Importance of mangrove plantations for climate change mitigation in Bangladesh

Mangroves have been identified as blue carbon ecosystems that are natural carbon sinks. In Bangladesh, the establishment of mangrove plantations for coastal protection has occurred since the 1960s, but the plantations may also be a sustainable pathway to enhance carbon sequestration, which can help Bangladesh meet its greenhouse gas (GHG) emission reduction targets, contributing to climate change mitigation. As a part of its Nationally Determined Contribution (NDC) under the Paris Agreement 2016, Bangladesh is committed to limiting the GHG emissions through the expansion of mangrove plantations, but the level of carbon removal that could be achieved through the establishment of plantations has not yet been estimated. The mean ecosystem carbon stock of 5–42 years aged (average age: 25.5 years) mangrove plantations was 190.1 (±30.3) Mg C ha−1, with ecosystem carbon stocks varying regionally. The biomass carbon stock was 60.3 (±5.6) Mg C ha−1 and the soil carbon stock was 129.8 (±24.8) Mg C ha−1 in the top 1 m of which 43.9 Mg C ha−1 was added to the soil after plantation establishment. Plantations at age 5 to 42 years achieved 52% of the mean ecosystem carbon stock calculated for the reference site (Sundarbans natural mangroves). Since 1966, the 28,000 ha of established plantations to the east of the Sundarbans have accumulated approximately 76,607 Mg C year−1 sequestration in biomass and 37,542 Mg C year−1 sequestration in soils, totaling 114,149 Mg C year−1. Continuation of the current plantation success rate would sequester an additional 664,850 Mg C by 2030, which is 4.4% of Bangladesh's 2030 GHG reduction target from all sectors described in its NDC, however, plantations for climate change mitigation would be most effective 20 years after establishment. Higher levels of investment in mangrove plantations and higher plantation establishment success could contribute up to 2,098,093 Mg C to blue carbon sequestration and climate change mitigation in Bangladesh by 2030.


| INTRODUC TI ON
Mangroves have been identified as blue carbon ecosystems that are natural carbon sinks Murray et al., 2011). These carbon rich ecosystems store carbon in both aboveground and belowground carbon pools, and have a mean global ecosystem carbon stock of 939 Mg C ha −1 (range 856-1023 Mg C ha −1 ) of which 49%-98% is stored in the soil (Alongi, 2012;Donato et al., 2011;Kauffman et al., 2020;Sanderman et al., 2018).
Management of coastal ecosystems such as mangroves can be an important component of climate change mitigation strategies (Arifanti et al., 2022;Macreadie et al., 2017Macreadie et al., , 2021Mcleod et al., 2011).
While mangroves cover a small part of the earth's surface (occupying only 0.5% of the global coastal zone), they have high levels of carbon stocks and sequestration per hectare. Unfortunately, 30%-50% of the world mangroves have already been lost and degraded due to human activities in the coastal zone (e.g., coastal aquaculture) (Giri et al., 2008;Hamilton & Casey, 2016;Lovelock et al., 2015).
Degradation of mangroves caused by land use conversion, over exploitation, impoundment and pollution has been widespread (Adame et al., 2021;Goldberg et al., 2020;Lovelock et al., 2018), which has contributed to CO 2 emissions and the reduction of carbon sequestration capacity (Adame et al., 2021;Pendleton et al., 2012). In response to the loss of these valuable ecosystems and because of their importance in carbon sequestration and other ecosystem services, many countries like Bangladesh have created mangrove plantations (Chowdhury, 2018;Saenger & Siddiqi, 1993).
The establishment of mangrove plantations (mangroves established from seedlings planted on newly accreted coastal lands where there was no previous vegetation or forests) and their sustainable management is a pathway that can help countries reduce loss in mangrove cover and enhance carbon sequestration that contributes to climate change mitigation while contributing to additional ecosystem services Herr et al., 2017;Taillardat et al., 2018).  (Friess et al., 2020;Herr & Landis, 2016;Martin et al., 2016;Taillardat et al., 2018).
Furthermore, mangroves have also been considered as an important ecosystem in climate change mitigation strategies within voluntary carbon markets and by Reducing Emissions from Deforestation and Forest Degradation (REDD+) initiative in developing countries (Friess, 2013;IPCC, 2007;Locatelli et al., 2014;MacKenzie et al., 2021). A range of countries have committed to reducing mangrove degradation and investing in mangrove conservation and restoration to achieve their NDC targets (Arifanti et al., 2022;Herr & Landis, 2016;Martin et al., 2016). Halting degradation of mangroves and enhancement of restoration are likely to be particularly significant pathways for reducing greenhouse gas (GHG) emissions in countries like Bangladesh that have large mangrove areas and relatively low emissions in other sectors (GOB, 2015;Sasmito et al., 2019;Taillardat et al., 2018).
In Bangladesh, blue carbon ecosystems cover 2.1 million ha (Mha), which includes mangroves, seagrasses, and saltmarshes. The natural mangroves (those that occur naturally and are not planted) comprise 0.44 Mha or 21% of Bangladesh's total blue carbon resources (Chowdhury et al., 2015). Levels of ecosystem ("biomass and soil") blue carbon stocks in natural mangroves of Bangladesh range between 115 and 256.6 Mg C ha −1 (Ahmed et al., 2017;BFD, 2015;Kamruzzaman et al., 2018;Majumder et al., 2019;Mitra et al., 2011;Nellemann & Corcoran, 2009). In addition to natural mangroves, the Bangladesh government has established mangrove plantations on newly accreted land since 1966 (Saenger & Siddiqi, 1993;Uddin et al., 2014Uddin et al., , 2022. The Delta experiences severe cyclones and tidal surges, which caused extensive loss of human lives and property.
The damage due to the cyclones was lower in the West coast of the Delta because of the presence of the Sundarbans natural mangroves.
Because of the coastal protection benefits of the Sundarbans natural mangroves, the Bangladesh Forest Department (BFD) took the initiative to plant mangrove species on newly accreted coastal land located to the east of the Sundarbans, which are characterized by their exposure to the open sea of the Bay of Bengal and the absence of coastal vegetation until the 1960s. In addition to coastal protection, the BFD has committed to establishing new plantations as a part of the Bangladesh Climate Change Strategy and Action Plan (BCCSAP) and the National Adaptation Plan (NAP) along with the NDC commitment (MoEFCC, 2009(MoEFCC, , 2022USDS, 2014). However, the contribution of plantation mangroves to blue carbon sequestration and climate change mitigation has not yet been assessed for the whole Bangladesh Delta. Therefore, the aim of this research was to evaluate the contribution of plantation mangroves to blue carbon sequestration and climate change mitigation in Bangladesh.
Mangroves occur along the 710 km coast of the Bangladesh Delta (CZPo, 2005;Moslehuddin et al., 2015;Sarwar, 2005). Based on ecological and geomorphic characterisation, the Bangladesh Delta has been divided into three main coastal regions-East, Central and West (Uddin et al., 2022). The West coast of the Delta is comparatively stable and colonised by the Sundarbans natural mangroves (Hossain et al., 2009;Iftekhar & Islam, 2004). In contrast, the geomorphology of the Central and East coasts has been more dynamic in recent decades, with high rates of sediment accretion delivered in the Ganges-Brahmaputra River system that have led to the formation of new islands (Rahman, 2012;Sarwar & Woodroffe, 2013) where the BFD has established mangrove plantations (Uddin et al., 2022).
The plantations were established between the built embankment that protects human settlements and agriculture and the sea or the river's edge, giving rise to plantations over different inundation zones (Choudhury et al., 2004;Uddin et al., 2014). Variation in geomorphology and plantation structure over geomorphic settings (e.g., East, Central and West coast) and intertidal zones (e.g., Landward, Middle and Seaside plantation) across the Delta (Uddin et al., 2022) may influence the rate of blue carbon sequestration Hayes et al., 2017;Kauffman et al., 2014), and hence give rise to variation in climate change mitigation offered by the established mangrove plantations.
While there has been a growing interest in planting mangroves to offset carbon emissions, there is a limited understanding of whether plantation mangroves contribute to blue carbon services that are similar to those of natural forests. In Indonesia, the carbon sequestered in 10 years old restored mangrove biomass and soil were similar (biomass: 16.2-16.6 Mg C ha −1 year −1 ; soil: 7.3-12.2 Mg C ha −1 year −1 ) to that of natural mangroves in Perancak estuary of Bali (Sidik et al., 2019). The rate of soil carbon accumulation (2.18 Mg C ha −1 year −1 ) in created mangroves in the Tampa Bay of Florida over 20 years of restoration was similar to the global mean soil carbon sequestration (2.26 Mg C ha −1 year −1 ) of natural mangroves (Osland et al., 2012). In Vietnam, the mean of total ecosystem carbon stocks (889 ± 111 Mg C ha −1 ) in planted mangroves of the Can Gio Mangrove Biospheres Reserve (CGMBR) were similar to that of naturally regenerated mangroves (844 ± 58 Mg C ha −1 ) in The Kien Vang Protection Forest (KVPF) after 35 years (Nam et al., 2016).
These previous assessments of carbon stocks and rates of C sequestration in restored or planted mangroves suggest that plantations can perform similarly to natural forests .
However, there are few assessments of carbon accumulation by afforested mangroves. We evaluated the carbon stocks and sequestration rates within the planted mangroves of the whole Bangladesh Delta. We also assessed variation in carbon stocks among the different regions of the Delta. We hypothesized that plantation mangroves in Bangladesh have a similar role in carbon sequestration and attain similar carbon stocks to that of natural mangroves in the same region (represented by a reference site). Finally, we estimated the likely contribution of mangrove plantations to achieving Bangladesh's emission reduction targets, assuming different levels of commitment by the BFD to creating mangrove plantations.
Our sites are located within the Bay of Bengal and the Ganges, Brahmaputra and Meghna River (GBM) networks and estuaries that span three coastal regions-the East, Central, and West coast of Bangladesh, that are differentiated by the levels of sediment accretion, tidal inundation, and salinity (Moslehuddin et al., 2015;Siddiqi, 2001).
The geomorphology of the region is highly dynamic and unstable due to continuous erosion, accretion, cyclonic attack and tidal surges, but the level of geomorphic change varies among the coastal regions (Siddiqi, 2001). The study area in the West coast and the reference site (on the eastern side of the natural mangrove of the Bangladesh Sundarbans) of the Delta is comparatively stable compared to the Central and East coast (Alam & Uddin, 2013;Rahman, 2012;Sarwar & Woodroffe, 2013).
The climate of the Delta is humid, with the mean daily tempera- Heritiera fomes (Barik & Chowdhury, 2014;Uddin et al., 2022). In the Central and East coasts, natural mangroves are rarely present. In these regions, the plantation mangroves are comprised of the species planted by the BFD (Sonneratia apetala and Avicennia officinalis) as well as six other species that were natural recruits, dispersed into the plantations from the Sundarbans natural mangroves. In addition to differences in diversity between the reference sites and the plantations, the plantations tended to have higher tree densities (Uddin et al., 2022).
There are approximately 28,000 ha of planted mangroves established (to the east of the Sundarbans natural mangroves) by the BFD in the Delta since 1966, using two native pioneer species-S. apetala and A. officinalis (Saenger & Siddiqi, 1993;Uddin et al., 2022), while approximately 200,000 ha of mangrove planting has been done by the BFD within the same period (BFD, 2018). These species were chosen for plantations because they are rapidly growing pioneer species and because they have a high tolerance of salinity (Saenger & Siddiqi, 1993).

| Study design
In this study, we assessed three carbon pools (aboveground and belowground biomass of trees, saplings, and mangrove soils), which represent the largest carbon pools in our study systems. We sampled trees and saplings from 72 nested circular plots (each plot consists of three sub-plots; total number of subplots: 216) from the plantations of ages 5-42 years in three geomorphic regions (24 plots for each region-East, Central and West) and 15 plots (total number of subplots: 45) from the natural mangroves (located in the West coast) in the Bangladesh Delta (Uddin et al., 2022). The mangrove plantations were established between the embankment and the sea (the Bay of Bengal) or river edge. The whole plantation, between the embankment and the sea or river edge, was divided into three intertidal zones called the Landward, Middle and Seaside plantations (Uddin et al., 2014). Tree and sapling attributes were assessed from 10 and 3 m radius plots, respectively, at selected sampling sites in 2020. Measurement and reporting of carbon stocks in mangrove forests followed the methods of Kauffman and Donato (2012).
Sampling of soils (from 216 subplots for plantations and 45 subplots for reference sites in natural mangroves) for assessing soil organic carbon (SOC) stocks was done by extracting uncompressed cores near the centre of each subplot using a 1-m long open-faced gouge auger. We collected our plot level field data for vegetation and soil from January to June 2020. and Chave et al. (2005) equations to those developed for Asia-Pacific mangrove species (Chave et al., 2005;Komiyama et al., 2008) indicated that Mahmood et al. (2019) and Chave et al. (2005) estimates were conservative and applicable for our sites and over a wide range of DBH ( Figure S1). We considered mangrove stems as trees which achieved a DBH at least 10 cm and above, whereas mangrove stems with DBH less than 10 cm and more than 1.37 m height were considered as saplings (Uddin et al., 2022).

| Calculation of ecosystem biomass carbon and SOC
Belowground biomass (BGB) was estimated using the equations from Komiyama et al. (2008) (Table S1). The AGB and BGB (in kg) of trees and saplings were then multiplied by 0.50 for stems and 0.39 for roots for the conversion of biomass to mass of carbon . The site average wood density (ρ, g cm −3 ) was estimated as 0.752 g cm −3 .
Assessment of soil carbon stocks to 1 m depth followed standard methods (Howard et al., 2014). We collected 216 cores (from 216 subplots) of soil samples from the established planted mangroves across the three coastal regions of the Bangladesh Delta.
We also collected 45 cores (from 45 subplots) of soil samples from the natural mangroves (reference site) located in the West coast of the Delta (Figure 1). The soil was sampled at intervals down core  with five subsoil depths (0-10 cm, 10-30 cm, 30-50 cm, 50-75 cm, 75-100 cm). We considered the subsamples F I G U R E 1 The study area showing the sampling plots for vegetation and soils in plantations (grey circles) and natural mangrove stands (yellow circles). The area that consists of mangroves is shown in the legend as "Mangroves" (both natural and plantations). There were 72 plantation plots and 15 natural reference mangrove plots.
at the approximate midpoint of each sample depth. Soil samples were oven dried at 60°C at the Chemistry Lab of the Institute of Forestry and Environmental Sciences, Chittagong University (IFESCU), until the attainment of constant dry mass. The dried samples were then ground and sieved through a 2 mm wire mesh.
Bulk density (BD, g cm −3 ) was determined by dividing the oven-dry soil sample mass (g) by the volume (cm 3 ) of the sample. We used the loss on ignition (LOI) method at a temperature of 550°C to obtain %LOI (Dean, 1974). (1 m), we summed the carbon in each soil segment (five segments) (Hayes et al., 2017). Soil OC stock for each soil segment was measured as SOC stock = BD × SDI × %OC, where BD is the bulk density (g cm −3 ), SDI is soil specific depth interval (cm) of the sample, and %OC is the percentage of organic carbon within the given sample.

| Estimating rates of carbon sequestration
The rates of carbon sequestration in biomass and soils of plantations were estimated based on the accumulation of biomass and organic carbon over a plantation chronosequence and the change in the area of plantations over time. The accumulation of biomass and soil carbon in plantations of known ages (between 5 and 42 years) was established using regression models from plot inventories. Accumulation of biomass carbon over time was modelled as an exponential function of the form y = ae bx (where a is the intercept, b is the slope and x is time in years), which is widely used for ecology data with nonlinear trends (Matula et al., 2015;Oddi et al., 2019). Furthermore, exponential models were frequently used for plantation growth data (e.g., biomass), although nonlinear growth models are flexible enough to account for varying growth rates (Paine et al., 2012). To calculate biomass carbon sequestration rate, firstly, biomass carbon stocks from 1966 to 2016 (e.g., 50 years) for individual coastal regions (e.g., East, Central, and West) were calculated from the fitted exponential equations (Table S2) based on plot biomass over plantations of different ages (e.g., 5, 6, 10, 11, 13, 19, 20, 24, 26, 30, 35, 39, and 42 years) recorded from the BFD field offices in 2020. Annual increments in biomass carbon for each coastal region were calculated as the difference between carbon stocks of consecutive years. Deltaic sediments contain organic carbon from the terrestrial environments.
Following the approach of Needelman et al. (2018), we deducted this baseline organic carbon from the plantation carbon stocks to calculate SOC attributable to the mangrove plantations. To calculate SOC sequestration rate, we estimated mean SOC stock for ecosystem age of 1-50 years (e.g., 1966-2016) for each of the three coastal regions (e.g., East, Central, and West) using regression equations developed from our plot level datasets (Table S3).
Mean accumulation of SOC (soil carbon sequestration rate) was calculated as the difference between SOC at each respective year and SOC prior to mangrove planting, divided by the age of the plantation (Equation 2).
where SOC n is the soil carbon stock at age n year, SOC 0 is the soil carbon stock at age zero (before plantation establishment), t n − 0 is the plantation total age after initial planting.
We considered baseline SOC stocks of 102.4, 71.5 and 115.5 Mg C ha −1 for the East, Central, and West coasts, respectively, for calculating regional rates of SOC sequestration from 1966 to 2016 (Table S3).

| Scaling up carbon sequestration in the plantations to 2016
We calculated regional (East, Central, and West) ecosystem carbon stock of biomass and SOC by multiplying aerial extent of each region, which were estimated from the Global Mangrove Watch dataset (https://data.unep-wcmc.org/datas ets/45) using ArcGIS Desktop 10.8 software (Uddin et al., 2022) by the mean carbon stocks for each region. Estimates of plantation success (%) were calculated from the regional successful area established per year (calculated from the Global Mangrove Watch dataset) divided by the regional planting area per year (Uddin et al., 2019) multiplied by 100 (Table S4).

| Estimating future mangrove carbon stocks by 2030 (2017-2030)
We calculated future stand biomass carbon and SOC (2017-2030) for existing and new plantations in each region. For existing plantations, we had no data describing the frequency distribution of plot ages (i.e., the plantation area in each plot age class was unknown), therefore, to estimate C sequestration, we assumed that in each region there was an equal area of plantations in each plot age class, which we modeled would accumulate biomass C on a growth trajectory to achieve biomass equivalent to the 40-year old plantations, following biomass accumulation observed in the chronosequence plots (Table S2), after which biomass accumulation was assumed ceased, in order to avoid estimating biomass accumulation beyond the range of our dataset. For SOC accumulation, a linear function was used to estimate SOC sequestration for the area of plantations in each region based on observations from the chronosequence plots (Table S3).
To estimate the potential for new mangrove plantations to sequester additional carbon in the future with a range of investments in plantation establishment, we calculated carbon sequestration by 2030 in three scenarios: (1) current rate of regional planting (East: 1018 ha year −1 ; Central: 2435 ha year −1 ; West: 554 ha year −1 , established from information published by Uddin et al., 2019, Table S4) and current rate of regional success at 6.8% for the East, 10.5% for the Central and 42.5% for the West coast; (2) current rate of planting and an increase in establishment success to 30% for the East and the Central coast region; and (3) current rate of planting and a maximum of 50% establishment success for all coastal regions, which was assumed the highest success rate for plantations (Bayraktarov et al., 2016). We used the biomass growth trajectory from the chronosequence plots to estimate biomass accumulation for each successful annual planting cohort and a linear function to estimate SOC accumulation for the different scenarios. We predicted aerial extent until 2030 by the following equation (Equation 3), assuming continuation of the present rate of regional planting and by varying rates of regional planting success under three scenarios (current rate of regional success, success rate increased to 30% and 50%).

| Data analysis
We tested the normality of our datasets (Gaussian normal distribution, α = .05) using Shapiro-Wilk normality test methods before statistical analysis of the data. We used analysis of variance to assess variation in different C stocks where the main effects were coastal region and intertidal zone. Uncertainty for each carbon pool and total stand level carbon stock was calculated as a percentage (%) of the mean at 95% confidence interval following the method of Kauffman and Donato (2012). We used regression analyses to describe trajectories of carbon accumulation in biomass and soil carbon across the plantation ages. To assess differences in carbon stocks over time and among coastal regions and intertidal zones and their interactions, we used two-way analysis of covariance where the main effects were coastal region and intertidal zone, and covariate was plot age. We used R packages (R version 4.1.0) "tidyverse," "ggplot2," "rstatix," "broom," "caret," and "car" for data analysis.

| Variations in carbon stocks across the coastal regions and intertidal zones
We found variations in carbon stocks for all carbon pools-AGB and BGB of trees, saplings and mangrove soils across the coastal regions (Table 1). Mean biomass carbon stock for trees and saplings over the whole delta, including plots of all ages, was 48 (±3.8) Mg C ha −1 and 12.3 (±2.6) Mg C ha −1 , respectively (total biomass carbon was 60.3 ± 5.6 Mg C ha −1 ; uncertainty: ±19%) ( Table 1; Table S5). In the top 1 m of soils, the mean carbon stock was 129.8 (±24.8) Mg C ha −1 (uncertainty ±38.2%) of which 43.9 Mg C ha −1 was added to the soil as a result of mangrove plantation establishment (i.e., higher than the background stocks). The soil carbon stock (1 m depth) was 2.2 times higher than the total biomass carbon stock (60.3 Mg C ha −1 ) ( Table 1). The mean whole ecosystem carbon stock was higher in the West coast (243.7 ± 30.4 Mg C ha −1 ; uncertainty: ±24.9%), which was approximately 1.4 times higher than the East coast (174.5 ± 33.1 Mg C ha −1 ; uncertainty: ±37.9%) and 1.6 times higher than the Central (152 ± 27.5 Mg C ha −1 ; uncertainty: ±36.2%) region.
Of the total ecosystem carbon stock (190.1 ± 30.3 Mg C ha −1 ; uncertainty: ±26.8%), ~68% was the carbon stock from the top 1 m of soils (Table 1; Table S5). Mean ecosystem carbon stock in the plantations (over all plot ages) was lower than the mean ecosystem carbon stock of the Sundarbans natural mangroves of Bangladesh, which was 369.1 ± 41.8 Mg C ha −1 (uncertainty: ±19%) for the reference site (Table 1; Table S5). SOC in the top 1 m of soil was 2.4 times higher in the reference site (305.7 ± 35.2 Mg C ha −1 ; uncertainty: ±23%) than in plantations (129.8 ± 24.8 Mg C ha −1 ). We found that the mean carbon stock of all plantations (age range 5-42 years) achieved ~52% of the mean ecosystem carbon stock calculated for the reference site (Table 1). (3) Regional area (ha) = Regional average planting area per year × success rate.

TA B L E 1
Mean carbon stock of trees, saplings, and the top 1-m soils of plantation mangroves across the three coastal regions of Bangladesh over all plantation ages.

| Carbon stocks over plantation ages
Our regression model showed that regional biomass carbon stocks significantly (p < .05) increased with plantation age, but soil carbon stock accumulation was variable particularly in the east region (Figure 2a-d; Tables S2 and S3). The whole ecosystem carbon stock in each coastal region increased with plantation age (Figure 2a-c).
Whole ecosystem carbon stock (biomass + SOC stock) in plantations of all coastal regions (East, Central, and West coast) did not achieve similar carbon stocks to the Sundarbans natural mangroves after 40 years when compared with the natural reference site (Figure 2a-d).
Analysis of covariance test found that biomass carbon significantly increased with plantation age (F 1,39 = 11.56, p = .0016).

| Accumulation of carbon in the plantations since 1966
Accumulation of total carbon (Mg C) in plantations was higher in  554 ha year −1 ) and the regional rate of success (current rate) at 6.8% for the East, 10.5% for the Central and 42.5% for the West coast respectively, and a scenario of increased success rate at 30% (for East and Central coast) and 50% for all coastal regions (Table 6).  (Table 6). The reason of the higher carbon stocks and rates of the West coast of the Delta is the higher plantation establishment success rate in this region (i.e., 42.5% success) (Table S4).

| Predicted carbon stocks by 2030-Existing and new plantations
With annual rates of regional planting and current success, the total aerial extent of new plantations in the three coastal regions by 2030 is expected to be ~969 ha for the East coast, ~3579 ha for the Central coast and ~3296 ha for the West coast ( TA B L E 5 Future contribution (carbon stocks) of existing and new plantations to CO 2 removal by 2030 (2017-2030) at current regional rate of planting and plantation success at ~6.8%, 10.5% and 42.5% for the East, Central, and West coast respectively. Adopting this ambitious management option by the BFD, the ecosystem carbon sequestration rate is expected to be approximately nine times higher (24,523 Mg C year −1 ) in the East coast, four times higher (59,946 Mg C year −1 ) in the Central coast and 1.2 times higher (16,349 Mg C year −1 ) in the West coast by 2030 when compared with current management scenario (Table 6).

| Contribution of afforested mangroves to carbon stocks
Ecosystem carbon stocks in afforested plantation mangroves of the Delta were comparable to planted mangroves in other mangrove rich deltas in Asia (Table 7), but lower than those in other geomorphic settings, reflecting lower mangrove biomass carbon per hectare compared to tropical humid regions (Simard et al., 2019) and lower soil carbon, likely due to high levels of sediment supply from the Ganges and Brahmaputra Rivers (Sarwar & Woodroffe, 2013 Kauffman et al., 2020;Murdiyarso et al., 2021;Rahman et al., 2015). Our findings indicate that total carbon stocks of planted mangroves in the Delta were ~37% of the carbon stock of the IPCC Tier 1 default total ecosystem carbon stock (511 Mg C ha −1 ) for natural mangroves (Hiraishi et al., 2014;Kauffman et al., 2020). However, whole ecosystem carbon stocks calculated in this study (~190.1 Mg C ha −1 ) are within the range of global carbon stocks reported for natural mangroves (79-2208 Mg C ha −1 ) (Kauffman et al., 2020), but lower than the reference site (52% of Sundarbans natural mangroves) (this study: ~369.1 Mg C ha −1 ) and on the lower edge of the global range (Kauffman et al., 2020). The lower carbon stock in Bangladesh plantations might be linked to the environmental conditions of the plantations that do not support the same level of productivity and carbon accumulation as natural forests in the Sundarbans, which tend to be less saline and less exposed to wind, waves, and currents (Rahman et al., 2015).

| Accumulation and variations of carbon stocks across coastal regions and intertidal zones
Variations in mangrove carbon stocks across the coastal regions of the Delta are influenced by natural (e.g., climate) and anthropogenic (e.g., land use change) factors (Kauffman et al., 2020). Climate, vegetation structure, salinity regime, geomorphology and land use change may influence carbon stocks, resulting the variations across the regions (e.g., East, Central and West coast of Bangladesh) (Kauffman et al., 2020;Rahman et al., 2015). Higher carbon stock in the West coast and lower values in the Central and East coast of the Delta is linked to differences in the geomorphic settings and associated differences in biophysical conditions among the coastal TA B L E 6 Future contribution of ecosystem regional carbon stock (in Mg C) and sequestration rate (in Mg C year −1 ) by 2030 at the current plantation success rate for existing plantations and under three management scenarios for new plantations at the current rate, the success rate increased to 30% and 50% (Bayraktarov et al., 2016)  regions (Rovai et al., 2018;Sasmito et al., 2020;Uddin et al., 2022).
The higher carbon stocks per hectare, as well as the higher level of plantation success in the West coast of the Delta compared to the Central and East coast is likely linked to environmental conditions that support higher mangrove productivity associated with fresh water influx from the rivers, high levels of precipitation, lower salinity levels, and higher species richness in the West coast (Alongi, 2014;Hayes et al., 2017;Uddin et al., 2022).
In addition to variation in productivity, variation in mangrove plantation success and carbon stocks over the Delta are influenced by disturbance regimes (e.g., wind and waves and intense storms) and anthropogenic drivers (e.g., deforestation, grazing) (Ewers Lewis et al., 2020;Kauffman et al., 2020). For example, we observed that the Central and East coasts of the Bangladesh Delta had lower carbon stocks than the West coasts, which may be caused by frequent disturbances associated with intense storms .
Cyclone frequency is higher in the East and Central coast (e.g., Noakhali and Chittagong region of Bangladesh) compared to the Western coastal region (Jakobsen et al., 2006). Additionally, anthropogenic factors (e.g. deforestation) and deer/cattle grazing may also influence carbon stocks (Uddin et al., 2022). We suggest that further study of the influence of disturbances, both natural and human activities, on mangrove carbon stocks in the plantations would enhance our understanding of the variation in carbon stocks observed among the different coastal regions.  Table S4).

| Present contribution of plantations to CO 2 removal
Although Bangladesh is one of the countries that includes creation of mangrove plantations as part of their LULUCF mitigation actions (Herr & Landis, 2016), the magnitude of the reduction target by afforested mangroves was not specified in Bangladesh's first NDC report (Herr & Landis, 2016) The future management scenario of increasing plantation success rate at ~30%-50% might be an ambitious goal for BFD, when anthropogenic (e.g., deforestation) and natural (e.g., climate change) pressures negatively act on mangrove cover (Giri et al., 2007). To implement enhanced plantation establishment a community forestry management approach may be needed (Gevaña et al., 2019;Pulhin et al., 2017), which may reduce human disturbances on plantations (e.g., deforestation) and help to achieve the plantation success rates of 30%-50%. To achieve ambitious planting success rates, BFD could adopt some additional management innovations. For example, improved ways of identifying newly accreted coastal lands which might be suitable for mangrove planting through remote sensing, improving assessments of site suitability, species selection, and replanting where trees are damaged by storms or deforested. Deforestation of afforested mangroves could be prevented through implementing current laws (e.g., Forest Act, 1927), and policies (e.g., National Forest Policy, 2016), including those related to the United Nation's SDGs goals no. 13-15 (Chowdhury, 2018;General Assembly, 2015), all of which help meet the first objective and relevant policy statements (e.g., statement no. 2. 7, 9.12, 9.13)  In our study, values of carbon stocks may be underestimated because woody debris, seedlings and herbaceous vegetation were not included in the carbon pool assessments . Underestimation of carbon stocks may also arise from measuring soil depth to 1 m (100 cm) when mangroves may add SOC deeper in the soil profile (Kauffman et al., 2020). Additionally, we calculated carbon stocks and sequestration rates of the existing plantations based on the aerial extent of 28,000 ha (baseline in 2016) (Uddin et al., 2022). We calculated mean ecosystem carbon stocks (Mg C ha −1 ) in plantations across plots with ages from 5 to 42 years, which might be over-or underestimated depending on the age distribution of plantations. Future work could establish age distributions of plantations and thus help reduce uncertainty. Furthermore, our study did not assess human impacts on plantations (e.g., conversion of mangroves to agriculture, aquacultural practices and deforestation) that might influence the carbon accumulation in the Delta by influencing mangrove area and forest structure (Boone Kauffman et al., 2017;Pendleton et al., 2012;Sanders et al., 2016). Human uses of the coastal zone in the Bangladesh Delta are intense and it is possible that human activities influence the carbon stocks estimated in our study. We therefore suggest that future work assessing carbon stocks and rates of sequestration should assess the influence of human use of the mangrove plantations.

| CON CLUS ION
Plantation mangroves in Bangladesh contribute to carbon accumulation in biomass and soils at a similar rate as other plantation mangroves in deltas in Asia, but their per hectare accumulation is lower than global estimates of carbon accumulation from natural mangroves. The creation of mangrove plantations in Bangladesh for climate change mitigation will be most effective 20 years after establishment. Carbon accumulation in the plantation mangroves (5,417,609 Mg C or 19.9 MtCO 2 e) is ~22% of Bangladesh's GHG reduction target by 2030 (from all sectors) which could be enhanced in the future. Carbon stocks are potentially correlated with success rates of planted mangroves, therefore, increasing plantation success rates could potentially enhance additional GHG removal (2,098,093 Mg C or 7.7 MtCO 2 e) from the existing and new plantations by 2030, which would help to meet the international commitments of Bangladesh government, such as the NDC targets.