Variability in blue carbon stocks across mangrove forest types in Jamaica

Mangroves are among the most carbon-rich tropical forests, with much of their carbon stock located in underlying sediments. They play a vital role in climate mitigation and adaptation by offsetting carbon dioxide emissions linked to global warming. Despite their importance, blue carbon ecosystems remain poorly represented in climate policy across the Caribbean, with Jamaica and Belize being the only Caribbean countries to incorporate blue carbon stock values into their Nationally Determined Contributions (NDCs). Limited standardized information on mangrove carbon stocks across representative forest types in the region, further constrains the development of national baselines and their integration into climate policies. We conducted comprehensive mangrove carbon assessments across seven sites within four forest ecotypes: riverine, fringing, cay and lake to establish a national carbon baseline. The national mean total ecosystem carbon stock was determined to be 387.8 ± 18.6 Mg C ha -1. Recent island-wide mapping of mangrove forests determined Jamaica’s total mangrove extent to be 11,978 ha. Using these values, a total national carbon stock of 4.7 ± 0.2 Tg C was determined for Jamaican mangrove forests, with 3.5 Tg C stored as sediment stocks, 1.1 Tg C as above and below-ground tree biomass and 0.08 Tg C as standing dead tree biomass. Results indicate that forests within freshwater systems had the greatest carbon stored as living biomass (79.5 ± 15.7 Mg C ha -1), while forests on small mangrove cays, had the highest sediment carbon (344.6 ± 13.1 Mg C ha -1) despite influences from tidal flows. The random forest regression predictor identified sediment pH as important predictors across all carbon components, while forest type was important for standing dead and sediment carbon. Water salinity was a strong predictor of standing dead carbon and a weak and negative predictor for biomass carbon. This study provides valuable insight on the variability of mangrove carbon in different forest types potentially supporting sustainable management and conservation strategies necessary to inform policy. It represents the most extensive mangrove carbon assessment done in Jamaica with results informing the 2025 update of the island’s NDCs, a commitment under the Paris Agreement.

Coastal blue carbon ecosystems; mangroves, seagrasses, and saltmarshes, are vital for climate regulation, sequestering and storing three to five times more carbon per unit area than most forests (Donato et al., 2011; Pendleton et al., 2012; Choudhary et al., 2024; Kang et al., 2025). Mangroves, in particular, are among the most carbon-rich tropical forests, with much of their carbon stock located in underlying sediments (Donato et al., 2011; Kauffman et al., 2014; Choudhary et al., 2024). Globally total ecosystem carbon stock (TECS) ranges from 46 to 2205 Mg C ha −1 (Alongi, 2012; Bulmer et al., 2020). The long-term carbon sequestration potential of mangroves has led the Intergovernmental Panel on Climate Change (IPCC) to recognize them as a key nature-based solution for climate change mitigation.

Mangrove systems remain poorly quantified in the Caribbean (Munguía and Bhat, 2024) with studies providing mainly aboveground biomass carbon and a few studies giving total carbon stocks (Kauffman et al., 2014; Bolivar et al., 2018; Branoff and Martinuzzi, 2020; Perdomo Trujillo et al., 2021; Ávila-Acosta et al., 2024). Majority of the carbon stock studies in this region are primarily conducted in Latin American countries (e.g., Dominican Republic, Columbia and Mexico) with only one comprehensive large-scale assessment in a Caribbean Community (CARICOM) member state-Belize (Morrissette et al., 2023). Consequently, the insular Caribbean remains data deficient, constraining regional reporting to support credible blue carbon integration into Nationally Determined Contributions (NDCs) and carbon markets. For Jamaica, and other Caribbean islands, quantifying blue carbon offers a pathway to climate commitments, resilience, and sustained conservation of these valuable ecosystems. However, integrating mangroves into national and regional climate strategies requires robust, spatially explicit carbon estimates and credible monitoring, reporting and verification (MRV) systems.

The classification of mangrove forests based on physiognomy or physiography/geomorphology has been the subject of several studies (Lugo and Snedaker, 1974; Thom et al., 1984; Woodroffe, 1992; Ewel et al., 1998; Twilley and Rivera-Monroy, 2005; Twilley et al., 2019) with Worthington et al., 2020 being the most recent suggesting a new term ‘biophysical typology’. Mangroves were first classified into six physiognomy or ecotype categories: fringe, riverine, overwash, dwarf, basin and hammock (a variant of basin) by Lugo and Snedaker (1974), with Ewel et al. (1998) then reclassifying the forests into three main categories: fringe, basin and interior. Variability in carbon stocks across species, physiognomy, geomorphological settings, and disturbance histories remains a major knowledge gap, complicating their integration into national inventories and climate policies (Smith and Sarun, 2026). This variability necessitates site-specific Tier three assessments, which uses detailed field measurements and/or spatial modeling to provide nationally representative carbon stock estimates, rather than reliance on global averages (IPCC, 2014). Community composition, forest structure, geomorphological setting, hydrological regime and disturbance history can strongly influence the magnitude and distribution of carbon among living biomass (above and below ground) and sediments. Research consistently indicates that physiognomic forest type is an important determinant of carbon storage, with riverine forests typically containing the highest sediment carbon (Kauffman et al., 2014; Adame et al., 2015; Murdiyarso et al., 2015; Marchio et al., 2016; Sasmito et al., 2020; Herrera-Silveira et al., 2020; Choudhary et al., 2024; Dharmayasa et al., 2025).

Jamaica is an archipelagic state in the Caribbean with mangroves along much of its coastline. Though relatively modest in extent, Jamaica’s mangroves are ecologically and economically significant. The mangroves for the most part have been surveyed for areal extent, species composition, above-ground biomass and a range of forest parameters (McDonald et al., 2003; Webber et al., 2019). Jamaica has a range of mangrove forest types including riverine, fringe, overwash islands, scrub (possibly) basin forests as well as mangroves associated with a freshwater/brackish lake (St. George’s Lake in Discovery Bay, St. Ann). It is therefore important to assess the range of forest types so as to generate representative carbon values that can be applied to the mangrove areas across the island. To the best of our knowledge, only two site level studies focusing on estimating carbon stocks of mangrove forests have been conducted in Jamaica. The first conducted by Mandal et al. (2019) assessed the biomass and sediment carbon of three fringe mangrove sites; Portland Cottage, Clarendon; Salt Marsh, Trelawny; and Bogue, St. James. Another reported on the vegetation structure and carbon stock in an urban influenced forest on the southeast coast of Jamaica (Francis et al., 2026). However, the variability of carbon stocks based on forest ecotypes across Jamaica have not yet been studied or documented. Integrating blue carbon into national climate strategies through policy mechanisms such as NDCs requires a clear understanding of the spatial and temporal variability of carbon stocks across Jamaica’s diverse mangrove landscapes.

Our study presents the first comprehensive assessment of Jamaica’s mangrove carbon stocks, quantifying above-ground, below-ground, and sediment carbon across four forest types. We hypothesize that riverine and freshwater-associated forests will hold the greatest carbon, due to increased biomass due to the presence of freshwater and more sediment input (similar to Donato et al., 2011; Kauffman et al., 2020), while saline fringe forests are hypothesized to store the least. Carbon pools within overwash or “mangrove cay” forests, influenced by Jamaica’s micro-tidal environment, are expected to vary based on location and hydrological inputs, and are expected to yield values more comparable to island mangroves with organic-rich peat substrates. By addressing a major regional data gap, this research establishes a critical baseline for CARICOM island states seeking to incorporate mangrove conservation targets into climate frameworks, strengthen NDCs, and pursuing evidence-based decisions on land use, conservation, and climate action.

Approach

To assess sediment carbon stocks in mangrove ecosystems, we followed the IPCC Guidelines for Wetland Supplement (2014) and the Blue Carbon Initiative’s Manual for Coastal Blue Carbon Field Measurements (Howard et al., 2014). This approach allowed for reliable estimation of sediment carbon pools in diverse mangrove settings across various sites.

Site selection

Sites were selected based on forest type and to ensure broad coverage of dominant ecological conditions across Jamaica (Figure 1). The selected sites represented fringe forests (Rosey Hole, in the Port Royal mangroves on the south coast of Jamaica and Rock-Falmouth on Jamaica’s North coast), overwash/mangrove cay forests (Goat Island in Portland Bight and Refuge Cay in the Port Royal mangroves), riverine forests (Duhaney River in Hunt’s Bay and Martha Brae River in Falmouth), and an anchialine lake forest (St. George’s Lake) in Discovery Bay on Jamaica’s North coast.

Field sampling

Sampling spanned May to December 2024 and March 2025. A total of 1,681 trees were assessed in 49 plots across the seven sites. Additionally, 48 sediment cores and 522 sub-cores were collected for carbon storage analysis across these forest types.

Two transect lines (100 m apart) were extended in each forest up to 125 m where possible. Transects ran perpendicularly from sea to land, or from north to south on cays, with the centre of plot one placed 15 m from the water’s edge. From this point onward, the transect was subdivided into 25 m segments, where 7 m circular plots with nested 2 m subplots were further established from each centre stake (Kauffman and Donato, 2012; Morrissette et al., 2023). GPS coordinates were taken at the beginning and end of each transect line, at the center stake of each plot and at the location where each sediment core sample was taken. This was repeated at each location. The total number of plots per transect ranged from 2 to 6, due to variability in forest extent between sites.

Mangrove biomass–above-ground and standing dead biomass

Above-ground mangrove biomass parameters were collected in each 7 m radius plot. Measurements included tree density, species composition and status (live/dead tree decay status 1, 2 & 3; Howard et al., 2014). Height and diameter at breast height (DBH; 1.3 m above-ground) were collected using telescoping poles and diameter tapes, respectively. Basal diameter was also determined for all standing dead trees. Within each 2 m nested subplot, all mangrove trees were measured for DBH and tree height while from 2 to 7 m, only trees whose DBH was 5 cm and greater were measured.

Sediment carbon

At each centre stake, a 2-m rebar was used to determine the depth of sediment within each plot/area. Two sediment cores were taken (in sequence) to ensure a 1 m (IPCC, 2014) sediment core was obtained. The sequential cores were obtained by inserting the Russian peat borer and extracting the upper 50 cm of sediment. The instrument was then reinserted into the same borehole with a 1.5 m extension attached to retrieve the lower 50 cm of sediment or until core refusal (the point where no further downward vertical advancement is possible due to bedrock or impenetrable material (Cox et al., 2022). Subsamples were placed in labelled pre-weighed bags and stored in a cool container for transport to the laboratory. This procedure was repeated for each plot along the transects across the study area.

Physicochemical parameters

Physicochemical parameters such as depth of water on the forest floor (where present), sediment pH, sediment temperature and salinity were also recorded, usually near the centre stake of each plot. Salinity was determined using a refractometer while sediment pH and moisture were measured using a Kelway soil pH and moisture meter (accuracy: pH, ±0.2; moisture ±10%). Temperature was measured using a Bel-Art Durac B61310-3400 thermometer (±2 °C).

Laboratory processing

Wet weight, dry weight and loss on ignition (LOI)

In the laboratory, each sediment sample was weighed using an Ohaus Adventurer AX422/E precision balance for wet weight and placed in a Thermo Scientific Heratherm OMH180-S lab oven for drying at 60 °C for 72 h or until completely dry. Samples were then reweighed for dry weight and calculation of dry bulk density. Each sample was then either placed in a Fritsch Pulverisette two grinding mill or a mortar and pestle and manually pulverized. Homogenized samples were then sieved using a 1 mm mesh sieve. A 1 g subsample of each dried and homogenized sample was placed into labelled pre-weighed crucibles and combusted in a Thermo Scientific Lindberg/Blue M LGO 1200 Box Furnace at 450 °C for 6 h in order to obtain LOI. Samples were then allowed to cool and reweighed for determination of ash free dry weight and calculation of carbon values following methods as detailed in Howard et al. (2014).

Data manipulation

Mangrove tree biomass was categorized into three distinct pools or components: above-ground biomass (AGB), below-ground biomass (BGB), and standing dead biomass (SDB). Above-ground biomass estimates were converted to carbon values using the equation: 0.362DBH 1.930 (Laguncularia racemosa), 0.722DBH 1.731 (Rhizophora mangle) or 0.403*DBH 1.934 (Avicennia germinans) while below-ground biomass was estimated using the equation: Below-ground tree biomass (kgC) = 0.199 * (wood density (g/cm 3) 0.899) * (tree diameter at breast height (cm)) 2.22. In order to calculate the carbon content of each component (carbon pool), a conversion factor of 48% was used for above-ground; 50% for standing dead (Kauffman and Donato, 2012) and 39% for below-ground as suggested in Jaramillo et al. (2003).

The resultant values were then scaled to the standardized unit area of g C cm -2 then converted to megagrams of carbon per hectare (Mg C ha -1). In order to determine a national mangrove carbon stock value for Jamaica, mangrove carbon stock collected for each site was calculated and averaged to give a mean carbon stock. This average was then multiplied by the total mangrove forest area of 11,978 ha obtained through a contemporaneous study by The Nature Conservancy. The Nature Conservancy is a global environmental non-profit organization which works through science-based conservation and partnerships to address biodiversity and climate-related challenges (The Nature Conservancy, 2026).

Statistical testing was conducted on the summarized dataset (Supplemental Datasheet) and included the assessment of data normality and equality of variance using the Shapiro-Wilk and Levene’s tests respectively. For normally distributed data (log-height, log-dbh and log-standing-dead), Analysis of Variance (ANOVA) followed by Tukey’s Honest Significant Difference (HSD) test were used to assess differences among the four forest types. For non-normal data, the Kruskal–Wallis test was applied, followed by Dunn’s post hoc test for pairwise comparisons. Random forest method using a set seed of 123 was used to model a non-linear multiple regression based on the mean square error (MSE). Additionally, Spearman’s rank correlation was used to evaluate relationship strength between variables. All statistical analyses were conducted using SPSS 22 and R-4.3.3 statistical software (RStudio 2025.09.1 + 401: R Core Team, 2025) with a significance threshold set at p < 0.05.

Plot observations

Tree species, dominance, sediment core characteristics

Across the four forest types, clear differences were observed in sediment composition and vegetation dominance.

Mangrove cay forests occurred as compact mangrove islands rather than typical sandy systems, with sediment cores overwhelmingly dominated by organic material (approximately 98.1%, based on observations Supplementary File 2). Substrates on the mangrove cays were peat-like, dense and highly consolidated. Minor contributions were observed from organic matter with sand/silt/clay mixtures (1.3%) and core refusal (0.7%). The dominant tree species was Avicennia germinans (47.5%) followed by Rhizophora mangle (35.5%) and Laguncularia racemosa (17.0%).

Fringe forests displayed a heterogeneous sediment structure, comprised of 59.7% organic matter and organic matter intermixed with gravel (28.4%). Additional components included organic matter intermixed with sand/silt/clay (2.8%), gravel alone (4.0%) and discrete sand/silt/clay layers (0.6%). Air spaces and core refusal accounted for 2.8% and 1.7% respectively. Fringe forests were dominated by L. racemosa (58.7%), with additional presence of R. mangle (26.7%) and A. germinans (14.6%).

Riverine forests are strongly influenced by fluvial processes (Woodroffe et al., 2016; Sasmito et al., 2020) and were characterized by mixed sediment composition, with 79.6% consisting of organic matter combined with sand/silt/clay sediments. Discrete organic matter deposits within these forest sediments accounted for 15.2% while sand/silt/clay only layers, accounted for 4.6% of core compositions. There was minimal organic matter with gravel (0.76%). Vegetation data indicate that the dominant tree species in riverine forests were A. germinans (39.4%), followed by L. racemosa (30.7%) and R. mangle (29.5%).

Sediments at St. Georges Lake, the only lake forest on the island of Jamaica, and an anchialine system, exhibited high organic content (81.8% organic matter), with additional contributions from organic matter mixed with sand/silt/clay (7.6%), air spaces (6.1%), and core refusal (4.6%). No discrete sand/silt/clay or pebble layers were recorded. Here, the dominant tree species was R. mangle (82.3%), with small amounts of L. racemosa (17.7%).

Current national mangrove carbon stocks

A mean total ecosystem carbon stock (TECS) value of 387.8 ± 18.6 Mg C ha -1 was generated for the four forest types across seven sites. From this, a national value of 4.7 ± 0.2 Tg C was estimated for Jamaican mangrove forests. The sediment carbon at the 1 m depth accounted for the greatest carbon proportion with an estimated stock of 3.5 Tg C followed by above-ground living biomass with 0.7 Tg C, below-ground living biomass with 0.4 Tg C and standing dead biomass of 0.08 Tg C (Figure 2A; Supplementary Table S2). The mean contribution from each carbon component can be separated (Figure 2B) as follows: Above-ground tree carbon 58.1 ± 5.4 Mg C ha -1, Below-ground tree carbon 35.1 ± 3.7 Mg C ha -1, Standing dead carbon 5.5 ± 1.3 Mg C ha -1 and Sediment carbon 295.1 ± 14.5 Mg C ha -1.

Relationship between forest type and carbon content

Total carbon stock values ranged between 38–865 Mg C ha -1 in fringe, 234–489 Mg C ha -1 in riverine, 265–485 Mg C ha -1 in lake and 298–494 Mg C ha -1 in cay forests. Of the four forest types, the highest mean carbon (422.7 ± 15.8 Mg C ha -1) was found within cay forests (Figure 3A). This was followed by fringe (406 ± 45.9 Mg C ha -1), lake (373.3 ± 34.2 Mg C ha -1) and riverine (328.5 ± 26.3 Mg C ha -1) forests. Despite some variability observed in mean carbon between forest types, the differences were not statistically significant (p = 0.110).

Mean above and below ground plant carbon stocks were highest in lake forest (79.5 ± 15.7 Mg C ha -1 and 55.7 ± 10.0 Mg C ha -1; Supplementary Figure S1) and least in cay forests (40.5 ± 4.7 Mg C ha -1 and 24.5 ± 3.2 Mg C ha -1) respectively, while standing dead carbon stocks were highest at cay forests (13.1 ± 3.5 Mg C ha -1) and lowest at fringe forests (1.6 ± 0.4 Mg C ha -1). Sediment carbon stocks across these sites ranged from 222.2 ± 10.4 Mg C ha -1 in riverine to 344.6 ± 13.1 Mg C ha -1 in cay forests (Figure 3B). Statistically significant differences among forest types were found in below-ground tree carbon (adjusted p = 0.044), sediment carbon (adjusted p = 0.001), and standing dead carbon (log values; p = 0.013). Dunn’s post hoc test revealed no statistical significance between specific forest types for below-ground but statistically significant differences between riverine and cay in terms of sediment carbon, while Tukey HSD indicated differences between fringe and cay forests in terms of standing dead carbon.

Relationship between physicochemical parameters and carbon content

Physicochemical parameters

Water salinity was the only parameter to vary statistically between forest types (p < 0.001) with significant differences highlighted between riverine, cay and fringe forests. There were no significant differences noted between cay and fringe forests. Water salinity values ranged from 0 ppt to 40 ppt across sites with a mean of 1.7 ± 1.7 ppt in lake, 9.7 ± 0.9 ppt in riverine, 28.6 ± 3.1 ppt in cay and 30.8 ± 1.2 ppt in fringe (Table 1).

Physiochemical parameters (sediment pH and sediment temperature) were not statistically different between forest types. There was very little variability in sediment pH within and across forest types, which was relatively neutral except in the case of riverine forests which had records from 5.6–7.0 units compared to those in cay forests (6.6–7.0 units), fringe (6.5–7.0 units) and lake (6.8–7.0 units). Similarly, temperature ranges in the forests largely fell between 22 °C–30 °C with the largest temperature difference being observed in the fringe (23 °C–30 °C) and cay forests (22 °C–28 °C) compared to lake and riverine forests at (24 °C–26 °C) and (24 °C–28 °C) respectively.

Regression predictor of carbon content

The random forest regression predictor (Figure 4; Supplementary Table S3) showed sediment pH was positively important in all carbon components with sediment carbon having the highest mean squared error (MSE) value of 7.6%. Forest type was very important for standing dead (9.8%) and sediment carbon (11.4%) pools. Water salinity was only a strong predictor of standing dead carbon (8.8%) but weak and negative for above-ground (−3.3%) and total carbon (−2.0%).

All carbon stocks (Figure 5), except total plant carbon had a ρ (rho) value between 0.73 and 0.82, indicating a strong non-linear relationship between predicted and observed carbon stocks. The random forest model (R2) was able to explain 61% of the variability in above-ground carbon (Figure 5A), 63% in below-ground tree carbon (Figure 5B), 77% in standing dead carbon (Figure 5C), 58% in total plant carbon (Figure 5D), 73% in sediment carbon (Figure 5E) and 62% in total carbon stock (Figure 5F).

A weak correlation was shown to exist between forest type with above-ground carbon (0.12, Figure 6A) and standing dead carbon (0.32, Figure 6C) while a similar weak but significant correlation was shown to exist between forest type and below-ground carbon (0.32 Figure 6B). Forest type and sediment carbon (−0.37) as well as sediment temperature (−0.36) which showed a negative significant correlation (Figure 6D). A moderate negative correlation was observed between forest type and water salinity (−0.57) and a weak negative correlation between forest type and total ecosystem carbon (−0.18; Figure 6E). Moderate positive correlation also existed between sediment carbon with sediment temperature (0.36) and salinity (0.4) at the critical alpha p < 0.05. DBH (0.55) and height (0.76) showed moderate to high significant positive correlation respectively with forest type at p < 0.001.

Background

Mangroves as coastal blue carbon ecosystems are increasingly recognized for their potential to contribute to climate change mitigation, yet the variability of their carbon stocks across species, physiognomy and geomorphological settings remains a critical knowledge gap. Understanding this variability is essential for designing effective conservation and restoration strategies, and for ensuring the scientific integrity of carbon accounting frameworks. This is especially relevant in the Caribbean, where small island developing states like Jamaica, highly vulnerable to climate change, are actively exploring blue carbon as a pathway to meet climate commitments, enhancing coastal resilience while balancing sustainable development.

Carbon stock

National comparisons

National estimates of mangrove carbon storage are strongly influenced by the spatial extent of mangrove forests within a given country. In Jamaica, mangroves cover approximately 11,978 ha (McNulty et al., 2025) and were found to store a total ecosystem carbon stock of 4.7 Tg C. This estimate exceeds previous national values by approximately 1 Tg C as reported by Mandal et al. (2019). The difference in carbon stock estimates between the current study and the Mandal et al. (2019) study may be attributed to differences in number of study sites, the total mangrove area assessed and sediment core depth. Only three (3) study sites and total mangrove area of 9,715 ha were evaluated and reported in the 2019 study, compared to seven (7) study sites and 11,978 ha investigated in the current study, likely yielding higher carbon stock estimates. Furthermore, Mandal et al. (2019) analyzed only the upper 30 cm of sediment at each site, while this study analyzed up to 100 cm where possible, potentially capturing additional stored carbon, resulting in increased carbon stock estimates.

Regional and global comparisons

On average, the mangroves in Jamaica store 387.8 ± 19 Mg C ha -1 with sediment stocks accounting for 75% of the total carbon. The mean value is slightly higher than but comparable to the IPCC default mean of 386 Mg C ha -1. Our value is however less than the findings of 444.1 ± 21 Mg C ha -1 (57,854 ha) in Belize (Morrissette et al., 2023), 477.0 Mg C ha -1 (2,195 ha) in Seychelles (Wartman et al., 2025) as well as the mean Central, North America and Caribbean value of 730.1 Mg C ha -1 (Alongi, 2022) or the global mean value of 856 Mg C ha -1 (Kauffman et al., 2020). These findings along with the range of country-specific mangrove extent further support the call for each nation to complete in-country tier two or three estimates instead of relying on default values (tier 1), which may result in over- or under-estimation of carbon stocks.

When each carbon component or carbon pool is examined, the above-ground (58.1 Mg C ha -1), below-ground (35.1 Mg C ha -1) biomass and sediment carbon (295.1 Mg C ha -1) averages for Jamaica are higher than findings in Florida (Radabaugh et al., 2017) where 38.3 Mg C ha -1 was above-ground and 104.2 Mg C ha -1 in sediment. These averages, however, are lower than the Central America, North America and Caribbean regional value of 92.5 Mg C ha -1 (above-ground biomass), 41.4 Mg C ha -1 (below-ground biomass) and 577.4 Mg C ha -1 (sediment) (Alongi, 2022) but within the global range of 1.4–2052 Mg C ha -1 (Kauffman et al., 2020). It should be noted that the regional and global studies provide results for a variety of geomorphic settings, forest types and anthropogenic influences which would be different from those in Jamaica. Collectively, these studies demonstrate that mangrove carbon stocks cannot be reliably represented by a single global or regional value. Differences in forest type, geomorphology, hydrological connectivity, sediment characteristics, species composition, areal extent, climate and disturbance history can produce substantial variation in both TECS and relative contribution of individual carbon pools. Consequently, the transfer of carbon stock estimates from geographically distant mangrove systems may introduce substantial uncertainty into national stocks, greenhouse gas inventories and climate mitigation assessments. This reinforces the need for site-specific, spatially explicit assessments that capture local environmental and forest characteristics.

Relationships between physicochemical parameters, forest type and carbon pools

Our results indicate that sediment pH and salinity are the main drivers of carbon stocks, with additional modulation by geomorphological setting and sediment characteristics. While the direct correlation between total carbon, sediment stocks and sediment pH in mangroves are not strongly correlated, sediment pH is an important regulator of organic matter decomposition and preservation (Asante et al., 2024). The recorded sediment pH range (5.6–7.0) falls within the typical slightly acidic conditions found within mangrove systems (Hemati et al., 2014; Sofawi, 2017; Dewiyanti et al., 2021; Cinco-Castro et al., 2022; Cinco-Castro et al., 2023). Mean pH values were consistent across forest types (∼6.9), with slightly lower values in riverine systems (6.7), aligning with values reported by Eddiwan and Hendrizal, 2024. These conditions are conducive to microbial activity and organic matter turnover, facilitating carbon accumulation over time.

Vegetative carbon pool

The hypothesis that highly productive riverine or freshwater-influenced forests would have the greatest carbon stores was only partially supported when carbon pools were considered independently. Distinct patterns emerged among above-ground biomass, below-ground biomass and sediment carbon, highlighting the importance of evaluating mangrove carbon as a function of multiple pools. Above-ground biomass carbon (AGC) and below-ground biomass carbon (BGC) were highest in the freshwater influenced system at St. George’s Lake, consistent with expectations for low salinity, highly productive environments. Reduced salinity stress allows for increased photosynthetic productivity and biomass accumulation (Twilley et al., 2018). The higher carbon stock in the living biomass of St. George’s Lake forest may also be attributed to the presence of trees with large diameters (Dharmayasa et al., 2025) and the tallest trees (Adame et al., 2013) among all forest types. The dominance of Rhizophora mangle (82.3%) within the lake forest, alongside field observations of mature and structurally developed stands supports the interpretation that both tree AGC and BGC are enhanced by favourable growth conditions. Similarly, riverine forests, influenced by fluvial inputs also exhibited relatively high tree carbon.

Fringe forests displayed relatively high AGC (63.38 Mg C ha -1) but comparatively low BGC (32.51 Mg C ha -1). This may reflect their more dynamic coastal setting, where heterogeneous sediments and exposure to wave action may constrain root expansion relative to more stable environments (Wang et al., 2024). The dominance of L. racemosa in these areas may further indicate adaptation to such conditions, with carbon allocation favouring above ground structures. In contrast, mangrove cay forests exhibited lowest AGC and BGC despite supporting the highest sediment stocks. This suggests that biomass accumulation may be limited and is likely due to stressors including natural and anthropogenic influences such as solid waste buildup, wave exposure, hurricanes and observed highest standing dead trees (13.06 Mg C ha -1). The comparatively high standing dead carbon in mangrove cay forests could possibly indicate reduced forest vitality, which may constrain live biomass accumulation while contributing organic inputs to the sediment pool. These patterns suggest that biomass carbon (AGC and BGC) is maximized in freshwater-influenced systems where growth conditions are favourable, while more exposed or structurally constrained environments, such as mangrove cay forests, exhibited reduced living biomass despite potentially high total ecosystem carbon.

Sediment carbon pool

Sediment composition is suggested to be more relevant than the quantity of sediment being transported (Asante et al., 2024). In riverine settings, our findings showed that the predominance of mixed sediments (organic matter combined with sand/silt/clay) indicated substantial mineral input and sediment reworking. This may be due to an increase in the deposition of upstream, terrigenous sediments which are likely to displace autochthonous organic-rich material. Although riverine forests exhibited high AGB, this pattern did not extend to sediment organic carbon (SOC), with both riverine and lake forests showing comparatively low SOC values. This contrasts with our initial hypothesis, as well as the findings of other studies such as Kauffman et al. (2020) and highlights a possible decoupling between vegetative biomass productivity and long-term carbon burial. At St. George’s Lake, although SOC was high, the presence of air spaces in some plots, points to discontinuities in sediment structure that may facilitate localized oxidation. These characteristics likely reduced the efficiency of carbon stabilization within the sediment, despite high organic inputs from biomass. This interpretation is consistent with Adame et al. (2015), who proposed that freshwater influenced mangroves may accumulate sediment carbon more slowly, even where biomass is high.

In contrast, overwash (mangrove cay) forests exhibited the highest SOC, despite evidence of disturbance, the highest number of standing dead trees and a comparatively lower biomass carbon. This suggests that sediment conditions, rather than productivity, are the dominant driving factor for long term carbon storage within this forest type. The near-complete dominance of organic material and the presence of dense, peat-like substrates indicate limited mineral (sand/silt/clay) inputs and restricted oxygen penetration, both of which favour carbon preservation. These findings are supported by Twilley et al. (2018) who suggests that tide and wave influenced mangroves can support substantial sediment carbon stocks. Fringe forests displayed intermediate characteristics, with moderate SOC corresponding to their observed heterogeneous sediment composition. The presence of gravel, mixed substrates, and air spaces suggests increased permeability and oxygen diffusion. These conditions are likely to enhance decomposition relative to the more consolidated, organic-rich settings seen in mangrove cays as discussed above. Contrastingly, Asante et al. (2024) suggests that refractory organic carbon (ROC), which is supplied by woody mangrove materials (a key constituent of peat) may resist rapid decomposition in the presence of available oxygen. This organic carbon can therefore remain buried within even disturbed sediments.

Physicochemical parameters

Physicochemical parameters, particularly pH appear to play a secondary role in shaping carbon storage patterns. While the slightly acidic to neutral conditions observed are conducive to organic matter decomposition, they do not sufficiently explain the variability in carbon storage across forest types. Instead, the interaction between hydrology, sediment composition, and vegetation structure emerges as the primary control, determining whether carbon is retained in living biomass or preserved within sediments.

Overall, carbon storage in mangrove ecosystems is not uniformly maximized within a single forest type but is instead partitioned across different carbon pools depending on environmental conditions. Freshwater influenced systems maximize carbon in living biomass, whereas overwash settings maximize long term sediment carbon storage. This reinforces the need for a multi-pool approach when evaluating blue carbon potential, as reliance on total carbon storage alone may obscure important ecological and geomorphological controls.

Study implications and future steps

The study reinforces the importance of all mangrove carbon components, especially sediments, in storing large quantities of carbon. Jamaica has committed to incorporating natural ecosystems for the first time in its 2025 update to the Nationally Determined Contributions (NDCs) under the Paris Agreement. This study has provided the baseline for carbon stock for the island so that in the next 5 years and beyond, another inventory event should take place to monitor changes in the biomass and carbon stock. This will also allow for measurements of greenhouse gas accounting estimates from natural ecosystems to be included in the 2030 NDCs.

Twilley et al. (2018) suggests that countries should estimate carbon stock for each area along its coast instead of using mean values, since environmental setting is naturally variable. Future studies should therefore aim to focus on estimating the carbon stock in all mangrove locations around the island. With carbon stocks calculated and verified, the application of improved technologies such as remote sensing can be applied to facilitate the expansion of Jamaica’s carbon inventories. A critical ‘next step’ for Jamaica would be the determination of sediment accretion rates within each forest type using radiocarbon dating as well as stable isotope C and N analysis. This would allow us to determine accumulation rates and identify sources of carbon in each mangrove type.

Jamaica’s mangroves store 4.7 Tg of carbon with 75% being in the sediment carbon pool. It was hypothesized that riverine or freshwater-influenced forest would store more carbon however, mangrove cay/overwash forests were found to contain greatest accumulation of carbon within the sediment, with lake and riverine forests storing more carbon in the above-ground living biomass. These findings highlight substantial variability in carbon storage among forest types and demonstrate the importance of considering both ecosystem characteristics and individual carbon pools when assessing national mangrove carbon stocks.

A few limitations were encountered throughout this study, including gaps in spatial coverage and soil profile inconsistencies. Furthermore, the 1-year project timeframe prevented multiannual sampling and resulted in the exclusion of two proposed sampling sites. Despite these limitations, this study provided the first tier two baseline national carbon stock value for Jamaica, a new addition to the blue carbon data-limited Caribbean region. In the absence of a Jamaica-specific organic carbon conversion factors for mangrove soils, our dataset was derived from loss on ignition (LOI) calculations for this study’s carbon stocks. This may have resulted in either under or overestimation of the carbon stock estimates. Yet, the carbon variability assessed within the four forest types, as well as their carbon pools, underscores the importance of adopting a comprehensive approach to mangrove conservation in Jamaica. These efforts should not be limited to a specific forest type, or any specific geographical area as carbon pools vary in contribution based on type and location.

Jamaica’s microtidal regime may facilitate carbon accumulation through reduced flushing of the forest, but the research suggests that local sediment dynamics may override broader tidal constraints. The results further demonstrate that mangrove carbon storage is influenced by environmental and forest characteristics. Sediment pH emerged as an important predictor across carbon pools, while forest type and water salinity were particularly influential for sediment and standing dead carbon stocks. The dataset generated from this study provided baseline data for the island’s updated NDCs allowing coastal ecosystems (nature-based solutions) to be part of the report for the first time.