Blue carbon ecosystems, including mangroves, are crucial for climate change mitigation. Since the 1970s, significant mangrove restoration efforts have been undertaken globally, categorized as reforestation (restoring mangroves in previously colonized areas) and afforestation (establishing mangroves in new areas). While both methods aim to increase carbon sequestration, their relative effectiveness remains unclear at a global scale. Reforestation might avoid habitat conversion conflicts but faces land tenure issues. Afforestation on mudflats might be cheaper but faces lower seedling survival rates. This study addresses the knowledge gap by comparing the carbon sequestration potential of reforestation and afforestation at a global scale, aiming to inform effective mangrove restoration policies for climate change mitigation.

Previous research has explored mangrove restoration strategies focusing on tree species selection and restoration methods (natural processes vs. active facilitation, monoculture vs. mixed species, planting density). However, few studies have comprehensively assessed the impact of establishment location and silvicultural approaches on carbon sequestration. While climate factors influence mangrove growth, regional and local factors like ecogeomorphic settings and environmental conditions also play crucial roles. Prior land use significantly impacts restoration trajectories in terrestrial forests, and a similar influence is hypothesized for mangroves. Studies suggest that mangroves restored in previously productive sites (e.g., aquacultural ponds) might exhibit higher biomass carbon sequestration rates than those in less productive sites. This study builds upon this existing research by conducting a large-scale comparative analysis of carbon accumulation in reforestation and afforestation projects.

This study compiled data on standing stocks of aboveground biomass carbon (AGC), belowground biomass carbon (BGC), and sediment carbon (SCS) from 379 mangrove restoration sites across the globe. Data were extracted from 106 scientific publications, covering various mangrove regions (Asia Pacific Ocean, Asia Indian Ocean, Africa Indian Ocean, American Atlantic Ocean, and American Pacific Ocean). Sites spanned a latitudinal range of 38°S to 28°N, and mangrove age ranged up to nearly 80 years for afforestation projects. A four-step critique process was used to screen articles based on data availability (at least one of the four mangrove carbon pools), forest age specification, prior land use description, and other criteria. Reforestation sites included those previously used for agriculture, aquaculture, or affected by deforestation, while afforestation sites were those where mangroves had not previously existed, such as mudflats or seagrass beds. Where necessary, data was digitized from graphs. Biomass carbon data were obtained using harvesting or allometric methods, and a standard carbon conversion factor was applied when necessary. Sediment carbon density was calculated for the top 1 m. Climate data (mean annual temperature and precipitation) were sourced from WorldClim datasets. Linear mixed models were used to analyze the influence of restoration type and time on different carbon pools, using age and restoration type as fixed factors and restoration region as a random factor. Nonlinear models were also fitted to explore carbon accumulation trajectories. The global carbon sequestration potential was estimated using different scenarios of reforestation completion rates (1 year, 5 years, 10 years, and varying rates across countries).

Over the first 40 years, both reforestation and afforestation projects showed gradual increases in AGC, but reforestation sites exhibited a larger increase, especially after 15 years. No significant difference was found between reforestation and afforestation for BGC increases. However, SCS density in reforestation sites was almost double that of afforestation sites throughout the study period, significantly contributing to greater total ecosystem carbon stocks in reforestation projects. Reforestation supported 232.8–407.0 Mg C ha⁻¹ and afforestation supported 119.2–213.7 Mg C ha⁻¹ over 40 years. Mean annual precipitation positively influenced AGC in both types of projects, but it wasn't the primary reason for the difference in carbon sequestration. Reforestation sites had greater initial sediment TOC and TN, and these properties remained higher throughout succession. Reforestation sites were mainly located in higher intertidal zones, while afforestation sites were in lower intertidal zones, potentially contributing to differences in salinity and sediment composition. Higher salinity in afforestation sites likely reduced seedling growth and carbon accumulation. Higher aboveground biomass in reforestation sites led to increased leaf litter input and sediment carbon accumulation. Differences in leaf biochemistry between species used in reforestation and afforestation may also have influenced litter decomposition rates and sediment carbon accumulation. Globally, reforestation was projected to sequester 60% more carbon per hectare than afforestation over 40 years. Under a scenario where all feasible reforested areas are restored within 1 year, the cumulative carbon sequestration potential over 40 years is estimated at 688.8 Tg CO2-eq, about 259 Tg CO2-eq higher than afforestation. Indonesia, Mexico, and Myanmar are projected to have the greatest carbon mitigation potential through reforestation. Restoring all potentially available mangrove areas through reforestation would increase annual CO2 uptake by 4.3–5.1%.

This study's findings highlight the significant advantage of mangrove reforestation over afforestation in terms of blue carbon sequestration. The superior performance of reforestation is mainly attributed to favorable site conditions (intertidal position, nutrient availability, salinity) and enhanced sediment carbon accumulation. These results emphasize the importance of considering site selection and silvicultural approaches when designing nature-based solutions for climate change mitigation. The significantly higher carbon sequestration potential of reforestation suggests that prioritizing reforestation in previously deforested areas could substantially enhance the climate change mitigation potential of mangrove restoration efforts. The study's findings directly address the need for optimized mangrove restoration strategies and offer valuable insights for policy-makers and practitioners. Avoiding habitat conversion associated with afforestation also supports the prioritization of reforestation.

This global analysis demonstrates that mangrove reforestation on previously degraded sites significantly outperforms afforestation in carbon sequestration. Prioritizing reforestation could substantially enhance blue carbon sink strength and contribute to more effective climate change mitigation. Future research should focus on refining regional estimates, incorporating socioeconomic factors into restoration planning, and investigating the long-term dynamics of carbon accumulation in both reforestation and afforestation projects.

While the study included a large dataset of mangrove restoration sites, data availability was limited for certain regions (Africa, Australia, North America). The concentration of reforestation sites in tropical zones and the scattered distribution of afforestation sites might lead to climate as a covariate. Within-pathway variation in silvicultural practices might lead to misestimations of carbon sink effect. The study focused on carbon storage benefits but did not fully consider socioeconomic factors and land values that might influence the feasibility of reforestation projects.