Coastal ecosystems, such as mangrove forests, are among the world’s most productive ecosystems, maintaining high levels of biodiversity (Thompson and Rog, 2019) and delivering substantial ecosystem services to support local- to global-scale human well-being relative to their spatial coverage (Donato et al., 2011; McLeod et al., 2011; Curnick et al., 2019). Particularly pertinent in the current global climate emergency is their ability to support climate change mitigation and adaptation (CCMA) across the world’s coasts, due to high relative sequestration and storage of “blue carbon,” and protecting coastal communities and infrastructure from increasingly frequent storm conditions (Donato et al., 2011; Lee et al., 2014; Duncan et al., 2016; Hochard et al., 2019). Despite their importance, mangroves remain in global decline due to high coastal land-use demand and extractive dependency (Richards and Friess, 2016; Thomas et al., 2017; Bunting et al., 2018; Friess et al., 2019), facing substantial future challenges from abiotic climate change processes (Lovelock et al., 2015; Ward et al., 2016). In some regions, extensive historical mangrove loss to production land-uses such as aquaculture ponds (Richards and Friess, 2016; Kauffman et al., 2017; Bunting et al., 2018; Goldberg et al., 2020) has led to their rehabilitation being high on national coastal management and conservation agendas (Primavera and Esteban, 2008; Phan et al., 2015; Lee et al., 2019). To promote incentives for conservation intervention, increasing empirical scientific focus is now placed on identifying investible mangrove ecosystem service benefits, in particular for blue carbon projects (e.g., The Blue Natural Capital Financing Facility, 2021; Zeng et al., 2021). Such opportunities could unlock sustainable conservation financing, particularly in low- and middle-income nations where international carbon markets could offer comparatively attractive payments for environmental stewardship (Thompson et al., 2014). However, high perceived risk in blue carbon permanence, uncertainty in creditable emissions forecasting in the absence of blue carbon-specific quantification methodologies, large project costs and political risk have meant that mangroves’ high CCMA potential has historically been largely unrealized in terms of operational blue carbon projects (Locatelli et al., 2014; Wylie et al., 2016; Herr et al., 2017).

In 2020, a major milestone was reached with the first blue carbon emissions reduction/sequestration quantification methodology approved under the Verified Carbon Standard that now enables inclusion of disproportionately large carbon sequestration in the soil compartment of blue carbon ecosystems (Verra, 2020a). This is likely to now herald a wealth of emerging blue carbon projects globally. To capitalize on this opportunity, a challenge to the design and implementation of blue carbon projects now lies in the ability of practitioners to assess potential return-on-investment (ROI), and hence their viability, prior to embarking on extensive and costly project registration and verification processes. While a major driver of potential blue carbon project costs is likely to be spatial scale, variation in initiation and on-going budget requirements will also depend on project implementation design. This is particularly true for rehabilitation-oriented mangrove management, where multiple intervention options exist with diverse associated costs and probability of long-term success (Bayraktarov et al., 2016; Lee et al., 2019; Wodehouse and Rayment, 2019; Su et al., 2021). Managers may adopt Ecological Mangrove Rehabilitation, with interventions to reinstate former hydrology in converted lands (e.g., abandoned aquaculture ponds, salt ponds, and agricultural land) to facilitate natural mangrove regeneration [hereafter “natural regeneration” (NR); Lewis and Brown, 2014], or more costly Assisted Natural Regeneration (hereafter “ANR”). ANR employs (a) (community-based) out-planting of nursery-reared or naturally available wildings to supplement NR in sites with more challenging or sub-optimal conditions, and (b) optional restoration of hydrology [e.g., breaching of aquaculture pond banks (“dikes”) where required] (Primavera et al., 2012b; Mangrove Action Project, 2021). Either strategy may be optimal for a given site, with faster and denser mangrove regeneration enhancing credit generation, associated ecosystem services (e.g., coastal protection, fisheries enhancement), and therefore pricing under co-benefit accreditation schemes (Plan Vivo, 2013). However, quantitative evaluation of which rehabilitation strategy may be financially optimal in a given context remains challenging. Research to date has quantified investible blue carbon opportunities and ROI utilizing broad mean project costs across large spatial scales and contexts (e.g., Jakovac et al., 2020; Taillardat et al., 2020; Zeng et al., 2021). However, the relative ecological outcomes of mangrove NR and ANR strategies are rarely considered in empirical studies, and these authors know of no existing study to date that explores variation in blue carbon credit generation potential and ROI under differing rehabilitation intervention strategies. This empirical data gap presents a particular challenge for managers designing potential small-scale (and community-based) blue carbon projects with minimal implementation budgets. There is thus a significant gap between the theory and practice of mangrove blue carbon projects that may represent a barrier to the development of operational projects to date despite the high interest.

The guiding principles underpinning successful implementation of NR-oriented rehabilitation are tidal connectivity to a viable source of mangrove propagules, favorable intertidal position [i.e., elevation above mean sea level (a.m.s.l.)] and low tidal energy exposure for rapid propagule establishment and regeneration in former mangrove areas (Lewis and Brown, 2014). By contrast, ANR can bolster NR rates in favorable sites (see also Huxham et al., 2010) and/or supplant NR where one or more of these conditions is sub-optimal, thus enhancing mangrove regeneration rates and strengthening coastal protection (“greenbelts”) in the face of rapidly advancing climate change (Primavera et al., 2012b; Mangrove Action Project, 2021). However, ANR can incur substantial costs (Primavera et al., 2012b; Bayraktarov et al., 2016), which must be weighed against the relative potential carbon credit returns vs. NR alone (i.e., ROI) to assess whether the approach is justified. Drivers of mangrove NR have been extensively established in experimental and site-specific field studies (e.g., Balke et al., 2011; Kamali and Hashim, 2011); however, their empirical quantification across larger scales (e.g., between sites) is absent, hindering out ability to predict (rates of) site-specific carbon sequestration potential from NR relative to ANR. Furthermore, ANR may buffer natural inter-annual variability in propagule availability and/or stochastic perturbations, thereby reducing uncertainty in emissions reduction forecasting and associated credit reductions (see Verra, 2019), but this remains empirically unquantified across rehabilitation projects. These data gaps may be a reason that all currently registered blue carbon projects do not fully account for NR in their ex ante project emissions reduction forecasts [ANR strategies only: Blue Ventures, 2019; Mikoko Pamoja, 2020; rewetting/soil stabilization from reinstated hydrology only: Conservation International, 2021]. Remote sensing has long been applied in largely inaccessible mangroves to successfully track changes in their distribution and functioning (Wang et al., 2019), and high resolution imagery has recently been employed to track fine-scale (tree-specific) landward mangrove expansion (Whitt et al., 2020). Similar high resolution monitoring of NR and ANR rates combined with spatial analysis across landscape-scales may enable quantification of NR rate drivers and relative rehabilitation and emissions reduction potential under similar conditions. However, to date, these approaches have not been employed to enhance our predictive capacity in predicting carbon sequestration in areas under different management regimes and in evaluating potential blue carbon project scenario options.

To close these identified data gaps and investigate the relative efficacy of ANR over NR-oriented mangrove blue carbon projects, here we model potential carbon sequestration and credit generation potential in a hypothetical project using a case study abandoned aquaculture pond in West Visayas, Philippines. The Philippines has experienced substantial mangrove loss: approximately 50% of the former 500,000 ha (Spalding et al., 2010) disappeared over the last century, due primarily to “extensive” (large area but shallow depth) brackish-water pond aquaculture development in former mangrove areas (Primavera, 2005). Some of the highest pond densities occur in the West Visayas region (Primavera and Esteban, 2008; Department of Agriculture of the Philippines – Bureau of Fisheries and Aquatic Resources, 2020). Development is largely unregulated, and despite laws mandating 50–100 m greenbelts (Primavera et al., 2012a), ponds are often built to the shoreline. Abandonment is high (Samson and Rollon, 2011; Primavera et al., 2012a), due primarily to bank (“dike”) breaches in sea-facing ponds over low productivity (Primavera et al., 2014). Legal mandates in the Philippines dictate that government-leased ponds that are abandoned must be reverted to public lands and rehabilitated to mangrove forest by the relevant government department. However, a myriad of confounding factors, centered predominantly on low political will, means cancelation and reversion rarely occurs: large areas of ponds built in former mangrove lie fallow and the few canceled leases are often absorbed and re-tenured or operated illegally, thus the long-term persistence of any naturally regenerated (NR) mangroves within abandoned ponds is not secured (Primavera et al., 2014). If better regulated and enforced, abandoned pond tenure reversion could afford a major rehabilitated-oriented emissions reduction opportunity in the Philippines, with minimal physical intervention required to generate blue carbon additionality (natural dike breaching; Primavera et al., 2014). Herein, we explore the potential for this abandoned pond blue carbon opportunity under NR and assisted NR (ANR; e.g., planting) strategies. First, we map mangrove NR (n = 8) and ANR (n = 1) at annual time-steps at identified abandoned aquaculture ponds, using a novel image classification approach with open access high resolution imagery. Second, we use open access spatial data to model the main drivers of these time-series NR rates of mangrove areal coverage across abandoned ponds. Third, we employ model-derived and observed NR and ANR rates (and uncertainty) to predict mangrove regeneration across a hypothetical 10-year time period at our case study abandoned pond, and apply field- and literature-derived rehabilitated mangrove carbon sequestration data to forecast potential rehabilitated carbon stocks at the site under these scenarios. Finally, we apply approved emissions reduction quantification protocols (Verra, 2020b) under potential accreditation methodologies, alongside ANR project cost data, varied potential voluntary market carbon prices and discount rates, to explore variability in relative ROI across NR and ANR scenarios at our case study site. Here we hypothesize slower (and more variable) areal regeneration and carbon sequestration under NR, and greater ROI with ANR over NR after deduction of biophysical project costs.

Our analysis of identified abandoned ponds in West Visayas found within-pond natural mangrove regeneration (following dike breaching) from as early as 2005 (Leganes Katunggan) to 2013 (Panay, Capiz) (see Table 2, Figure 2, and Supplementary Figures 1A–F). Abandoned pond characteristics varied widely, with total pond area ranging from 1.58 (Basiao, Ivisan, Capiz) to 17.90 ha (Buenavista, Guimaras, Iloilo), adjacent mangrove extent (within a 2.5 km buffer) from 0.56 (Balagon and Napnud, Leganes, Iloilo) to 108.67 ha (Barotac Nuevo, Iloilo), distance to source population (nearest 10 ha patch) from 0.02 (San Dionisio, Iloilo) to 11.64 km (Roxas City, Capiz), proportional dikes remaining from 0.36 (Leganes Katunggan) to 0.95 (Buenavista, Guimaras), relative dike protection from 89.98 (Basiao, Ivisan, Capiz) to 526.35 m² m^–1 (Leganes Katunggan), SRTM DEM-derived mean pond elevation from 0.42 (Roxas City, Capiz) to 3.63 m (San Dionisio), and variation (CV) in pond elevation from 17.34% (San Dionisio) to 341.39% (Balagon and Napnud, Leganes) (see Table 2).

“Rate of areal increase” under NR was highly variable across abandoned pond time-series, ranging from 0.077 ± 0.007 (1 s.e.) ha year^–1 at Basiao, Ivisan to 1.434 ± 0.146 (1 s.e.) ha year^–1 at Balagon and Napnud, Leganes. Among the highest “rate of areal increase” was observed during assisted natural regeneration [ANR (planting): 2009?2013] at the Leganes Katunggan pond [1.327 ± 0.157 (1 s.e.) ha year^–1]; however, this rate was exceeded under NR at three sites [Balagon and Napnud, Leganes; Barotac Nuevo: 1.339 ± 0.215 (1 s.e.) ha year^–1; Buenavista, Guimaras: 1.337 ± 0.198 (1 s.e.) ha year^–1] (Table 2, Figures 2, 3, and Supplementary Figures 1A–F). “Leveling” of rates of areal increase was not reached in three non-case study NR abandoned ponds, possibly owing to recent (2012) and slow regeneration at San Dionisio, slow early regeneration at Roxas City, and very slow regeneration at Basiao, Ivisan (Table 2, Figures 2, 3, and Supplementary Figure 1). “Time-to-leveling” for the remaining ponds was lowest (4 years) at the small (6.40 ha) and moderately quickly regenerating Panay abandoned pond and under ANR at the Leganes Katunggan pond (Table 2 and Figure 3).

Little empirical guidance currently exists to enable mangrove rehabilitation practitioners to evaluate the potential ROI and viability of blue carbon project options. Our study provides the first to these authors’ knowledge that explores drivers and variation in mangrove regeneration rates in converted coastal areas at multi-site scales, and to quantify relative ROI in potential rehabilitation-oriented blue carbon project scenario options. We observed substantial variation in mangrove NR across abandoned aquaculture ponds, with the best-fitting model explaining faster NR rates in lower-lying (pioneer low- to mid-intertidal zone species) and less exposed ponds (greater dike retention). Mangrove recolonization was substantially faster and less variable (lower risk) in our case study pond under assisted natural regeneration (ANR; planting 2009–2013) than under either observed (2005–2009) or best-fitting model-predicted NR according to site conditions (Tables 2, 4). This translated to 3.7- to 5.2-fold greater carbon sequestration and 2.5- to 3.4-fold greater greenbelt regeneration (coastal protection: predicted pond mangrove coverage) over our hypothetical 10-year forecasting period with ANR. However, deducting generated potential carbon credit finance for realized project costs, ROI was low under all scenarios at current mean Afforestation/Reforestation project voluntary market carbon prices and ANR afforded a more optimal (higher ROI) rehabilitation strategy over NR only where mangrove soil compartment carbon sequestration was included in potential accreditation methodologies in our case study pond site. Projected financial returns (ROI) from ANR with autochthonous soil carbon inclusion (potential accreditation methodology BAS) were moreover negative and similar to those under NR at higher discount rates, highlighting that neither option may be appropriate under a small-scale blue carbon project financing-only lens at current credit prices. However, where site conditions (i.e., elevation, exposure) are less optimal, our findings reveal the relative merit of ANR over slower and more variable NR to strengthen coastal protection greenbelts more quickly in the face of accelerating global climate change, which may attract additional prospective investors to inflate realized project credit payments (Plan Vivo, 2013; Beeston et al., 2020; The Blue Natural Capital Financing Facility, 2021) toward those that also substantially maximize ROI over NR alone (Turner et al., 2021). We therefore urge prospective managers to consider these context–dependencies in local conditions constraining rapid NR, as well as to identify co-benefit credit pricing opportunities to ensure positive and greater relative ROI from blue carbon project. Overall, our approach provides a means to quantify mangrove regeneration and its drivers in converted lands across relevant (i.e., multi-site) geographic contexts and utilize these to evaluate rehabilitation-oriented blue carbon project scenarios that take in to account these enabling processes in local conditions, and “plug-in” forecasted project-specific costs.

While, as mangrove regeneration rates for our Leganes Katunggan, Iloilo case study were greatest under the forecasted ANR scenario (1.327 vs. 0.247 ha year^–1 in best-fitting model-derived NR scenario 1 and 0.352 ha year^–1 in observed NR scenario 2), we observed wide variability in NR rates across the seven studied NR abandoned ponds that were in three cases greater than the case study ANR scenario (Table 2). Our employed methodology and an absence of ground truth data meant we were unable to conduct time-series classification accuracy assessment. However, we are confident that our mapping approach combining manual visual classification of high-resolution imagery segmentation produced high accuracy (with low variability due to a single visual classifier analyzing each image) in estimated NR rates, and therefore that the rates quantified reflect context-dependence in NR drivers across sites. Depending on pond-specific conditions, ANR could still enhance some high NR rates observed by enhancing stem density, soil stability and inter-individual facilitation (Huxham et al., 2010) even under their apparently optimal local conditions. Our study could identify only one case study pond in which multiple regeneration processes (early NR followed by ANR: Primavera et al., 2012b) had occurred in the region, and further empirical and/or experimental study of mangrove NR vs. ANR rates at ponds with varying site-specific conditions is needed to identify the conditions at which ANR becomes redundant to background NR rates (see also Wodehouse and Rayment, 2019). Further mapping and quantification of NR across varied and more numerous abandoned ponds would also be beneficial in refining driver models: only eight abandoned ponds were identified here and we did not observe wide ranges of propagule supply within our model dataset (Table 2). We observed high NR rates in ponds with lower intertidal elevation (∼0.4–0.1 m a.m.s.l.) and low exposure (∼90% sea-facing dike retention) (Table 2). However, it is possible that the high NR rates observed at some abandoned ponds (Table 2) are driven by other known NR-influencing factors such as propagule supply (Balke et al., 2011; Lewis and Brown, 2014), where low sample size and variation in our dataset did not enable detection of a relationship, contrary to our hypothesis (Table 1). Interactions with factors that it was not possible to include in our landscape-level analysis, such as substrate type and salinity, could also have further driven variation in observed NR rates. We searched the high pond abandonment West Visayas coastal zone (∼31% of production from the Philippines total 239,323 ha of coastal aquaculture ponds: Department of Agriculture of the Philippines – Bureau of Fisheries and Aquatic Resources, 2020) exhaustively for study ponds, but a lack of high resolution imagery pre-2009 resulting in non-inclusion of many NR ponds for which we could not establish time of NR initiation. Future quantification of pond abandonment and NR could be evaluated against, and potentially supplemented by, mapping with longer-term moderate resolution imagery (e.g., Landsat) (Duncan et al., 2018; Baloloy et al., 2020); however the fine-scale of mangrove regeneration (i.e., individual tree establishment) may reduce the utility of these methods. Going forward, the generation of greater quantities of high resolution RGB imagery with existing and to-be-deployed satellite constellations will substantially enhance our ability to map and monitor newly abandoned coastal areas and refine NR driver models (Curnick et al., 2021).

Higher regeneration rates under ANR translated to substantially greater carbon sequestration and credit potential than under slower and more variable NR (Table 4). However, contrary to our predictions, application of blue carbon Verified Carbon Standard accreditation protocols (Verra, 2020b) and deduction of project costs (Supplementary Table 2) at current mean Afforestation/Reforestation project voluntary market carbon prices (USD $9.70 Mg CO₂e; Forest Trends’ Ecosystem Marketplace, 2021), resulted in ROI not only lower than NR scenarios but negative (net losses of USD $1,905 to $2,198) under the ANR scenario at Leganes Katunggan under biomass-only potential accreditation methodology (B). Biomass carbon sequestration at the case study ANR abandoned pond with sub-optimal conditions (elevation, exposure) is at the lower end of empirical estimates for rehabilitating mangroves (Sasmito et al., 2019), and biomass-only carbon emissions reductions may be sufficient to return positive ROI relative to ANR project costs in different contexts. However, our case study illustrates that biomass-only accreditation, accounting for only a small proportion of full emissions reduction portfolios, returns lower relative ROI in ANR rehabilitation-only blue carbon project scenarios. Thus NR may be the appropriate strategy under biomass-only accreditation in small-scale potential blue carbon projects at current mean credit prices (Afforestation/Reforestation projects). Relative ROI under ANR increased to marginally to substantially greater with methodologies with increased inclusion of soil compartment emissions reductions at a 1.5% discount rate (Table 4); yet, projected ROI under ANR with autochthonous soil carbon inclusion (methodology BAS) was negative and similar to that projected under NR at higher discount rates ≥ 3.5%. Thus neither NR nor ANR rehabilitation option may be appropriate under a small-scale blue carbon project financing-only lens with soil carbon inclusion at current credit prices, where soil compartment monitoring and verification inhibits ROI, unless additional co-benefits can drive higher credit pricing (see Plan Vivo, 2013; Mikoko Pamoja, 2020). Importantly, ANR was the only rehabilitation scenario that forecasted full mangrove cover within the 10-year time frame at the 9.58 ha Leganes Katunggan case study pond (upper NR estimate: 3.90 ha cover; Table 4). ANR therefore establishes wider greenbelts more quickly in abandoned ponds with sub-optimal NR conditions (i.e., elevation and exposure), which is critical for surge reduction in typhoon-prone areas and where national greenbelt mandates are a long way from being met (Primavera et al., 2012a). This additional CCMA benefit could furthermore strengthen ANR blue carbon projects with accreditors enabling higher payments for “bundled” co-benefits (e.g., Plan Vivo, 2013), allowing them to further recoup implementation costs from enhanced coastal resilience. This could see the relative ROI of ANR approaches outweighing that of NR alone at levels similar to those reported here at USD $25.00 Mg CO₂e prices: ∼4.8- to 8.3-fold greater relative ROI under ANR with autochthonous soil carbon inclusion (BAS methodology) (Table 4). ANR project scenarios moreover have the potential to be designed to further enhance site-specific carbon sequestration (Bai et al., 2021; Rahman et al., 2021) and associated co-benefits (coastal protection: Duncan et al., unpublished data) through introducing greater species and taxonomic diversity at early rehabilitation stages. The limited number of and wide variation in empirical studies on carbon sequestration in NR abandoned ponds in SE Asia furthermore drove high uncertainty in emissions reductions under NR scenarios (see section “Materials and Methods”; Table 4). Further on-ground research to quantify carbon sequestration in NR rehabilitating mangroves and their drivers may serve to reduce uncertainty in potential emissions reductions and therefore potential buyer confidence in ex ante NR-generated blue carbon credits. At present, this uncertainty, combined with uncertainty in model-derived NR rates (Figure 4), triggered deductions in project scenario emissions under NR for all potential accreditation methodologies (5.01–11.28% deductions). Coupled with increased natural risk from soil erosion at the exposed site (5% additional non-permanence buffer applied; Verra, 2019), this could sum to substantial perceived project risk to credit-buyers. Instead, blue carbon project developers could seek to increase potential project viability in terms of credit sales by prioritizing reduced uncertainty (ANR rehabilitation strategies) over slow and variable NR alone. Here, a combination of perceived risk reduction (i.e., lower discounting) alongside faster rates of mangrove greenbelt regeneration may attract substantial interest from other relevant industry investors to drive up credit pricing and realized ROI (see Beeston et al., 2020; Sumaila et al., 2021). This reduction of credit-buyers’ perceived risk with ANR rehabilitation efforts may also be combined with less costly conservation actions (i.e., avoided deforestation/degradation) to maximize both pricing and carbon sequestration in larger-scale blue carbon project planning (see Mikoko Pamoja, 2020; Conservation International, 2021; The Blue Natural Capital Financing Facility, 2021).

There remain other substantial practical considerations that should be factored in to the development of blue carbon projects beyond those that could be explored in our case study. First, we conducted case study-specific non-permanence risk assessment (Verra, 2019), where low exposure in coastal configuration (adjacent island) reduces risk from highly frequent typhoon events in the country, resulting in moderate non-permanence credit buffers (30–35%). Natural disaster risk, and others such as political risk, in other potential sites may exceed certification thresholds (Verra, 2019) or mean that non-permanence buffers drastically restrict potential ROI without active intervention to reduce stochastic threats to permanence in susceptible regions. Second, our study considered low-intensity, sea-facing abandoned ponds for which the need for physical rehabilitation interventions to reinstate hydrology are minimal (natural storm-driven dike breaching: Primavera et al., 2014). On-site physical intervention costs on initiation to make projects additional (Verra, 2020b) (e.g., dike breaching in less exposed sites and/or additional hydrological interventions) for NR rehabilitated-oriented projects in other contexts may be substantial (Su et al., 2021). With potential physical intervention costs applied, our hypothetical 9.68 ha) case study project would not be financially viable under either rehabilitation strategy at current voluntary carbon market prices (USD $9.70 Mg CO2e^–1: Forest Trends’ Ecosystem Marketplace, 2021; see also Thompson et al., 2014). Furthermore, costs to both NR and ANR blue carbon project strategies may increase substantially where community engagement for project activities is low or absent (Primavera et al., 2012b). Variation in such project-specific biophysical costs may tip the balance of relative ROI between rehabilitation strategy options (NR vs. ANR), and high costs may in some cases make all options prohibitively expensive. Most importantly, our quantification of project costs is simplistic, focusing only on biophysical costs. In reality, any blue carbon project requires large budgets for certification standard fees (registration and verification) and any technical capacity not held by the proponent organization. These costs can be substantial (∼USD 40,000 for our hypothetical case study: Verra, 2020c), and realization of an operational blue carbon project, let alone sufficient financial ROI, is often not possible without early impact or seed investment from a non-credit purchasing body (see Mikoko Pamoja, 2020; The Blue Natural Capital Financing Facility, 2021). Put together, these cost and risk context-dependencies invite application and extension of the approaches developed herein to other and wider mangrove blue carbon rehabilitation contexts, urge caution to prospective blue carbon project developers and conduct of driver analyses such as performed here to evaluate real-world project scenario options.

With the arrival of the approved blue carbon Verified Carbon Standard methodology (Verra, 2020a,b) we are likely to see a rapid increase in demand for blue carbon credits in the coming years. To maximize blue carbon’s potential to advance global CCMA efforts and unlock sustainable conservation financing, it will now be imperative to evaluate the economic viability of potential blue carbon project scenario options prior to embarking on lengthy and costly project registration processes. With this study, we provide an approach to quantify the ROI and viability of potential rehabilitation-oriented blue carbon project scenarios, using projected biophysical project costs and multi-site scale ecological models to predict rates of transactional credit generation against likely biophysical cost. We employ a case study in which biophysical rehabilitation intervention requirements are minimal, revealing potentially substantial carbon sequestration and reduced uncertainty in such moderate-scale ANR rehabilitation blue carbon projects now able to incorporate significant soil carbon compartments. While ANR rehabilitation costs in other contexts are highly variable and can be considerable (Su et al., 2021), our case study is representative of coastal mangrove rehabilitation opportunities across large parts of SE Asia where coastal aquaculture and pond abandonment is extensive (Richards and Friess, 2016; Goldberg et al., 2020). We thus highlight a potential opportunity for additional ANR blue carbon projects in SE Asian countries such as the Philippines, where much former mangrove extent has already been cleared, rates of loss are decreasing (Spalding et al., 2010; Goldberg et al., 2020), national mandates forbid further unplanned deforestation (Primavera et al., 2012a), and coastal greenbelt rehabilitation for CCMA is high on the political agenda. Instead, ANR rehabilitation-oriented blue carbon strategies in abandoned, sea-facing ponds under non-private ownership, and for which tenure reversal and political will strengthening are the limiting factors to additionality, represent a widespread, comparatively low-cost CCMA opportunity. Generating the co-benefit of reducing perceived risk to credit-buyers and more rapidly strengthening coastal greenbelts in typhoon-prone regions, such a strategy could moreover attract higher credit pricing and advance national CCMA mitigation strategies, where ANR strategies could have substantially greater (∼4.8- to 8.3-fold) relative ROI than NR alone. It is important to stress that any implemented ANR rehabilitation approach for blue carbon credit generation should employ scientifically-founded guiding principles, and that traditionally-performed widespread mangrove planting (inappropriate species, intertidal locations) is highly unlikely to produce successful blue carbon projects (Primavera et al., 2012b; Lee et al., 2019) or positive financial returns (as yet unstudied). Approaches such as those developed herein can guide optimal site-specific rehabilitation implementation options within such a strategy to maximize potential sustainable conservation financing.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

CD, JP, and HK conceived the research ideas. CD collected the data, performed analyses, and wrote the first draft of the manuscript. All authors contributed critically to wrote the final draft of the manuscript and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.