Mangrove reforestation provides greater blue carbon benefit than afforestation for mitigating global climate change

Blue carbon ecosystems, including mangroves, are recognized as important nature-based solutions for climate mitigation due to their significant carbon sequestration potential. Since the 1970s, global mangrove restoration has proceeded via two principal silvicultural pathways: reforestation (restoring mangroves where they previously existed) and afforestation (establishing mangroves in areas without prior mangroves). Each pathway involves trade-offs: reforestation may avoid conversion of other habitats but face land tenure issues, while afforestation on tidal flats can be cheaper but often suffers low seedling survival due to unsuitable hydrodynamics. Prior land use and local ecogeomorphic conditions likely shape restoration trajectories and carbon flux dynamics. The study poses the hypothesis that carbon accumulation trajectories differ significantly between reforestation and afforestation due to differences in preexisting versus novel site conditions, and tests this at a global scale using compiled field data.

Previous work on maximizing mangrove carbon sequestration has focused on species selection, planting strategies (natural regeneration vs. active planting, monoculture vs. mixed species, planting density), and climate influences on growth. Regional and local factors—hydrogeomorphic settings, nutrient availability, salinity, and prior land tenure/use—also strongly influence restoration outcomes. Studies indicate prior land use affects forest recovery trajectories; in mangroves, sites with historical hypersalinity or altered hydrology may show distinct sediment organic matter and nutrient dynamics post-restoration. Restoration in previously productive aquaculture ponds can yield higher biomass carbon than less productive sites. However, a comprehensive comparison of carbon benefits between reforestation and afforestation across global sites had been lacking.

The authors compiled carbon stock data from 106 peer-reviewed publications and databases, yielding 379 mangrove restoration sites worldwide, spanning Asia-Pacific, Indian Ocean regions, Africa, and the Americas, with sites ranging from 38°S to 28°N. Carbon pools analyzed included aboveground biomass carbon (AGC), belowground biomass carbon (BGC), and sediment carbon to 1 m depth (SCS). Stand ages for reforestation were mostly under 40 years; afforestation chronosequences extended up to ~80 years in some cases. Single-species plantings predominated. Data extraction included numeric values or digitized values from figures; biomass was derived via harvest or allometric methods, with standard carbon conversion factors (AGC 0.47, BGC 0.39) if study-specific values were not provided. Sediment carbon stocks were standardized to 1 m depth; for cores less than 1 m, the deepest measured layer’s density was extended to 1 m. Climate variables (mean annual precipitation and temperature) were extracted from WorldClim2 within 1 km. Restoration projects were classified by prior land use into reforestation (e.g., abandoned aquaculture/agriculture converted from mangroves, clear-felled areas, hydrologically altered mortality) and afforestation (e.g., mudflats, seagrass, saltmarsh, coral reef, denuded zones). Statistical analysis used linear mixed models with log(age), restoration type, and their interaction as fixed effects and region as a random effect, chosen by AIC; ANOVA evaluated significance. Age-specific comparisons used 5-year classes (20–40 grouped); Wilcoxon tests assessed differences when assumptions were unmet. Space-for-time paired control sites estimated sediment carbon increments; negative increments were corrected where 210Pb rates were available. Nonlinear growth models (Von Bertalanffy, Gompertz, Chapman–Richards) were tested, with Gompertz selected to model AGC, BGC, SCS, and total ecosystem carbon (TECS) trajectories; SCS and TECS included non-zero baselines. Global sequestration potential was estimated using Global Mangrove Watch deforestation (1996–2016), excluding irrecoverable areas (erosion, settlements). Country-specific feasible reforestation extents were calculated by loss drivers. Four restoration rollout scenarios (1-year completion; 5-year; 10-year; varying rates by country) were applied to convert modeled TECS increments to CO2-eq using 44/12 molecular weight conversion.

- AGC increased over time in both pathways, with reforestation showing a steeper increase, especially after 15 years (P≤0.001 for log(age) and interaction). - BGC accumulation did not differ significantly between pathways over 40 years. - SCS was nearly twice as high in reforestation vs. afforestation within 0–5 years (220.7 ± 38.5 vs. 108.9 ± 5.8 Mg C ha−1; P<0.01) and remained more than double at 20–40 years (293.4 ± 26.4 vs. 128.8 ± 14.4 Mg C ha−1; P≤0.001). Sediment carbon increments since restoration were also higher for reforestation. - TECS over the first 40 years was higher for reforestation (232.8 ± 48.0 to 407.0 ± 23.7 Mg C ha−1) than afforestation (119.2 ± 20.3 to 213.7 ± 30.7 Mg C ha−1; P≤0.05 interaction), driven largely by AGC and SCS. - Modeled 40-year per-hectare stocks: AGC 127.7 (110.7–144.7) vs. 88.7 (70.2–107.2) Mg C ha−1; BGC 38.7 (30.1–47.3) vs. 37.0 (22.9–51.2) Mg C ha−1; SCS 139.2 (136.4–142.1) vs. 65.0 (−3.7–133.7) Mg C ha−1 for reforestation vs. afforestation, respectively. - Annual SCS increments: 3.5 vs. 1.6 Mg C ha−1 yr−1 for reforestation vs. afforestation, within ranges observed for natural mangroves. - Reforestation sequesters ~60% more carbon per hectare than afforestation over 40 years. - Global feasible reforestation (biophysically constrained deforestation area: 614,467 ha) could yield 688.8 (624.8–752.7) Tg CO2-eq over 40 years if completed in 1 year; afforestation of the same area would yield ~259 Tg CO2-eq less. - Under slower rollout: 5-year scenario 687.2 (627.3–747.1) Tg, 10-year scenario 684.2 (631.5–736.9) Tg, varying-rate scenario 671.5 (637.1–706.0) Tg CO2-eq. - Top country contributions (1-year scenario): Indonesia 188.7 Tg, Mexico 110.4 Tg, Myanmar 61.3 Tg CO2-eq over 40 years. - Controlling factors: Reforestation sites typically had higher sediment TOC and TN, positioned higher in the intertidal zone with lower porewater salinity, promoting AGC and SCS; MAP influenced AGC similarly across pathways; salinity constraints likely reduced growth at lower intertidal afforestation sites. Species leaf chemistry (e.g., higher tannin/lignocellulose in Rhizophora spp. common in reforestation) may slow litter decomposition, enhancing SCS.

The findings support the hypothesis that reforestation delivers greater blue carbon benefits than afforestation, mainly via enhanced sediment carbon stocks and faster AGC development. Positioning in mid-to-upper intertidal zones, higher sediment fertility (TOC, TN), and lower salinity at reforestation sites favor growth and carbon preservation, while afforestation on lower intertidal flats faces hydrodynamic and salinity stresses that reduce biomass accumulation and carbon burial. Given that SCS dominates the difference, ensuring adequate hydrologic function and site selection is critical. Reforestation also avoids conversion of other valuable habitats (mudflats, seagrasses) and can be more cost-effective, strengthening its prioritization as a nature-based climate solution. While BGC patterns were similar across pathways, potential shifts in biomass allocation under salinity stress may explain this parity. The results imply that policies should prioritize reforestation of previously forested mangrove areas where biophysically feasible, integrate silvicultural practices that optimize AGC and SCS, and consider rotation ages for wood production in managed systems with caution due to emissions from harvest residues. Country-specific opportunities are substantial, particularly in Indonesia, Mexico, and Myanmar.

This global synthesis demonstrates that mangrove reforestation yields substantially greater carbon sequestration benefits than afforestation over the first 40 years, primarily through elevated sediment carbon accumulation and faster aboveground biomass growth. Prioritizing reforestation aligns climate mitigation goals with reduced habitat conversion conflicts and potential cost efficiencies. As global mangrove losses slow and some regions experience gains, optimizing silvicultural pathways becomes increasingly important for blue carbon strategies. Future work should broaden geographic representation (e.g., Africa, Australia, North America), extend beyond 40-year chronosequences, refine within-pathway classifications (e.g., hydrogeomorphic settings, sediment properties), and integrate socio-economic and land tenure constraints to estimate achievable, not just biophysically feasible, sequestration potential.

Site representation was uneven, with fewer observations from Africa, Australia, and North America; some available studies lacked key metadata (age, land-use history). The 40-year window captures common rotation ages but not century-scale dynamics. Reforestation sites were more tropical, while afforestation spanned tropical to subtropical zones, introducing climate covariates; MAP influenced AGC similarly across pathways, and BGC was affected by MAP but not pathway. Within-pathway variation (species, hydrology, sediment texture, elevation changes in rehabilitated ponds) can produce outcomes that deviate from mean trends. Sediment stock standardization to 1 m and extension of deepest measured layer may slightly overestimate SCS in some cases; space-for-time controls introduce uncertainties, partially corrected using 210Pb-derived rates. Socio-economic factors (land tenure, opportunity costs, policy frameworks) and local land values constrain achievable restoration; estimates considered only biophysical feasibility and deforestation since 1996, potentially undercounting earlier losses that could be reforested.