life-history strategies.

On the contrary, seagrass NPP per unit area in persistent genera was 2–fold lower (3126 ± 205 kg C ha−1 yr−1) than in meadows of opportunistic and colonizing life-history strategies (10,276 ± 1096 and 7271 ± 716 kg C ha−1 yr−1, respectively; p-value < 0.001; Fig. 1 and Supplementary Table 1). Remarkably, Phyllospadix showed an exceptionally high NPP, reaching 32,986 ± 3248 kg C ha−1 yr−1, far exceeding the NPP observed in other genera, which ranged from 1380 ± 250 in Enhalus to 11,877 ± 701 kg C ha−1 yr−1 in Syringodium (p-value < 0.001; Fig. 1). Above-ground NPP varied up to three orders of magnitude, from 12 kg C ha−1 yr−1 in Thalassodendron meadows to 19,129 ± 2122 kg C ha−1 yr−1 in Phyllospadix. Below-ground NPP was lowest in Posidonia and Enhalus (49 ± 16 and 324 kg C ha−1 yr−1, respectively), and highest in Phyllospadix (13,612 ± 1290 kg C ha−1 yr−1), whereas other genera exhibited relatively similar below-ground NPP (ranging from 1047 to 3871 kg C ha−1 yr−1; p-value < 0.001, Supplementary Fig. 2).

Carbon stocks in seagrass biomass and NPP among bioregions

The Mediterranean bioregion holds the highest seagrass biomass carbon stock per unit area (2328 ± 189 kg C ha−1; p-value < 0.001), followed by the Tropical Indo Pacific and Atlantic bioregions (1420 ± 88 and 1346 ± 85 kg C ha−1, respectively). The Temperate Southern (1102 ± 97 kg C ha−1), North Pacific (736 ± 36 kg C ha−1), North Atlantic East (660 ± 33 kg C ha−1), and the North Atlantic West (594 ± 45 kg C ha−1) bioregions had lower carbon stocks in seagrass biomass (Fig. 1 and Supplementary Table 2). Carbon stocks in below-ground biomass followed the same pattern as in total biomass, whereas carbon stocks in above-ground biomass were lowest in the Tropical Indo Pacific and Atlantic and Temperate North Atlantic East bioregions (299 ± 17, 404 ± 26, and 372 ± 22 kg C ha−1, respectively) compared to Temperate North Atlantic West (418 ± 27 kg C ha−1 yr−1), Temperate Southern (551 ± 28 kg C ha−1 yr−1), Mediterranean (582 ± 36 kg C ha−1 yr−1), and the Temperate North Pacific (729 ± 38 kg C ha−1 yr−1) bioregions (p-value < 0.001; Supplementary Fig. 3 and Supplementary Table 2).

Conversely, the opposite pattern was observed for NPP, with the Mediterranean (2554 ± 323 kg C ha−1 yr−1), Tropical Atlantic (2843 ± 289 kg C ha−1 yr−1), Temperate North Atlantic West (4413 ± 1280 kg C ha−1 yr−1), and Tropical Indo Pacific (5050 ± 328 kg C ha−1 yr−1) bioregions showing from 2– to 3–fold lower NPP than Temperate Southern (11,917 ± 1403 kg C ha−1 yr−1), North Pacific (14,765 ± 26,769 kg Cha−1 yr−1) and North Atlantic East (15,858 ± 1490 kg C ha−1 yr−1) bioregions (p-value < 0.001; Fig. 1 and Supplementary Table 2). Above- and below-ground NPP exhibited similar patterns to total NPP across bioregions (Supplementary Fig. 3).

Global and national carbon inventories in seagrass biomass and NPP

On a global scale, we estimated the biomass carbon stocks in seagrass meadows at 24–40 Tg C, and NPP at 83–137 Tg C yr−1, distributed across an estimated extent of 160,387–266,500 km2. Total carbon stocks in biomass widely varied across countries, ranging from 0.014 Gg C in Slovenia to 12,718 Gg C in Australia (Fig. 2 and Supplementary Table 3). Similarly, total carbon stocks in biomass per unit area varied widely among countries, ranging from 20 ± 2 kg C ha−1 in Vietnam to 12,483 ± 1770 kg C ha−1 in Malta. We observed substantial variation in total seagrass NPP summed across the total spatial extent of seagrass meadows of countries, ranging from 0.013 Gg C yr−1 in Slovenia to 58,805 Gg C yr−1 in Australia (Fig. 2 and Supplementary Table 3). NPP per unit area varied up to two orders of magnitude among countries, ranging from 144 ± 35 kg C ha−1 yr−1 in Italy to 32,985 ± 3248 kg C ha−1 yr−1 in Mexico. Clear patterns in seagrass biomass carbon stocks and NPP per unit area among life-history strategies within countries were observed, with up to one order of magnitude higher carbon stocks and NPP in persistent compared to colonizing genera within the same country (Fig. 3 and Supplementary Table 3).

Fig. 2: Global seagrass biomass carbon stocks and seagrass net primary production (NPP) across countries.
Fig. 2: Global seagrass biomass carbon stocks and seagrass net primary production (NPP) across countries.
Full size image

a Global map of total carbon stocks in seagrass biomass per country. b Global map of annual net primary production (NPP) per country. Countries represented with stripes have low-confidence extent estimate7. Gray represents countries with no data, and white represents countries where seagrass is absent. Maps legends represent equal quantiles. Basemap outlines were obtained from Opendatasoft (https://public.opendatasoft.com/) under the Open Government Licence v3.0.

Fig. 3: Distribution of mean biomass carbon stocks per country among seagrass life-history strategies.
Fig. 3: Distribution of mean biomass carbon stocks per country among seagrass life-history strategies.
Full size image

Global maps of seagrass biomass carbon stocks per country and unit area categorized by life-history strategies: a persistent, b opportunistic and c colonizing. Countries with a white striped pattern represent those with relatively poor data quality (<4 observations). Countries shown in white indicate areas with no data or no seagrass presence of any life-history strategy. Maps legends represent equal quantiles. Seagrass icons extracted from IAN Media Library (https://ian.umces.edu/media-library), used under the Creative Commons Attribution-ShareAlike 4.0 International license (CC BY-SA 4.0). Basemap outlines were obtained from Opendatasoft (https://public.opendatasoft.com/) under the Open Government Licence v3.0.

Discussion

Patterns in seagrass biomass carbon stocks and NPP among genera, life-history strategies and bioregions

Seagrass carbon stocks and NPP differ markedly across seagrass genera, life-history strategies, countries, and bioregions. Persistent genera store carbon more efficiently due to their extensive below-ground biomass that can remain in the system for decades, whereas opportunistic and colonizing genera have shorter lifespans10, and therefore lower carbon storage potential. Long-lived, large, and persistent seagrass species within the genera Posidonia, Enhalus, and Thalassia predominate in bioregions like the Mediterranean and the Tropical Atlantic, contributing to higher biomass carbon stocks per unit area compared to bioregions like the Temperate North Atlantic and Pacific that are dominated by relatively small opportunistic (Zostera, Cymodocea) or colonizing (Halophila, Ruppia) genera. In contrast, although the Tropical Indo-Pacific bioregion hosts vast seagrass meadows, the prevalence of colonizing and opportunistic genera resulted in lower above-ground biomass carbon stock per unit area. These patterns underline the strong influence of life-history strategies on carbon dynamics in seagrass biomass.

NPP patterns are also related to the different life-history strategies present within each bioregion. While persistent genera such as Posidonia and Enhalus have relatively lower NPP per unit area (3126 ± 205 kg C ha−1 yr−1), their stability ensures sustained contributions to the carbon cycle and the maintenance of the biomass carbon pool over time. In contrast, the significantly higher total NPP in opportunistic (e.g., Zostera or Syringodium) and colonizing (e.g., Halophila or Ruppia) genera (10,276 ± 1096 and 7271 ± 716 kg C ha−1 yr−1, respectively) is likely driven by higher photosynthetic rates, rapid above-ground growth, but also high turnover7. As a result, their potential to contribute to carbon storage pathways beyond the seagrass meadow boundaries (e.g., storage of particulate and dissolved carbon in coastal and deep-sea regions) is likely higher than in persistent meadows28. Bioregional patterns could further emphasize this dichotomy. The Mediterranean and Tropical Atlantic bioregions, dominated by persistent genera (e.g., Posidonia and Thalassia), exhibit lower total NPP compared to regions like the North Pacific and Temperate Southern, where dynamic opportunistic and colonizing species contribute to higher NPP in the latter. For example, the Mediterranean bioregion exhibits the highest biomass carbon stock per unit area (2328 ± 189 kg C ha−1) but lower NPP (2554 ± 323 kg C ha−1 yr−1), compared to the Temperate North Atlantic East bioregion that holds moderate biomass carbon stocks but exhibits exceptionally high NPP (15,858 ± 1490 kg C ha−1 yr−1). These patterns highlight the important role of opportunistic and colonizing species in the carbon cycle, despite their being transient and dynamic carbon reservoirs.

The fact that annual seagrass NPP is more than 3-fold higher than the standing biomass carbon stock showcases the large carbon flux from seagrass ecosystems28,29,30. Seagrasses contribute significantly to the carbon dynamics of adjacent marine and terrestrial ecosystems through the export of organic carbon through various pathways, including export of particulate organic carbon (POC) and dissolved organic carbon (DOC), which can be advected across substantial distances in coastal waters31,32,33. Studies estimate that about 30–50% of the total NPP by seagrasses may be exported and stored in adjacent habitats such as mangroves, coral reefs, and even the open ocean, while 7–26% will be consumed by fauna associated to seagrass ecosystems, highlighting their role as critical nodes in blue carbon networks28,30,34. Additionally, about 20% of the NPP will be buried in the seagrass sediment, contributing to long-term sequestration, with burial efficiencies varying based on environmental and geomorphological conditions16,28. This dual role of high productivity and efficient export emphasizes the pivotal importance of seagrasses in carbon cycling and storage.

Global perspective on seagrass biomass carbon stocks and NPP

When scaled globally, the total biomass carbon stock in seagrasses amounts to approximately 24–40 Tg C, with an annual NPP of 83–137 Tg C yr−1. Per unit area, seagrass NPP (6696 ± 557 kg C ha−1 yr−1) is comparable to that of tropical rainforests (ranging from 6000–10,000 kg C ha−1 yr−1) and exceeds terrestrial grasslands (2000–4000 kg C ha−1 yr−1)35,36. Compared to other blue carbon ecosystems, seagrass NPP is similar or slightly lower than in mangroves and tidal marshes (6000–9000 kg C ha−1 yr−1)37,38. However, the biomass carbon stocks are substantially lower for seagrasses (1551 ± 40 kg C ha−1) compared to mangroves (15,000–30,000 kg C ha⁻¹) and salt marshes (8000–12,000 kg C ha−1) due to their smaller above-ground biomass and shorter-lived plants23,39. Seagrasses also differ significantly from seaweeds in carbon cycling. While seaweeds exhibit higher NPP per unit area (up to 20,000 kg C ha−1 yr−1 in some kelp forests), their carbon storage is transient, as they lack the below-ground biomass necessary for long-term sequestration40. These comparisons highlight the substantial contribution of seagrass ecosystems to the global carbon cycle, albeit with significant variability across bioregions and countries.

Globally, seagrass soil carbon stocks per unit area in the top 20 cm have been estimated at (mean ± SE) 22,000 ± 800 kg C ha−19, while in this work we estimated that the biomass carbon stock contributes approximately an extra 1551 ± 40 kg C ha−1. This indicates that seagrass biomass accounts for roughly 7% of the carbon stocks within seagrass ecosystems. However, this percentage would differ when considering the range of soil thicknesses (from 0 m when seagrasses grow on top of rock to more than 10 m of below-ground material in P. oceanica) reported for seagrass ecosystems41. Despite its small proportion, the seagrass biomass is critical in maintaining the integrity of seagrass soil carbon stocks through the input of new carbon, but also by preventing the erosion of the soil and by enhancing the deposition of allochthonous carbon28. The above-ground and below-ground biomass act as natural barriers against hydrodynamic forces, reducing sediment resuspension and enhancing deposition42. In contrast, the loss of biomass accelerates sediment erosion, which can lead to the release of ancient, sequestered carbon back into the atmosphere17,43,44. Therefore, the role of biomass in carbon sequestration goes beyond the standing stocks, with the protection of soil carbon from erosion constituting an inherent function that can result in avoided emissions from soil carbon in denuded soils after meadow loss.

Global biomass carbon emissions derived from seagrass meadow loss

Based on previous estimates of seagrass loss trends over the past century calculated for each country18, and the mean country-specific seagrass total biomass obtained in this study, we estimated annual global emissions from seagrass loss at 154–256 Gg CO2-eq yr−1 (based on high (160,387 km2) and low (266,500 km2) confidence world extent estimates, respectively7). These estimates assumed that 90% of the biomass would ultimately be mineralized following seagrass loss (see Methods). Although the focus on emissions associated with seagrass loss had mostly been on the carbon stocks in the denuded sediments, the results obtained in this study underscore the potential contribution of seagrass biomass losses to carbon emissions globally.

Annual global emissions from seagrass biomass loss exhibit substantial variations across bioregions, showcasing which countries could benefit the most from conservation and restoration actions to avoid further losses. The highest emissions were estimated for the Tropical Atlantic and the Tropical Indo Pacific bioregions (70.4 and 64.5 Gg CO2-eq yr−1, respectively; Fig. 4). Countries within these regions (e.g., Australia, Cuba, and Indonesia) could avoid further losses and benefit from carbon and emerging nature-crediting schemes through the management of eutrophication, coastal development, and trawling fishing threats18. Emissions from the Mediterranean and Temperate Southern (10 and 7.4 Gg CO2-eq, respectively) bioregions were up to 10-fold lower, likely linked to climate change threats in the occidental Mediterranean Basin and the Temperate Southern44,45, and threats linked to coastal development and eutrophication in the oriental Mediterranean Basin. In contrast, although net seagrass loss rates reported for the Temperate North Atlantic East bioregion are high (~69%)18, the emissions from seagrass biomass in this bioregion were relatively low (2.6 Gg CO2-eq, Fig. 4) owing to the limited biomass carbon storage capacity of the dominant seagrass genera in this region (i.e., opportunistic Zostera spp.) and the limited extent compared to other bioregions. Emissions within the Temperate North Pacific and North Atlantic West bioregions were negligible due to the absence of net seagrass meadow loss in these areas18.

Fig. 4: Distribution of annual carbon emissions (CO2 equivalents) only from seagrass biomass linked to habitat loss per country.
Fig. 4: Distribution of annual carbon emissions (CO2 equivalents) only from seagrass biomass linked to habitat loss per country.
Full size image

a Global map annual emissions (CO2 equivalents) from seagrass biomass and seagrass habitat loss per country. Countries represented with stripes have low-confidence extent estimates7. Gray represents countries with no data, and white represents countries where seagrass is absent. Legends represent equal quantiles. b Scatter plot illustrating the relationship between seagrass extent and seagrass biomass annual carbon emissions. The x- and y-axes were 4th-root transformed for improved visualization of differences among countries. c Bar plot showing seagrass biomass annual carbon emissions per bioregion. Basemap outlines were obtained from Opendatasoft (https://public.opendatasoft.com/) under the Open Government Licence v3.0.

The top five countries that experienced the highest potential of CO2-eq emissions per year due to seagrass biomass loss were Australia (83.7 Gg yr−1), Spain (14.4 Gg yr−1), Mexico (13.6 Gg yr−1), Italy (12.3 Gg yr−1), and the United States (7.4 Gg yr−1) (Fig. 4 and Supplementary Table 3). Together these countries account for ∼82% of the total potential emissions with seagrass biomass loss, with Australia alone accounting for 52%. These findings suggest the emissions are overwhelmingly a function of the area of seagrass (82% explained) and therefore, the rate or extent of loss had a relatively minor relevance in driving emissions compared to total area. Thus, implying that limited seagrass loss in countries with large seagrass extent (i.e., Australia, Mexico and the United States) are currently driving the bulk of global CO2-eq emissions from seagrass biomass loss (Fig. 4). Indeed, among these key contributors, countries like Spain, Italy and France host significant biomass carbon stocks per unit area (>4000 kg C ha−1), yet they experience relatively moderate rates of seagrass loss (Fig. 4). Seagrass conservation efforts targeting the highest potential sources of emissions would likely yield the most significant benefits for mitigating CO2-eq emissions.

Role of seagrass biomass carbon stocks in carbon crediting frameworks

Blue carbon has been defined as the carbon captured by marine coastal ecosystems3 that can remain undecomposed and unmineralized for long periods of time (up to several thousand years)46. Given that seagrass blue carbon research efforts have largely focused on soil storage16,47, this study provides the global, regional, and national data needed to include living biomass as blue carbon in carbon crediting frameworks, as is the case in VERRA and the Australian carbon methodology for Wetlands25,26,27, to fully capture their potential for climate change mitigation and adaptation. The dynamic nature of seagrass biomass necessitates adaptive management strategies that prioritize both conservation and restoration to maintain meadow resilience and carbon sequestration potential. Although seagrass biomass is a dynamic pool with turnover rates ranging from months to years, conservation actions contributing to the maintenance of the meadows in a steady state condition, or restoration actions generating new biomass, can contribute to carbon sequestration. Yet, similar to biomass in terrestrial forests and other carbon accounting methodologies, it seems difficult to maintain this steady state condition for >100 years, but other approaches allow shorter permanence periods (often at a discounted crediting rate)19, which may be more appropriate for opportunistic and colonizing species.

The life-history-driven variability in biomass carbon stocks emphasizes the need for region-specific conservation strategies that account for the ecological characteristics of local seagrasses and the need to account for different life-history strategies in carbon crediting accounting. Specifically, the biomass of persistent and long-lived seagrasses should be included in carbon accounting if conservation and restoration actions contribute to maintaining these carbon pools in a steady state condition, which may otherwise be lost. Also, by focusing seagrass conservation efforts on countries with the highest seagrass biomass carbon stocks and emissions potential (e.g., Australia, Spain, and Mexico), together with the much larger soil carbon stocks, the climate mitigation impact of blue carbon initiatives could be optimized. Importantly, regions with relatively higher biomass and soil carbon stocks per unit area, such as the Mediterranean, offer unique opportunities to maximize the climate benefit of blue carbon strategies through tailored conservation actions.

The growing emphasis on blue carbon science reinforces the recognition of seagrasses as critical contributors to climate change mitigation and adaptation16,24. The estimates of annual NPP provided here contribute to further understanding the key role of seagrasses in the carbon cycle, consolidating seagrasses as some of the most productive ecosystems in the biosphere, while the global estimates of biomass carbon stocks presented in this study allows seagrass biomass to be included in national carbon inventories and, along with actions to conserve and restore seagrass meadows, in National Determined Contributions (NDCs) for climate change mitigation. Ultimately, integrating seagrass biomass into global carbon accounting can enhance the accuracy of blue carbon estimates and boost the implementation of seagrass restoration and conservation projects.

Limitations, assumptions, and data gaps

This study increased the number of biomass data reported in Fourqurean et al.16 by 91% (from 251 to 2743 data points). Our estimates of total seagrass biomass were up to 1.5–fold lower than those reported by Fourqurean et al.16, whereas median estimates are comparable (1.1 and 1.0 Mg OC ha⁻¹, respectively). This study reduced the uncertainties associated with previous estimates by increasing the dataset, including data from previously underrepresented regions, and by obtaining total seagrass biomass estimates by applying weighted averages based on life-history strategies, accounting for the disproportionate number of observations associated with opportunistic and colonizing genera.

This study relies on assumptions that limit the representativeness of the global estimates of carbon stocks and NPP provided. Global estimates of carbon stocks in seagrass biomass and NPP were obtained throughout multiple years and seasons, and therefore, we assumed that mean values per category are representative of the annual variability in biomass and NPP. Also, it seems likely that global biomass, NPP, and emissions calculated with global extent estimates may be underestimated due to lack of data or low confidence estimates on seagrass extent for some regions7. The poor seagrass mapping in several regions around the world (17% of the seagrass-harboring countries and territories remain unmapped)7 underestimates their contribution in the global carbon cycle, and the potential emissions resulting from historic seagrass losses. Current and future seagrass losses (and associated carbon emissions) should be reassessed owing to the strict management measures currently in place in many countries that slowed down or even reversed the declining trends in seagrass extent48.

Our study highlights significant data gaps in assessing seagrass biomass carbon stocks and NPP globally. Data were available for only 66 countries, representing 44% of seagrass-harboring sovereign states7. Regions including small island nations such as Tonga, Timor-Leste, and Vanuatu, and African nations like Somalia and Senegal, were not represented, while others like Sudan, Russia, or Chile were underrepresented in terms of limited data compared to other countries. Similarly, total biomass data were unavailable for 92 countries, and NPP data were missing for 132 countries, representing 58 and 82%, respectively, of the seagrass-harboring countries. Future research should expand the geographic and seasonal scope of data collection, particularly in underrepresented regions, to help disentangle the drivers behind the high variability observed in biomass seagrass carbon stocks across bioregions and life-history strategies. Indeed, the standardization of biomass and NPP data acquisition facilitates comparisons among studies.

Region-specific and seasonal variability should be embedded to improve global estimates and support equitable policy development, including the integration of seagrass biomass carbon stocks into ecosystem-based approaches for climate change mitigation and carbon accounting. The accurate mapping of seagrass worldwide remains a critical gap that precludes unraveling the full potential of seagrass ecosystems for climate adaptation and mitigation. Collaboration among governments, researchers, and conservation organizations will contribute to fulfilling current gaps.

Methods

Data collection and quality control

Data on seagrass above-, below-ground, and total biomass and NPP was sourced from “Global dataset on seagrass meadow structure, biomass and production”49,50. This published dataset was complemented with a new literature review search conducted in Web of Science (Clarivate™; webofknowledge.com) and Google Scholar (Google™; scholar.google.com) to encompass posterior and previously unidentified peer-reviewed literature. We conducted a search across titles, keywords, and abstracts, for the 1980-2024 time period, using the terms: “seagrass*” AND (“biomass” OR “carbon” OR “nutrient*” OR “product*” OR “NPP” OR “growth”). We identified 128 new papers and extracted relevant data. We also extracted data from 33 peer-reviewed manuscripts on organic carbon content (%) in above- and below-ground biomass for the 12 seagrass genera.

To evaluate and select relevant seagrass studies, we established specific criteria to ensure robust and reliable data extraction. The studies selected were required to report above-ground, below-ground, or total biomass, or NPP per unit area explicitly on a per-area basis in the text, figures, or tables, excluding those presenting only relative measures. Selected studies had to provide geolocation (latitude and longitude) and specify seagrass genera; those lacking these details or involving mixed meadows without genus-level resolution were excluded. Experiments conducted in mesocosms or natural environments with substantial modifications (e.g., nutrient enrichment or acidification) were not considered, focusing instead on in situ measurements. Additionally, studies needed to include original data not previously published, ensuring the use of novel information. These criteria facilitated the selection of studies offering accurate, spatially explicit estimates of seagrass biomass and/or NPP.

After applying the selection criteria, we compiled and reviewed 351 peer-reviewed scientific papers, encompassing 8576 data points on seagrass biomass and NPP from 791 coordinates distributed worldwide (Fig. 5). Of these, 342 papers contributed 8077 data points on biomass, including above-ground (N = 3353), below-ground (N = 1981), and total biomass (N = 2743). In contrast, only 38 papers provided the 499 data points compiled on NPP (N = 196 for above-ground, N = 93 for below-ground, and N = 210 for total NPP), with some of these manuscripts also containing biomass data. Additionally, 33 papers reported 761 data points on carbon content (%OC) in above-ground (N = 404) and below-ground (N = 357) biomass.

Fig. 5: Global distribution of data on seagrass biomass and net primary production (NPP) among seagrass bioregions8.
Fig. 5: Global distribution of data on seagrass biomass and net primary production (NPP) among seagrass bioregions8.
Full size image

Distribution of the data points containing information of seagrass above-, below-ground or total biomass (green), above-, below-ground or total NPP (orange) and both biomass and NPP (light blue) among seagrass bioregions. TS Temperate Southern, TIP Tropical Indo Pacific, TA Tropical Atlantic, TNAE Temperate North Atlantic East, TNAW Temperate North Atlantic West, M Mediterranean, TNP Temperate North Pacific. Basemap outlines were obtained from Opendatasoft (https://public.opendatasoft.com/) under the Open Government Licence v3.0.

Global, bioregion- and country-specific estimates of seagrass extent were sourced from McKenzie et al. 7, which provides estimates categorized by confidence levels: high confidence (supported by both qualitative and quantitative data) and low confidence (based only on qualitative data). For each country, the high-confidence extent estimate was prioritized when available; otherwise, the low-confidence estimate was used (Supplementary Table 4).

Numerical analyses

Standardized units of living biomass (g DW m−2) and NPP (g DW m−2 day−1 or g DW m−2 yr−1) were extracted from the reviewed literature. The biomass and NPP of leaves and sheaths were considered above-ground, whereas roots and rhizomes were considered below-ground. The sum of above- and below-ground biomasses and NPP constituted total biomass and total NPP (when total biomass and NPP were not explicitly reported). The % organic carbon in above- and below-ground biomass was weight-averaged for each genus based on leaf and root/rhizome percentages sourced from published literature, respectively (Supplementary Table 5). Biomass and NPP expressed as dry weight (DW) data were transformed into organic carbon by multiplying each value by the seagrass genus-specific mean % organic carbon content (Supplementary Table 5). For seagrass genera where there were no data available for % organic carbon content in below-ground biomass (i.e., Amphibolis and Phyllospadix), the mean % organic carbon value in below-ground across all genera with available data (38%) was used. For the calculation of total carbon in total biomass and total NPP in each seagrass genera, a weighted calculation was performed based on the proportion of above- and below-ground biomass. This was run as the sum of the weighted values of above- and below-ground biomass multiplied by the % organic carbon content in above- or below-ground biomass, respectively, and then multiplying by the total biomass. Data on seagrass carbon biomass and NPP are reported as kg C ha−1 and kg C ha−1 yr−1, respectively.

We assigned each data point to a specific seagrass bioregion defined as Temperate North Atlantic East and West, Tropical Atlantic, Mediterranean, Tropical Indo Pacific, Temperate Southern, Temperate North Pacific8 (Fig. 5); and classified seagrass genera as colonizing (i.e, Halodule, Halophila, and Ruppia), opportunistic (i.e., Syringodium, Zostera, Cymodocea and Phyllospadix,) or persistent (i.e., Posidonia, Enhalus, Amphibolis, Thalassia, Thalassodendron) according to its life-history strategy10. Global, bioregional, and national estimates of biomass and NPP were obtained by applying weight averages based on life-history strategies and seagrass extents. Global biomass and NPP values were calculated by multiplying the mean global biomass and NPP value by the high and low confidence global extent estimate7.

Global, bioregional and national estimates of seagrass biomass carbon emissions from seagrass biomass loss were calculated using the data on seagrass loss extent, expressed as the percentage of net change per bioregion over a century18. Country-specific annual seagrass extent loss was calculated by proportionally allocating the net percentage change of each bioregion to individual countries based on their proportional coastline extent within each bioregion (Supplementary Table 4). This allocation was then multiplied by the total seagrass extent for each country7, and divided by the temporal period over which seagrass loss was assessed to calculate annual rates. Similarly, bioregion-specific annual seagrass extent loss was estimated by multiplying the net percentage change per bioregion by the total seagrass extent within that bioregion7 and dividing by the same temporal period of estimates.

To estimate carbon emissions from biomass, the annual extent of losses was multiplied by the mean seagrass total biomass for each country or bioregion. It was assumed that 90% of the biomass would ultimately be mineralized following seagrass loss, whereas the remaining 10% would be preserved elsewhere28. These biomass values were subsequently converted to carbon dioxide equivalents (CO2-eq) using a conversion factor of 3.67, reflecting the stoichiometric relationship between carbon and CO2. This methodological framework provides a robust, spatially explicit assessment of carbon emissions from seagrass biomass across global, bioregional, and national scales.

Statistical analyses

To evaluate the differences in above, below and total biomass and NPP for different life-history strategies, genera and bioregions (considered as fixed factors) we used generalized linear models (GLM). The biomass and NPP data were log- and square root-transformed, respectively, to meet model assumptions. These models were evaluated in R Studio software with the glmmTMB function from the glmmTMB package51,52. The significance of the fixed factors analyzed was evaluated with the Anova function of the car package53 that performs Wald chi-square tests. Post-hoc pairwise comparisons were conducted using Estimated Marginal Means (EMM) and were performed with the emmeans function of the emmeans package with Tukey adjustment54. The normality and homoscedasticity assumptions were assessed by visually inspecting the residuals in a Q-Q plot. These analyses were implemented using R studio software51 and the results were visualized using ggplot2 package55 with a combination of box-and-whisker plots and bar plots. The boxplots y-axes were transformed using the 4th-root to enhance visualization while maintaining the display of real values. Global heatmaps were generated in QGIS56 to show the distribution of the data compiled, and the geographical patterns of total and per unit area biomass carbon and NPP across life-history strategies and countries.