ARTICLE https://doi.org/10.1038/s41467-019-12176-8 OPEN Australian vegetated coastal ecosystems as global hotspots for climate change mitigation Oscar Serrano et al.# Policies aiming to preserve vegetated coastal ecosystems (VCE; tidal marshes, mangroves and seagrasses) to mitigate greenhouse gas emissions require national assessments of blue 1234567890():,; carbon resources. Here, we present organic carbon (C) storage in VCE across Australian climate regions and estimate potential annual CO2 emission benefits of VCE conservation and restoration. Australia contributes 5–11% of the C stored in VCE globally (70–185 Tg C in aboveground biomass, and 1,055–1,540 Tg C in the upper 1 m of soils). Potential CO2 emissions from current VCE losses are estimated at 2.1–3.1 Tg CO2-e yr-1, increasing annual CO2 emissions from land use change in Australia by 12–21%. This assessment, the most comprehensive for any nation to-date, demonstrates the potential of conservation and restoration of VCE to underpin national policy development for reducing greenhouse gas emissions. Correspondence and requests for materials should be addressed to O.S. (email: o.serranogras@ecu.edu.au). #A full list of authors and their affiliations appears at the end of the paper. NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 D estruction and degradation of natural ecosystems a accounts for 12–20% of the CO2 released into the atmo- Bioregion Tropical Tidal marsh sphere globally1. Despite their relatively small global Subtropical Semi-arid N Arid N extent (between 0.5 and 1 × 106 km2, equivalent to 0.2% of the Temperate Northern ocean surface), vegetated coastal ecosystems (VCE), tidal mar- Territory Western Queensland shes, mangroves and seagrasses, contribute ~50% of C seques- Australia tered in marine sediments2 (i.e., blue carbon), with their organic South Australia New carbon (C) sequestration rates exceeding those of terrestrial for- South Wales ests, per unit area, by 1–2 orders of magnitude3. Hence, con- Victoria servation and restoration of VCE has an important potential to 1000 km contribute to climate change mitigation4,5. 1000 km Tasmania Blue carbon ecosystems are among the most threatened eco- b systems on Earth. The global area occupied by VCE is being Mangrove Seagrass globally reduced at rates ranging from 0.03 to >1% per year, twice as high as those reported for tropical forests3,5. These losses led to Northern Northern the development of blue carbon strategies to prevent and mitigate Territory Territory Western Queensland Western Queensland greenhouse gas (GHG) emissions through the conservation and Australia Australia South South restoration of VCE. The development of programs like REDD+, Australia New Australia New the payments for ecosystem services6 and the inclusion of VCE South Wales South Wales within Nationally Appropriate Mitigation Actions7 aim to Victoria Victoria maintain the benefits these ecosystems provide to climate change mitigation and adaptation, fisheries, and other ecosystem services Tasmania Tasmania that support coastal communities and their livelihoods5,8. Blue carbon strategies are now being included within Nationally c 160 160 Determined Contribution to mitigate and adapt to climate 140 B 140 change. However, this requires strong scientific evidence, and Aboveground biomass C stock Aboveground biomass C stock 120 120 whereas reports of C stocks and sequestration rates in VCE have recently increased exponentially5,9–11, comprehensive estimates 100 100 (Mg C ha–1) (Tg C) of blue carbon storage at national and continental scales are 80 80 lacking, particularly for tidal marshes and seagrass. Uncertainties 60 60 on the extent of these ecosystems, their C stocks and sequestra- 40 40 tion rates, as well as limited available data on the loss and fate of 20 A 20 C after disturbance, hinder adoption of VCE into carbon trading 0 A 0 and national inventories5–7. Tidal marsh Mangrove Seagrass Tidal marsh Mangrove Seagrass Here, we pioneer the estimation of C stocks in aboveground 300 B biomass and soils, as well as soil C sequestration rates, in VCE at 1000 250 the national level, and do so for the Australian continent, one of Soil C stock (Mg C ha–1) 800 the major reservoirs of VCE in the planet. We estimate the 200 A Soil C stock (Tg C) potential for VCE conservation and restoration to mitigate GHG 600 150 emissions in Australia and demonstrate, therefore, the potential C of blue carbon strategies to support policies contributing to cli- 100 400 mate change mitigation at the national level. 50 200 0 0 Results Tidal marsh Mangrove Seagrass Tidal marsh Mangrove Seagrass Australian blue carbon. Total C stocks in aboveground biomass 3.50 B and the upper 1-m of VCE soils in Australia were 67–183 Tg C 1.5 3.00 and 1053–1542 Tg C, respectively, with annual soil C seques- Soil C sequestration rates Soil C sequestration rates tration rates of 3.5–5.5 Tg C year−1 (Fig. 1, Table 1). 2.50 (Mg C ha–1 yr–1) (Tg C yr–1) 1.0 The extent, geographic distribution and type of VCE determine 2.00 the distribution of C stocks and sequestration rates over the 1.50 continent (Fig. 1). Mangroves contain ~80% of total C in living 0.5 A A 1.00 aboveground biomass of Australian VCE, while seagrasses accounted for ~70% of total soil C stocks and sequestration 0.50 rates. Most C stocks in seagrass and tidal marsh ecosystems are 0.0 0.00 Tidal marsh Mangrove Seagrass Tidal marsh Mangrove Seagrass found in their soils (98% and 99%, respectively), while C stocks in mangrove ecosystems are distributed in both soil (62%) and aboveground biomass (38%) pools (Table 1). Australian man- with higher C storage capacity of mangroves and tidal marshes groves have up to 17-fold and 65-fold higher C stocks in compared to seagrasses on an areal basis3,5. aboveground biomass per unit area compared to tidal marshes Most tidal marsh and mangrove ecosystems in Australia occur and seagrasses, respectively (P < 0.001; Fig. 1), while mangrove in tropical regions (62% and 73%, respectively), while seagrasses soil C stocks and sequestration rates are 2-times and 3-times are distributed across subtropical (38%), tropical (32%), and arid higher than tidal marshes and seagrasses, respectively (P < 0.001; (16%) climate regions (Supplementary Table 2). Tidal marsh and Fig. 1, Supplementary Table 1). The soil C stocks per unit area in mangrove soil C stocks and sequestration rates are 3-fold to tidal marshes are 1.5-fold higher than in seagrass meadows (P < 13-fold higher in tropical Australia than in other climate regions, 0.01; Fig. 1). These values are consistent with global estimates, mainly due to their extensive coverage over broad intertidal 2 NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 ARTICLE Fig. 1 . Distribution of climate regions, vegetated coastal ecosystems (tidal large, but poorly mapped. Hence, the extent of seagrasses and marshes, mangroves and seagrasses) and organic carbon (C) storage in tidal marshes may be significantly larger than currently estimated, Australia. a Climate regions used to classify vegetated coastal ecosystems so their blue carbon contribution estimated here is a and scale blue carbon storage across Australia. Climate regions for conservative one. Australia were modified from the Australian Bureau of Meteorology’s Combined, the Australian VCE soil C stocks and sequestration “Koppen–Major Classess” climate classification for Australia based on rates per unit area are up to 3-fold higher in tropical regions temperature/humidity, vegetation and seasonal rainfall61. The original compared to other regions (P < 0.001; Fig. 2, Supplementary climate classification scheme was comprised of six classes: Equatorial, Table 3), while C stocks in aboveground biomass are significantly Tropical, Subtropical, Desert, Grassland and Temperate, but the number of higher in both subtropical and tropical regions (P < 0.05; climate regions was reduced into five categories: Tropical (includes Supplementary Table 1). The C stocks and sequestration rates Equatorial), Subtropical, Arid (instead of Desert), Semi-arid (instead of per unit area also differ, though not consistently, among the three Grassland) and Temperate. b Spatial distribution of tidal marsh25, ecosystem types (Fig. 2, Supplementary Table 3). For tidal mangrove26, and seagrass28 ecosystems within Australia. c Organic carbon marshes, the C stocks in aboveground biomass per unit area stocks in living aboveground biomass and soils (in the top meter), and C are up to 6-fold higher in in temperate regions compared to semi- sequestration rates per unit area (Mg C ha−1) and across Australia (Tg C). arid and subtropical regions (P < 0.05), while soil C stocks and The stacked bars represent the maximum and minimum estimates (s.d.). sequestration rates per unit area are not significantly different Source data are provided as a Source Data file among climate regions (P > 0.05; Supplementary Table 1). This likely reflects the influence of higher biomass species (e.g., Juncus spp. rushes) in many temperate marshes, compared to the lower biomass species (e.g., Sporobolus virginicus and Sarcocornia saltflats in tropical regions (Supplementary Table 2). Similarly, quinqueflora) typically found in tropical, arid and subtropical mangrove soil C stocks and sequestration rates are up to 60-fold climates. Variability in tidal marsh soil accretion rates and C higher in tropical compared to other climate regions, owing to stocks is often associated with differences in the position within their larger extent in the tropics12. Subtropical seagrasses within the intertidal zone14, porewater salinity, sediment inputs, and Australia hold 2-fold to 6-fold higher C stocks than seagrasses plant productivity15. from other Australian climate regions. However, knowledge of Tropical mangroves contain up to 2-fold higher C stocks in Australian seagrass extent is incomplete due to challenges in aboveground biomass per unit area compared to temperate mapping this ecosystem, as recently illustrated by the recent mangroves (P < 0.001), and soil C stocks and sequestration rates discovery of 35,000 km2 of tropical seagrass in the intensively are up to 2-fold higher in subtropical mangroves compared to studied Great Barrier Reef13. Similarly, the spatial extent of other climate regions (P < 0.05; Fig. 2, Supplementary Tables 1 tropical tidal marshes (including high intertidal saltflats) is likely and 3), which is in agreement with previous studies16,17. Table 1 Organic carbon (C) storage in Australian vegetated coastal ecosystems (i.e., tidal marshes, mangroves and seagrasses), per unit area (in Mg C ha−1 and Mg C ha−1 year-1) and Australia-wide (in Tg C) a Ecosystem Stock-aboveground biomass per unit area (Mg C ha−1) Total area (Mha) Stock-aboveground biomass (Tg C) N plots Mean Median SD Min Max Min Max Tidal marsh 52 7.5 6.4 6.1 1.4 1.5 2.3 2.6 Mangrove 37 125 94 90 0.3 1.1 50 158 Seagrass 52 1.9 1.5 2.0 9.3 12.8 16 22 Total 141 11.0 15.4 67 183 b Ecosystem Stock-soil (Mg C ha−1 in 1 m-thick) Total area (Mha) Stock-soil (Tg C) N cores Mean Median SD Min Max Min Max Tidal marsh 292 168 140 127 1.4 1.5 210 234 Mangrove 262 251 238 155 0.3 1.1 81 257 Seagrass 549 112 85 88 9.3 12.8 762 1051 Total 1103 11.0 15.4 1053 1542 c Ecosystem Sequestration rates-Soil (Mg C ha−1 year−1) Total area (Mha) Sequestration rates (Tg C year−1) N cores Mean Median SD Min Max Min Max Tidal marsh 292 0.39 0.3 0.3 1.4 1.5 0.48 0.54 Mangrove 24 1.26 0.9 0.9 0.3 1.1 0.4 1.4 Seagrass 36 0.36 0.3 0.3 9.3 12.8 2.5 3.5 Total 352 11.0 15.4 3.5 5.5 Mean and median ± SD C stock in (a) living aboveground biomass and (b) in the top meter of soil c: Soil C sequestration rates. Mha = 106 ha NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 a Soil C stock (Mg C ha–1) Tidal marsh Mangrove Seagrass N 1000 km ≤125 >125 to ≤151 >151 to ≤196 >196 b Biomass C stock (Mg C ha ) –1 Tidal marsh Mangrove Seagrass N 1000 km ≤1 >1 to ≤3 >3 to ≤85 >86 c –1 –1 Soil C sequestration rate (Mg C ha yr ) Tidal marsh Mangrove Seagrass N 1000 km ≤0.31 >0.31 to ≤0.40 >0.40 to ≤0.85 >0.85 Fig. 2 Scaled up estimates of organic carbon (C) storage in vegetated coastal ecosystems (tidal marshes, mangroves and seagrasses) across Australian climate regions. a Soil C storage (Mg C ha−1) in the top meter. b Living aboveground biomass C stock (Mg C ha−1). c Soil C sequestration rates (Mg C ha−1 year−1). The four ranges of data (indicated by different colors) are based on the lower quartile, median quartile, and upper quartile. Source data are provided as a Source Data file Subtropical and temperate mangroves in Australia are mainly unit area among climate regions (P < 0.001), while tropical comprised of a single species, Avicennia marina, while tropical seagrasses have the lowest soil C stocks (P < 0.001). However, Australian mangroves comprise more than 40 species18. The seagrass soil C sequestration rates are similar among climate species richness, together with climatic and environmental drivers regions (P > 0.05), suggesting higher remineralization rates of soil associated with variations in coastal geomorphology (i.e., C in tropical regions4. The relatively low C stocks of tropical temperature, rainfall, tides, sediments and nutrients) govern the seagrasses may be due to the predominance of small, ephemeral structure and function of mangroves and tidal marshes, and and fast-growing species in the tropics (e.g., Halophila spp. and thereby their C storage capacity, with relatively high plant Halodule spp.), which also support intensive grazing by sea turtles productivity and biomass in tropical, subtropical and some and dugongs, compared to the persistent meadows with large temperate regions compared to arid regions9,19,20-22. biomass found in arid and temperate regions (e.g., Posidonia spp. Seagrasses inhabiting arid climate regions have 5-fold to 9-fold and Amphibolis spp.), where herbivory rates are much lower10,23. higher C stocks in aboveground biomass per unit area than those Meadows formed by these larger temperate seagrass species in other climate regions (P < 0.001; Fig. 2, Supplementary Table 3). contain relatively high amounts of degradation-resistant organic Seagrasses in arid regions also exhibit the highest soil C stocks per compounds in their tissues (e.g., lignin and cellulose) compared 4 NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 ARTICLE to small tropical seagrass species, which have readily decom- nations with greatest potential to benefit from developing blue posable tissues24. In addition, the scarcity of riverine inputs in carbon schemes4,9,33. Yet, Australia is, through the estimates arid regions likely results in higher irradiance reaching the provided here, the only nation with a robust estimate of existing seafloor compared to other climate regions with higher run-off, blue carbon resources at the national scale. thereby enhancing aboveground and belowground seagrass Soil C stocks in Australian terrestrial ecosystems have been productivity13,23 and soil C stocks in arid regions. estimated at 29.7 Mg C ha−1 (in 30 cm-thick soils34), while Biomass and soil C stocks and sequestration rates for VCE vary Australian VCE contain on average ~4-fold higher C stocks among administrative jurisdictions (Supplementary Table 4). (123 Mg C ha−1 in 1 m-thick soils) than in terrestrial counter- Queensland, Northern Territory and Western Australia hold the parts. Note that this estimate is conservative since terrestrial and largest area of tidal marshes (39%, 28%, and 25%, respectively) VCE soil C stocks can reach thicknesses >1 m and >6 m depth, and mangroves (39%, 37%, and 20%, respectively) within respectively10,12,34. Furthermore, VCE can accrete vertically over Australia, while Queensland and Western Australia have the vast millennia without becoming saturated in C, and soil C stocks are majority of seagrass ecosystems (65% and 20%, respectively). This generally protected from fires. As a result, C turnover rates in information is an essential foundation for the implementation of VCE are an order of magnitude slower than most terrestrial blue carbon accounting processes and climate change mitigation soils5,35. Our estimate of the soil CO2 sequestration capacity of strategies (i.e., C policies, abatement and trading schemes) in Australian VCE (13–20 Tg CO2-e year−1) is equivalent to ~4–6% Australia. of the CO2 emissions from fossil-fuel burning, cement production and gas flaring in Australia (361 Tg CO2-e year−1 at 2014 rates36). In addition, the CO2 sequestration capacity of Australian Australia as a global blue carbon hotspot. Australia is among VCE equates to 70–140% of CO2 emissions from land use change the countries that hold the largest area of VCE (11–15 Mha, in Australia (estimated in 14.4–18.4 Tg CO2-e year−1 37,38). 9–32% of VCE worldwide), including 3–37% of global tidal Hence, VCE are a significant component of Australia’s C budget marshes (1.4–1.5 Mha3,11,25), ~2–8% of mangroves (0.3–1.1 and provide effective opportunities for C sequestration and Mha3,17,25,26) and ~15–43% of global seagrass meadows climate change mitigation strategies based on VCE conservation (9.3–12.8 Mha5,13,25,27,28; Table 2). Accordingly, Australia has and restoration. some of the world’s largest blue carbon storage capacity, with 5–11% of global blue carbon soil stocks and 2–7% of annual soil C sequestration, based on available global estimates3–5,10,11,17,29,30 Potential of Australian VCE for greenhouse gas mitigation. (Table 2). Soil C stocks per unit area in Australian VCE are Historic losses of VCE extent after European settlement in the similar to global estimates, but their C sequestration rates per unit 19th century in Australia have been estimated at 13,800 km2 for area are lower than global mean estimates3,5, particularly for tidal tidal marsh14 (47–50% loss of original extent), 11,500 km2 for marshes and seagrasses (Table 2). This difference can be partially mangroves33,39 (52–78% loss of original extent) and 32,000 km2 explained by the relative sea-level stability of Australian coastlines for seagrass40 (20–26% loss of original extent). In addition to the compared to other regions (i.e., North America and Europe) that loss of important ecosystems services (e.g., coastal protection, have experienced rapid sea-level rise over the last millennia, fisheries and biodiversity5), losses of Australian VCE result in a hence, enhanced soil C accumulation31. In addition, differences in loss of their CO2 sequestration potential, remineralization of C carbon accumulation rates among climate regions may arise from stocks in aboveground biomass as CO2, and the risk that erosion differences in methods used for estimating soil accumulation and remineralization of soil C accumulated over millennia con- rates (i.e., surface elevation tables, 210Pb, 239+240Pu, and 14C) and tributes to GHG emissions4. Management activities, which fall the periods of accumulation encompassed by the estimates32. outside business-as-usual scenarios, aiming to restore VCE can Nevertheless, the extent of VCE in Australia makes it a global enhance soil C sequestration and/or avoid GHG emissions fol- blue carbon hotspot, and the loss of VCE extent since European lowing disturbance, as demonstrated for Australian seagrass arrival provides unique opportunities for Australia to mitigate meadows41, and thus have the potential to be eligible for C emissions through blue carbon strategies. Indeed, Australia, along crediting under financial mechanisms. The Australian Govern- with Indonesia, Malaysia, USA and Brazil rank among the ment Emission Reduction Fund is a policy mechanism that Table 2 Extent, soil organic carbon (C) sequestration rates and stocks (in the top meter of soil) in vegetated coastal ecosystems a Global extension (km2) Global soil C sequestration rate Global C stock in soil (Pg C) (Tg C year−1) Ecosystem Min Max Min Max Min Max Tidal marsh 41,65711 400,0003 4.83 87.33 0.674,11 6.55 Mangrove 137,7603 166,00017 2329 2529 5.030 6.417 Seagrass 300,0005 600,0005 483 1123 4.210 8.410 Total 479,417 1,166,000 75 224 14.2 21.3 b Australian extension (km2) Australian soil C sequestration Australian C stock in soil (Tg C) rate (Tg C year-1) Ecosystem Min Max Min Max Min Max Tidal marsh 13,765 (3%) 15,329 (37%) 0.5 (0.5%) 0.5 (11%) 210 (3.2%) 234 (35%) Mangrove 3315 (2%) 10,509 (8%) 0.4 (1.8%) 1.4 (6%) 81 (1.3%) 257 (5%) Seagrass 92,569 (15%) 127,720 (43%) 2.5 (2.3%) 3.5 (7%) 762 (9%) 1,051 (25%) Total 109,649 (9%) 153,558 (32%) 3.5 (1.6%) 5.5 (7%) 1053 (5%) 1542 (11%) a: Global estimates based on studies providing data (mean or median values) based on global datasets b: estimates for Australia. The proportion (maximum and minimum) of Australian ecosystems compared to global estimates is presented in brackets NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 attempts to promote GHG abatement through the conservation, in Australia (1150 km2) would enhance soil C sequestration by restoration or creation of VCE42. For example, restoring tidal 0.65 ± 0.46 Tg CO2-e year−1, reducing annual emissions from flows to drained coastal areas has been identified as a feasible land use change in Australia by 4–5%. It is therefore imperative to activity for restoring VCE42. Evidence of the effectiveness of VCE develop policies that preserve and restore VCE to mitigate GHG restoration for enhancing biomass and soil C sequestration is emissions in Australia, where recovery of historic losses offers a available43–45, although Australian case studies are limited41. vast potential for C sequestration. Importantly, conservation and The potential of VCE conservation for climate change restoration of VCE also provide enhanced adaptation to climate mitigation relies on the preservation of their millenary soil C change through coastal protection and regulation of flooding, as stocks. Following disturbance or conversion of VCE, a portion of well as biodiversity and fisheries benefits5,47. the soil C becomes exposed to oxic conditions and decays at a Recent coastal development in Australia continues to result relatively fast rate (estimated at 0.183 year−1), resulting in the in a net decline in VCE extent, estimated at a minimum 0.03% remineralization of 85% of the soil C stock exposed to oxic year−1 for Australian mangroves (100–315 ha year−1; Table 3)38. conditions within a decade46. Since major historic losses of VCE The decline of tidal marsh extent in Australia remains unknown, extent in Australia occurred more than two decades ago, but it is likely similar or higher than in mangroves (415–460 ha restoration of these areas may have little benefit in terms of year−1)14. Similarly, current decline in seagrass extent in avoided GHG emissions. Australia has not been estimated, and here we assume a loss of Assuming a recovery of soil C sequestration after VCE 0.1% year−1 (i.e., 9300–12,800 ha year-1) largely resulting from rehabilitation, we estimate that the restoration of an area dredging, water quality deterioration and shoreline modification, equivalent to 10% of historic losses of VCE extent in Australia to estimate potential CO2 emissions resulting from habitat loss48. (5730 km2) would enhance soil C sequestration by 1.15 ± 0.91 Tg This assumption is conservative relative to global estimates, CO2-e year−1 (mean ± SD; Table 3), reducing annual emissions which are one order of magnitude higher40. Assuming that 50% from land use change in Australia by 6–8%. This can also result in of the aboveground biomass and soil C in the top meter are a benefit of US$ 11.5 ± 9.1 million per annum (assuming a remineralized after disturbance4,49, we estimate emissions of conservative C trading price of US$10 t CO2-e−1, the approx- 2.1–3.1 Tg CO2-e year−1 as a result of current losses of Australian imate value paid in Australia’s Emissions Reduction Fund VCE (Table 3), increasing emissions from land use change in auctions). The benefits of restoration vary among VCE and Australia by 12–21% per annum. The loss of VCE would also climate regions, averaging 1.4 ± 1.1 Mg CO2 ha year−1 for tidal result in a loss of future soil C sequestration of 13–19 Gg CO2-e marsh (13–16 US$ ha year−1), 4.6 ± 3.3 Mg CO2 ha year−1 for year−1. Thus, conservation actions resulting in avoided losses of mangroves (15–145 US$ ha year−1), and 1.3 ± 1.2 Mg CO2 ha VCE in Australia, falling outside business-as-usual scenarios, year−1 for seagrass (10–18 US$ ha year−1). These estimates are could result in avoided CO2 emissions and sustained C conservative since potential avoided GHG emissions from soil C sequestration valued at 22–31 million US$ per annum following the restoration of historic losses of VCE extent were not (3000–4000 US$ ha year−1 for tidal marsh, 2000–22,000 US$ accounted for, while the restoration of mangroves would also ha year−1 for mangroves, or 1500–3000 US$ ha year−1 for entail enhanced CO2 sequestration in aboveground biomass. seagrass). Targeting blue carbon hotspots for restoration can provide larger benefits per unit area. Mangroves, in particular those occupying tropical regions where there is potential for reversing Implementation of blue carbon strategies in Australia. Sus- large, historic losses in coastal wetland extent42, are hotspots for tainable management of VCE requires an informed under- C sequestration. For example, the restoration of an area standing of the ecological and economic significance of changes equivalent to 10% of historic losses in tropical mangrove extent in natural resources due to threats such as human activities and Table 3 Potential annual CO2 emissions from loss of Australian vegetated coastal ecosystems and economic valuation a Ecosystem Total stock (Soil + Habitat loss per C at risk of Potential CO2 Economic value of CO2 Biomass) (Tg C) year (ha year−1) remineralization emissions emissions per year (Tg C year−1) (Tg CO2-e year−1) ($10 t CO2−1) (106 US$) Tidal marsh 212–237 413–460 0.036–0.040 0.13–0.15 1.3–1.5 Mangrove 131–415 99–315 0.019–0.059 0.07–0.22 0.7–2.2 Seagrass 778–1,073 9,257–12,772 0.53–0.73 1.9–2.7 19–27 Total 1121–1725 9769–13,547 0.58–0.83 2.1–3.1 21–30 b Ecosystem Sequestration rates in Habitat loss per Lack of C Potential lack of CO2 Economic value of lack of CO2 Australia (Tg C year−1) year (ha year−1) sequestration sequestration sequestration per year (Gg C year−1) (Gg CO2-e year−1) ($10 t CO2−1) (106 US$) Tidal marsh 0.48–0.53 413–460 0.16–0.18 0.58–0.65 0.0058–0.0065 Mangrove 0.4–1.4 99–315 0.13–0.40 0.46–1.46 0.0046–0.015 Seagrass 2.5–3.5 9,257–12,772 3.3–4.6 12.2–16.9 0.12–0.17 Total 3.5–5.5 9,769–13,547 3.6–5.2 13.3–19.0 0.13–0.19 a: Potential gross annual emissions (Tg CO2-e year-1) from aboveground biomass and soils as a result of the decline in vegetated coastal ecosystems extent in Australia (0.03% year-1 for tidal marshes and mangroves, and 0.1% year-1 for seagrasses). Emission estimates assume that 50% of organic carbon (C) stocks in aboveground biomass and in the top meter of soil deposits are remineralized after ecosystem loss (at a rate of 0.183 year−1 46) b: Potential annual loss of CO2 sequestration capacity in blue carbon soils as a result of current losses in the extent of vegetated coastal ecosystems in Australia, assuming that soil C accretion does not occur after ecosystem loss. Carbon trading price of US$ 10 per ton of CO2. Economic value is expressed in 106 US$ 6 NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 ARTICLE climate change. Current assessments of CO2 emissions from VCE spectrometer. Percentage soil C on a mass basis was calculated for the bulk (pre- due to land use change in Australia have focused on emissions acidified) samples. Data on soil accumulation rates from 315 cores derived by means of 210Pb (last due to conversion of mangroves and tidal marshes to settlements century) and/or radiocarbon (last millennia) was compiled. Concentration profiles (Tier 1 estimates), and have not accounted for emissions asso- of 210Pb along the upper part of the sediment cores were determined by alpha ciated with losses of seagrass nor for conversion of mangrove and spectrometry through the measurement of 210Po using Passivated Implanted tidal marsh to pasture37. The Australian Government established Planar Silicon (PIPS) detectors (CANBERRA, Mod. PD-450.18 A.M) after acid the International Partnership for Blue Carbon after the Con- digestion of the samples, assuming radioactive equilibrium between the 210Pb and 210Po radionuclides. After alpha spectrometry, selected samples from each core ference of the Parties to the United Nations Framework on Cli- were analyzed for 226Ra by ultra-low background liquid scintillation counting mate Change conference in Paris 2016. The Australian (LSC, Quantulus 1220) or gamma spectrometry through the emission lines of its Government also commissioned a technical review of the inclu- daughter radionuclides 214Pb and 214Bi (295.2, 351.9, and 609.3 KeV). The sion of blue carbon projects in its domestic carbon abatement concentration profiles of excess 210Pb were determined by subtraction of 226Ra scheme (the Emissions Reduction Fund), through management of from total 210Pb concentrations along each core. Gamma spectrometry measurements were conducted in some cores using semi-planar intrinsic ecosystems towards the enhancement of C storage and/or avoided germanium high purity coaxial detectors with 40% efficiency, housed in a lead GHG emissions42. Our assessment of Australian national blue shield, coupled to a multichannel analyzer. 210Pb activity was determined by the carbon storage, accounting for the various climate regions and direct measurement of 46.5 KeV gamma peak. Sediment accumulation rates were administrative jurisdictions, provides a basis to estimate potential obtained by applying the Constant Rate of Supply (CRS53) or the Constant Flux: Constant Sedimentation models (CF:CS54). The 239+240Pu activities were measured CO2 abatement through restoration and conservation of VCE. in a sector ICPMS55. Our results show that Australia is a hotspot for VCE holding Samples of bulk soil, plant debris and shells along the cores were radiocarbon large quantities of blue carbon storage equivalent to 5–11% of dated following standard procedures56. The 14C dates from seagrass cores were blue carbon soil stocks worldwide, despite losses amounting to calibrated using the marine13 calibration curve57 considering a local Delta R ranging from 3 to 71 years as a function of study site58. The corrected ages were 47–78% of tidal marsh and mangrove extents, and 20–26% of used to produce an age-depth model (linear regression) to estimate sediment seagrass extent since European arrival. Therefore, Australia accumulation rates. stands to benefit from developing blue carbon-focused climate change mitigation schemes. Restoration of historic losses of VCE together with enhanced conservation of threatened VCE could Data analyses. To allow direct comparison among study sites, the C storage per unit area (cumulative stocks, mass C m−2) was standardized to 1 m-thick deposits constitute a mechanism to mitigate Australian CO2 emissions (i.e., length of the soil cores sampled). When necessary, we inferred C stocks below while enhancing ecosystem services and climate adaptation the limits of the reported data to 1 m, extrapolating linearly integrated values of C capacity. The estimates reported here provide a pioneer content (cumulative C stock per unit area) with depth. Correlation between demonstration of the approach required to deliver estimates that extrapolated C stocks and measured C stocks in sediment cores ≥1 m was r2 = 0.87 (P < 0.001). The C sequestration rates (mass C m−2 year−1) were calculated by can be incorporated into national carbon accounting and multiplying average C concentration by the sediment accumulation rate (mass m−2 underpinning the incorporation of robust blue carbon strategies year−1) in each core (where quantified). Estimates of aboveground biomass per within Nationally Determined Contribution to mitigate climate unit area were obtained by drying and weighing aboveground materials for tidal change. The baseline map of blue carbon in Australia provides an marshes and seagrasses, and using field measurements and allometric equations (specific to the region and species) for mangroves59,60. essential underpinning to assess the impact of land use changes All analyses were performed using Generalized Linear Model procedures in and climate change on blue carbon fluxes and stocks. The pioneer SPSS v. 14.0. A Generalized Linear Model was used to consider the potential non- assessment at a national, and continental, level reported here independence of samples taken within habitats. All response variables (C stocks in provides a methodology beyond the use of Tier 1 approaches aboveground biomass and soil C stocks and sequestration rates) were square-root currently available in the IPCC Wetland Supplement50 that transformed prior to analyses and had homogenous variances. Climate region (arid, semi-arid, temperate, subtropical, and tropical) and ecosystem type (tidal provides an exemplar of an approach toward estimating national marsh, mangrove and seagrass) were treated as fixed factors in all statistical models blue carbon resources elsewhere. (probability distribution: normal; link function: identity). Potential C stock losses (mass C) and CO2 emissions (mass CO2-e year−1) were estimated based on 0.03% annual ecosystem area loss for tidal marshes and Methods mangroves, and 0.1% year−1 for seagrass, and accounted for the sum of C stocks in Data acquisition. Data on C stocks and sequestration rates in Australian tidal aboveground biomass and the top meter of soils, assuming that 50% of total C marshes, mangrove forests and seagrass meadows were compiled from published stocks are lost and remineralized to CO2 after disturbance4,49. data. In addition, unpublished studies from the CSIRO Marine and Coastal Carbon The upscaling of each habitat polygon was performed by multiplying the Biogeochemistry Cluster project and other studies by the co-authors were included. average ± SD soil C stocks, sequestration rates, and standing C stocks in the The study sites included mono-specific and/or mixed tidal marsh, mangrove and aboveground biomass for each ecosystem within each climate region by the specific seagrass ecosystems within a variety of depositional environments (from estuarine ecosystem area to obtain blue carbon estimates at climate region scale (arid, semi- to exposed coastal areas, and supra-tidal to sub-tidal habitats) across five climate arid, temperate, subtropical and tropical, adapted from61; Fig. 1) and regions (arid, semi-arid, temperate, subtropical and tropical) in Australia. Data administrative jurisdictions within Australia (Northern Territory, Queensland, from 1553 study sites (593 from tidal marshes, 323 from mangrove forests and 637 New South Wales, Victoria, Tasmania, South Australia and Western Australia). from seagrass meadows) on soil C stocks (1103 cores in total), soil C sequestration The datasets on biomass C stocks (N = 37 for mangroves and N = 52 for both tidal rates (352 cores in total) and standing C stocks in aboveground biomass (141 marshes and seagrasses) and on soil C sequestration rates for mangroves (N = 24) measurements in total) were used in this study. and seagrasses (N = 36) were limited, which resulted in data gaps within climate Soil cores were sampled using different coring mechanisms, such as manual regions (Supplementary Fig. 1 and Supplementary Table 3). For example, estimates percussion and rotation of PVC pipes, vibracoring or using a Russian corer. The of biomass C stocks in tidal marsh are lacking for arid and tropical regions cores were sliced at regular intervals, each slice/sample was weighed before and (Supplementary Table 3). In order to estimate C storage in VCE around Australia, after oven drying to constant weight at 60–70 °C (i.e., dry weight, DW). C data from the nearest climate region was used when data was not available. The The organic C content of the soil organic matter was measured in milled extents of each ecosystem considered to scale up Australia-wide estimates of C subsamples from multiple slices along cores. The ‘Champagne test’ was used to stocks and sequestration (based on climate regions) are presented in Table 2 and determine whether soil samples contained inorganic carbon51. The soil core sub- Supplementary Table 2. samples containing carbonates were acidified with 1 M HCl, centrifuged (3500 Substantial data gaps for blue carbon stocks and sequestration rates exists in RPM; 5 min) and the supernatant with acid residues was removed using a pipette, parts of the country (Supplementary Fig. 1 and Supplementary Table 3). Most then washed in deionized water, centrifuged again and the supernatant removed. notably, there are limited data over much of northern Australia, where ecosystem These residual samples were re-dried (60–70 °C) before C elemental analyses. The extent is greatest for all three VCE. Stocks, sequestration rates, and losses of C are method used to remove inorganic C prior to organic C analyses may lead to the also poorly quantified in converted or modified systems and there are few studies of loss of part of the organic C (soluble fraction), thereby potentially leading to an C sequestration capacity in restored ecosystems49. The robustness of the global underestimation of sediment C content52. Where carbonates were absent (all living assessments presented here relies on a number of estimates, including the extent of plant samples and most tidal marsh and mangrove soil samples), bulk soil samples Australian VCE, annual loss rates, degree and fate of soil C loss after disturbance, were milled and encapsulated without acid pre-treatment before C analyses. The C differences in the impact of different types of disturbances or management content was obtained using a dry combustion elemental analyzer or mass activities, and the type of C trading methodology used. Therefore, the potential C NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 abatement and its economic value may vary across VCE, and with management 27. Lavery, P. S., Mateo, M. A., Serrano, O. & Rozaimi, M. Variability in the and political scenarios. carbon storage of seagrass habitats and its implications for global estimates of Blue Carbon ecosystem service. PLoS ONE 8, e73748 (2013). 28. Mount, R., et al. National Intertidal/Subtidal Benthic (NISB). In Habitat Map Data availability Data. (University of Tasmania, Department of Climate Change and the A Source Data File, containing the raw data underlying the research and all figures and National Land and Water Resources Audit, 2008). tables presented in our paper, is available in the Supplementary Information. The spatial 29. Breithaupt, J. L., Smoak, J. M., Smith, T. J. III, Sanders, C. J. & Hoare, A. datasets that support the findings of this study have been deposited in the Organic carbon burial rates in mangrove sediments: strengthening the global Commonwealth Scientific and Industrial Research Organisation portal with the identifier budget. Glob. Biogeochem. Cycles 26, GB3011 (2012). https://doi.org/10.25919/5d3a8acc9b598. 30. Jardine, S. L. & Siikamäki, J. V. A global predictive model of carbon in mangrove soils. Environ. Res. Lett. 9, 104013 (2014). Received: 8 April 2019 Accepted: 21 August 2019 31. Rogers, K. et al. Wetland carbon storage controlled by millennial-scale variation in relative sea-level rise. Nature 567, 91 (2019). 32. Baskaran, M., Bianchi, T. S. & Filley, T. R. Inconsistencies between 14C and short-lived radionuclides-based sediment accumulation rates: effects of long- term remineralization. J. Environ. Radioact. 174, 1–7 (2016). 33. Hamilton, S. E. & Friess, D. A. Global carbon stocks and potential emissions References due to mangrove deforestation from 2000 to 2012. Nat. Clim. Change 8, 240 1. Le Quéré, C. et al. Global Carbon Budget 2017. Earth Syst. Sci. Data 10, (2018). 405–448 (2018). 34. Viscarra Rossel, R. A. et al. Baseline map of organic carbon in Australian soil 2. Duarte, C. M., Middelburg, J. J. & Caraco, N. Major role of marine vegetation to support national carbon accounting and monitoring under climate change. on the oceanic carbon cycle. Biogeosci. Discuss. 1, 659–679 (2004). Glob. Change Biol. 20, 2953–2970 (2014). 3. Mcleod, E. et al. A blueprint for blue carbon: toward an improved 35. Barrett, D. J. Steady state turnover time of carbon in the Australian terrestrial understanding of the role of vegetated coastal habitats in sequestering CO2. biosphere. Glob. Biogeochem. Cycles 16, 55 (2002). Front. Ecol. Environ. 9, 552–560 (2011). 36. CDIAC. Carbon Dioxide Information Analysis Center. http://cdiac.ornl.gov 4. Pendleton, L. et al. Estimating Global “Blue Carbon” emissions from (2019). conversion and degradation of vegetated coastal ecosystems. PLoS ONE 7, 37. Haverd, V. et al. The Australian terrestrial carbon budget. Biogeosciences 10, e43542 (2012). 851–869 (2013). 5. Duarte, C. M. et al. The role of coastal plant communities for climate change 38. Department of the Environment and Energy. National Inventory Report 2015. mitigation and adaptation. Nat. Clim. Change 3, 961–968 (2013). Vol. 2. (Commonwealth of Australia, 2017). 6. Thomas, S. Blue carbon: knowledge gaps, critical issues, and novel approaches. 39. Friess, D. A. & Webb, E. L. Variability in mangrove change estimates and Ecol. Econ. 107, 22–38 (2014). implications for the assessment of ecosystem service provision. Glob. Ecol. 7. Ullman, R., Bilbao-Bastida, V. & Grimsditch, G. Including blue carbon in Biogeogr. 23, 715–725 (2014). climate market mechanisms. Ocean Coast Manag. 83, 15–18 (2013). 40. Waycott, M. et al. Accelerating loss of seagrasses across the globe threatens 8. Barbier, E. B. et al. The value of estuarine and coastal ecosystem services. Ecol. coastal ecosystems. Proc. Natl Acad. Sci. 106, 12377–12381 (2009). Monogr. 81, 169–193 (2011). 41. Marbà, N. et al. Impact of seagrass loss and subsequent revegetation on carbon 9. Atwood, T. B. et al. Global patterns in mangrove soil carbon stocks and losses. sequestration and stocks. J. Ecol. 103, 296–302 (2015). Nat. Clim. Change 7, 523 (2017). 42. Kelleway, J. et al. Technical Review of Opportunities for Including Blue Carbon 10. Fourqurean, J. W. et al. Seagrass ecosystems as a globally significant carbon in the Australian Government’s Emissions Reduction Fund. (CSIRO, Canberra, stock. Nat. Geosci. 5, 505–509 (2012). 2017). 11. Ouyang, X. & Lee, S. Y. Updated estimates of carbon accumulation rates in 43. Greiner, J. T., McGlathery, K. J., Gunnell, J. & McKee, B. A. Seagrass coastal marsh sediments. Biogeosciences 11, 5057–5071 (2014). restoration enhances “blue carbon” sequestration in coastal waters. PLoS ONE 12. Donato, D. C. et al. Mangroves among the most carbon-rich forests in the 8, e72469 (2013). tropics. Nat. Geosci. 4, 293–297 (2011). 44. Osland, M. J. et al. Ecosystem development after mangrove wetland creation: 13. Coles, R. G. et al. The great barrier reef world heritage area seagrasses: plant–soil change across a 20-year chronosequence. Ecosystems 15, 848–866 managing this iconic Australian ecosystem resource for the future. Estuar. (2012). Coast. Shelf S. 153, A1–A12 (2015). 45. Burden, A., Garbutt, R. A., Evans, C. D., Jones, D. L. & Cooper, D. M. Carbon 14. Macreadie, P. I. et al. Carbon sequestration by Australian tidal marshes. Sci. sequestration and biogeochemical cycling in a saltmarsh subject to coastal Rep. 7, 44071 (2017). managed realignment. Estuar. Coast. Shelf S. 120, 12–20 (2013). 15. Rogers, K., Wilton, K. M. & Saintilan, N. Vegetation change and surface 46. Lovelock, C. E., Fourqurean, J. W. & Morris, J. T. Modeled CO2 emissions elevation dynamics in estuarine wetlands of southeast Australia. Estuar. Coast. from coastal wetland transitions to other land uses: tidal marshes, mangrove Shelf S. 66, 559–569 (2006). forests, and seagrass beds. Front. Mar. Sci. 4, 143 (2017). 16. Simard, M. et al. Mangrove canopy height globally related to precipitation, 47. Creighton, C., Boon, P. I., Brookes, J. D. & Sheaves, M. Repairing Australia’s temperature and cyclone frequency. Nat. Geosci. 12, 40 (2019). estuaries for improved fisheries production–what benefits, at what cost? Mar. 17. Sanderman, J. et al. A global map of mangrove forest soil carbon at 30 m Freshw. Res. 66, 493–507 (2015). spatial resolution. Environ. Res. Lett. 13, 055002 (2018). 48. Larkum, A. W., Kendrick, G. A. & Ralph, P. J. (eds). Seagrasses of Australia: 18. Duke, N., Ball, M. & Ellison, J. Factors influencing biodiversity and Structure, Ecology and Conservation. (Springer, 2018). distributional gradients in mangroves. Glob. Ecol. Biogeogr. 7, 27–47 (1998). 49. Lovelock, C. E. et al. Assessing the risk of carbon dioxide emissions from blue 19. Osland, M. J. et al. Climatic controls on the global distribution, abundance, carbon ecosystems. Front. Ecol. Environ. 15, 257–265 (2017). and species richness of mangrove forests. Ecol. Monogr. 87, 341–359 (2016). 50. Intergovernmental Panel on Climate Change. Climate Change 2007: The 20. Allen, J. R. L. Morphodynamics of Holocene salt marshes: a review sketch Physical Science Basis. Contribution of Working Group I to the Fourth from the Atlantic and Southern North Sea coasts of Europe. Quat. Sci. Rev. 19, Assessment Report of the Intergovernmental Panel on Climate Change (eds. 1155–1231 (2000). Solomon S., et al.) (Cambridge University Press, Cambridge, 2007). References 21. Rovai, A. S. et al. Global controls on carbon storage in mangrove soils. Nat. in Methods section Clim. Change 8, 534 (2018). 51. Burt, R. Soil Survey Staff. Soil Survey Laboratory Methods Manual. Soil Survey 22. Twilley, R. R., Rovai, A. S. & Riul, P. Coastal morphology explains global blue Investigations Report 42, Version 5.0. US Department of Agriculture, Natural carbon distributions. Front. Ecol. Environ. 16, 503–508 (2018). Resources Conservation Service (National Soil Survey Center, 2014). 23. Carruthers, T. J. B. et al. Seagrasses of south–west Australia: A conceptual 52. Brodie, C. R. et al. Evidence for bias in C and N concentrations and δ13C synthesis of the world’s most diverse and extensive seagrass meadows. J. Exp. composition of terrestrial and aquatic organic materials due to pre-analysis Mar. Biol. Ecol. 350, 21–45 (2007). acid preparation methods. Chem. Geol. 282, 67–83 (2011). 24. Trevathan-Tackett, S. M. et al. A global assessment of the chemical 53. Appleby, P. G. & Oldfield, F. The calculation of lead-210 dates assuming a recalcitrance of seagrass tissues: implications for long-term carbon constant rate of supply of unsupported 210Pb to the sediment. Catena 5, 1–8 sequestration. Front. Plant Sci. 8, 925 (2017). (1978). 25. Geoscience Australia. Australian Coastal Waterways Geomorphic Habitat 54. Krishnaswamy, S. et al. Geochronology of lake sediments. Earth Planet Sci. Mapping Data. http://www.ozcoasts.gov.au/search_data/datasets.jsp (2005). Lett. 11, 407–414 (1971). 26. Giri, C. et al. Status and distribution of mangrove forests of the world using 55. Sanders, C. J. et al. Elevated rates of organic carbon, nitrogen, and phosphorus earth observation satellite data (version 1.3, updated by UNEP-WCMC). Glob. accumulation in a highly impacted mangrove wetland. Geophys. Res. Lett. 41, Ecol. Biogeogr. 20, 154–159 (2011). 2475–2480 (2014). 8 NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 ARTICLE 56. Stuiver, M. & Reimer, P. J. Extended 14C data base and revised CALIB 3.0 14C version of the paper. O.S., C.E.L., T.A., P.I.M., R.C., S.P., A.A.-O., L.B., J.B., C.B., P.C., age calibration program. Radiocarbon 35, 215–230 (1993). R.M.C., P.D., A.E., C.J.E.L., B.D.E., M.A.H., P.H., L.B.H., C.R.J.K., J.J.K., G.A.K., K.K., 57. Reimer, P. J. et al. IntCal13 and Marine13 radiocarbon age calibration curves A.L., S.Y.L., P.S.L., D.T.M., N.M., P.M., M.A.M., R.M., P.J.R., C.R., M.R., R.R., C.S., 0–50,000 years cal BP. Radiocarbon 55, 1869–1887 (2013). J.S.-V., J.S., C.S., I.S., C.S., A.D.L.S., T.C., S. M.T.-T. and C.M.D. contributed by providing 58. Ulm, S. Australian marine reservoir effects: a guide toΔ R values. Aust. their data and improving the paper. Archaeol. 63, 57–60 (2006). 59. Kauffman, J. B. & Donato, D. Protocols for the Measurement, Monitoring and Additional information Reporting of Structure, Biomass and Carbon Stocks in Mangrove Forests. Supplementary information accompanies this paper at https://doi.org/10.1038/s41467- (CIFOR, Bogor, 2012). 019-12176-8. 60. Guerra-Santos, J. J. et al. Estimation of the carbon pool in soil and above- ground biomass within mangrove forests in Southeast Mexico using allometric Competing interests: The authors declare no competing interests. equations. J. Forestry Res 25, 129–134 (2014). 61. Australian Government Bureau of Meteorology. Climate classification of Reprints and permission information is available online at http://npg.nature.com/ Australia. http://www.bom.gov.au/jsp/ncc/climate_averages/climate- reprintsandpermissions/ classifications/index.jsp?maptype=kpngrp#maps (2019). Peer review information Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available. Acknowledgements This project was supported by the CSIRO Marine and Coastal Carbon Biogeochemical Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in Cluster, CSIRO Oceans and Atmosphere, the ECU Faculty Research Grant Scheme and published maps and institutional affiliations. Early Career Research Grant Schemes, UTS Plant Functional Biology and Climate Change Cluster, NSW Southeast Local Land Services, Department of Environment, Land, Water and Planning (DELWP), Parks Victoria, Victorian Coastal Catchment Manage- ment Authorities (GHCMA, CCMA, PPWCMA, WGCMA, EGCMA), University of Open Access This article is licensed under a Creative Commons Queensland Centennial Scholarship, Hodgkin Trust Scholarship, Australian Institute of Attribution 4.0 International License, which permits use, sharing, Nuclear Science and Engineering, Northern Territory Government Innovation Grant, adaptation, distribution and reproduction in any medium or format, as long as you give Australian Research Council (DE130101084, DE140101733, DE150100581, appropriate credit to the original author(s) and the source, provide a link to the Creative DE160100443, DE170101524, DP150103286, DP150102092, DP160100248, Commons license, and indicate if changes were made. The images or other third party DP160100248, DP180101285, LE140100083, LE170100219, LP150100519, LP160100242 material in this article are included in the article’s Creative Commons license, unless and LP110200975), the Generalitat de Catalunya (MERS 2014 SGR-1356), the ICTA indicated otherwise in a credit line to the material. If material is not included in the ‘Unit of Excellence’ (MinECo, MDM2015-0552), Obra Social “LaCaixa”, SUMILEN, article’s Creative Commons license and your intended use is not permitted by statutory CTM 2013-47728-R, Ministry of Economy and Competitiveness and UKM-DIP-2017- regulation or exceeds the permitted use, you will need to obtain permission directly from 005. The authors are grateful to G. Bastyan, D. Kyrwood, G. Davis, J. Bongiovanni, A. the copyright holder. To view a copy of this license, visit http://creativecommons.org/ Jesse, Q. Hua, A. Zawadzki, J. Gudiño, P. Bray, H. Markham, M. Lepore, K-le Gómez- licenses/by/4.0/. Cabrera, and J. Pandolfi for their help in field and/or laboratory tasks. © The Author(s) 2019 Author contributions C.M.D. and C.E.L. conceived and designed the research, O.S. coordinated data acquisi- tion with contributions from all co-authors, analyzed the data, and drafted the first Oscar Serrano 1, Catherine E. Lovelock 2,3, Trisha B. Atwood 3,4, Peter I. Macreadie 5, Robert Canto3,6, Stuart Phinn 3,6, Ariane Arias-Ortiz 7, Le Bai8, Jeff Baldock 9, Camila Bedulli10,11, Paul Carnell5, Rod M. Connolly 12, Paul Donaldson13, Alba Esteban1, Carolyn J. Ewers Lewis5, Bradley D. Eyre 14, Matthew A. Hayes2,3,12, Pierre Horwitz1, Lindsay B. Hutley 8, Christopher R.J. Kavazos 1,15, Jeffrey J. Kelleway 16, Gary A. Kendrick 10,17, Kieryn Kilminster17,18, Anna Lafratta1, Shing Lee12,19, Paul S. Lavery 1,20, Damien T. Maher 14, Núria Marbà 21, Pere Masque 1,7,10,22, Miguel A. Mateo1,20, Richard Mount23, Peter J. Ralph24, Chris Roelfsema6, Mohammad Rozaimi 1,25, Radhiyah Ruhon10,26, Cristian Salinas1,27, Jimena Samper-Villarreal3,28,29, Jonathan Sanderman 9,30, Christian J. Sanders31, Isaac Santos 31, Chris Sharples 23, Andrew D.L. Steven 32, Toni Cannard32, Stacey M. Trevathan-Tackett5 & Carlos M. Duarte 10,33 1 School of Science and Centre for Marine Ecosystems Research, Edith Cowan University, Joondalup, WA 6027, Australia. 2School of Biological Sciences, University of Queensland, St. Lucia, QLD 4072, Australia. 3The Global Change Institute, University of Queensland, St. Lucia, QLD 4072, Australia. 4Department of Watershed Sciences and Ecology Center, Utah State University, Logan, UT 84322, USA. 5Centre for Integrative Ecology, School of Life and Environmental Sciences, Deakin University, Geelong, Burwood Campus, Geelong, VIC 3125, Australia. 6Remote Sensing Research Centre/Joint Remote Sensing Research Program, School of Earth and Environmental Sciences, University of Queensland, Queensland, QLD 4072, Australia. 7Institut de Ciència i Tecnologia Ambientals and Departament de Física, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. 8 Research Institute for the Environment and Livelihoods, Charles Darwin University, Casuarina, NT 0810, Australia. 9CSIRO Agriculture and Food, Locked Bag 2, Glen Osmond, SA 5064, Australia. 10UWA Oceans Institute, The University of Western Australia, Crawley, WA 6009, Australia. 11 Instituto de Biociências de Botucatu, Universidade Estadual Paulista, Botucatu 18618-970, Brazil. 12Australian Rivers Institute—Coast and Estuaries, School of Environment andScience, Griffith University, Gold Coast, QLD 4222, Australia. 13BMT Environment, Newcastle, NSW 2292, Australia. 14Centre for Coastal Biogeochemistry, School of Environment, Science and Engineering, Southern Cross University, Lismore, NSW 2480, Australia. 15School of Biological, Earth and Environmental Sciences, University of New South Wales, Kensington, NSW 2052, Australia. 16School of Earth, Atmospheric and Life Sciences, University of Wollongong, Wollongong, NSW 2522, Australia. 17School of Biological Sciences, The University NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications 9 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12176-8 of Western Australia, Crawley, WA 6009, Australia. 18Department of Water and Environmental Regulation, Locked Bag 10, Joondalup DC, WA 6027, Australia. 19Simon FS Li Marine Science Laboratory, Chinese University of Hong Kong, Shatin, Hong Kong. 20Centre d’Estudis Avançats de Blanes-CSIC, 17300 Blanes, Spain. 21Global Change Research Group, IMEDEA (CSIC-UIB), Institut Mediterrani d’Estudis Avançats, Miquel Marquès 21, 07190 Esporles, Spain. 22School of Physics, The University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia. 23Discipline of Geography and Spatial Sciences, School of Technology, Environments and Design, University of Tasmania, Hobart, TAS 7001, Australia. 24 Climate Change Cluster, University of Technology Sydney, PO Box 123, Broadway, NSW 2007, Australia. 25Centre for Earth Sciences and Environment, Faculty of Science and Technology, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia. 26Faculty of Marine Science and Fisheries, Hasanuddin University, Jl. Perintis Kemerdekaan Km.10, Tamalanrea, Makassar 90245, Indonesia. 27Marine and Coastal Research Institute “José Benito Vives De Andréis” INVEMAR, Calle 25 No. 2-55, Santa Marta, Colombia. 28Centro de Investigación en Ciencias del Mar y Limnología (CIMAR), Ciudad de la Investigación, Universidad de Costa Rica, San Pedro, San José 11501-2060, Costa Rica. 29Marine Spatial Ecology Lab, University of Queensland, St Lucia, QLD 4072, Australia. 30Woods Hole Research Center, Falmouth, MA 02540, USA. 31National Marine Science Centre, Southern Cross University, PO Box 4321, Coffs Harbour, NSW 2450, Australia. 32CSIRO Oceans and Atmosphere, Queensland Biosciences Precinct, 306 Carmody Rd, St. Lucia, QLD 4067, Australia. 33Red Sea Research Center (RSRC) and Computational Bioscience Research Center (CBRC), King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia 10 NATURE COMMUNICATIONS | (2019)10:4313 | https://doi.org/10.1038/s41467-019-12176-8 | www.nature.com/naturecommunications