Changes in soil organic matter content and quality in Amazonian mangrove forests converted to shrimp farms

Mangroves are highly efficient carbon sinks, yet they are increasingly threatened by aquaculture expansion. Moreover, the long-term impacts of shrimp farming on soil organic matter (SOM) quality and carbon sequestration in Amazonian mangroves remain poorly understood, particularly regarding changes in SOM composition and their consequences for ecosystem resilience. This study compares soil organic matter (SOM) in pristine mangroves, mangroves converted to shrimp ponds, and mangroves adjacent to shrimp ponds in the Brazilian Amazon. We showed that pristine mangroves soils had higher soil organic carbon (SOC) ranging from 25 to 32 g.kg^-¹, higher C/N ratio (15.6–20.9), and stable δ¹³C (-27.5 to -27.0 ‰) and δ¹⁵N (3.6 to 3.3 ‰), with SOM rich in lignin, carbohydrates, and polyaromatic compounds. In contrast, mangroves converted to shrimp ponds had lower SOC (<10 g.kg^-¹), lower C/N (~12), and enriched δ¹³C and δ¹⁵N, indicating shrimp feed inputs and nitrogen enrichment. The SOM was also lipid-rich and thermally less stable. Soils in mangroves adjacent to shrimp ponds exhibited only minor alterations, suggesting high resilience; however, effluent-driven SOM mineralization and reduced thermal stability are latent risks. Overall, shrimp farming decreases SOM content and alters its quality, undermining long-term carbon storage in both converted and nearby mangrove areas. Our findings highlight the importance of regulating effluent discharge and promoting more sustainable aquaculture practices to mitigate SOM degradation and safeguard the carbon sequestration capacity of adjacent mangroves.

Introduction

Mangroves are known for their remarkable carbon storage capacity making them important contributors to global carbon budgets and climate regulation (1, 2). These ecosystems are known for their high soil organic carbon (SOC) contents, which constitutes more than 80% of their total ecosystem carbon stocks (1). Mangrove SOC stocks often exceed those of upland forests due to their high biomass productivity and complex soil biogeochemical processes (3a).

Despite their ecological importance, mangrove forests are being rapidly degraded and converted to alternative land uses, particularly for aquaculture. Shrimp farming has expanded considerably in tropical and subtropical regions since the 1980s, often at the expense of mangroves (4). This conversion alters hydrology, salinity, and nutrient dynamics, potentially reducing the carbon sequestration capacity of these ecosystems (5–7), thereby altering their role from being a carbon sink to becoming important carbon sources (8, 9).

Previous studies have documented substantial declines in SOC stocks following mangrove conversion to aquaculture and agricultural uses (5, 9). However, the long-term impacts of shrimp farming on soil organic matter (SOM) composition, quality, and persistence in Amazonian mangroves remain poorly understood. In particular, little is known about how the shift from plant- to animal-derived organic inputs alters SOM chemical signatures, and the resilience of mangrove soils to ongoing disturbance.

Amazon mangroves are among the largest and most intact coastal forests globally. Consequently, the expansion of human activities in pristine areas pose a serious threat to their role in climate change mitigation and local ecosystem sustainability (10). Addressing this gap is essential for understanding the vulnerability of Amazonian mangroves to the pressures of aquaculture. By linking SOC concentration, SOM composition, and morphology across pristine mangroves and shrimp ponds, we can provide new insights into how aquaculture reshapes belowground carbon dynamics and threatens the capacity of mangroves to function as long-term carbon sinks.

The objectives of this study are therefore to: (i) assess differences in SOC concentration and SOM composition between pristine mangroves and shrimp ponds; (ii) evaluate changes in particulate organic matter (POM) morphology and potential shifts from plant- to animal-derived inputs; and (iii) discuss the implications of these changes for SOC persistence and ecosystem resilience in Amazonian mangroves. To achieve this, we utilized an integrated approach combining isotopic analysis, pyrolysis gas chromatography-mass spectrometry (Py-GC/MS), Rock-Eval^® thermal analysis (RE), and scanning electron microscopy (SEM). This multidisciplinary methodology allowed us to comprehensively characterize SOM transformations and their impact on carbon dynamics in these ecosystems.

Material and methods

Study area, soil sampling and preparation

The study area is located in the city of Curuçá, east of the mouth of the Amazon river, in the state of Pará, Northern Brazil (Figure 1). Samples were collected from pristine mangroves (ManP; 0°44’54”S, 47°56’44”W), former mangrove forests converted to shrimp ponds (Shrimp; 0°41’11”S, 47°51’26”W), and mangroves adjacent to shrimp ponds (ManS; 0°41’11”S, 47°51’26”W), along the Furo do Maripanema channel in the Curuçá River estuary (Figure 1). Shrimp pond operations in the study area started in the late 1980s; therefore, at the time of sampling (January 2019) mangroves had been converted for nearly 30 years. Management practices in the ponds consist of three harvests per year. After each harvest, the ponds are drained and remain empty for approximately two months, meaning that over the course of a year ponds are filled for about half of the time and left fallow for the other half.

Figure 1

Figure 1. Location of the study areas in the Curuçá River estuary, northern Brazil (a). The study includes three areas: pristine mangroves (ManP) (b), mangroves adjacent to shrimp ponds (ManS), and shrimp ponds (Shrimp) (c).

A soil core adapted for waterlogged conditions was used. Sampling in the shrimp ponds was conducted while the ponds were empty. Two cores were collected from ManP, while three cores were collected from ManS and Shrimp, with samples taken at depths of 0–10 cm, 10–20 cm, 30–40 cm, 40–50 cm, and 80–100 cm. All samples were placed in hermetic plastic bags and transported to the laboratory until stored at -20 °C.

Prior to analysis, samples were rinsed twice with deionized water to remove the excess salt, which can interfere with analyses, particularly Rock-Eval pyrolysis (RE) (11–13). An aliquot of the rinsed samples was freeze-dried and ground for further analysis described below.

SOC, total nitrogen and stable isotopes (δ¹³C and δ¹⁵N) analyses

Soil samples were treated with 1 M HCl to remove carbonates, rinsed with deionized water to remove any remaining acid, and then dried at 60 °C. The acid-washed samples were ground and sieved to less than 0.075 mm (200 mesh). The samples were analyzed using a FlashEA 1112 autoanalyzer (Thermo Finnigan) coupled with a Deltaplus isotope ratio mass spectrometer (ThermoFinnigan). The isotopic composition was expressed relative to international standards (Vienna Pee Dee Belemnite for δ¹³C and atmospheric nitrogen for δ¹⁵N) and obtained as described in Equations 1 and 2. The precision of isotope measurements was monitored by analyzing internal laboratory standards calibrated against international reference materials, with typical precision better than ±0.2‰ for δ¹³C and ±0.3‰ for δ¹⁵N.

$\begin{array}{ll}{\delta }^{13}C=\left(\frac{{(\frac{{}^{13}C}{{}^{12}C})}_{sample}}{{(\frac{{}^{13}C}{{}^{12}C})}_{VPBD}}-1\right)\times 1000& (1)\end{array}$

$\begin{array}{ll}{\delta }^{15}N=\left(\frac{{(\frac{{}^{15}N}{{}^{14}N})}_{sample}}{{(\frac{{}^{15}N}{{}^{14}N})}_{athmosphere}}-1\right)\times 1000& (2)\end{array}$

Rock-Eval^® thermal analysis

Soil samples were analyzed on a RE6 Turbo analyzer (Vinci Technologies). We used the 0-10, 40–50 and 80–100 cm depths of each site. Approximately 40 mg of finely ground samples were weighed into stainless steel pods. The first step involved a pyrolysis phase under a N₂ atmosphere. Samples were heated to 200 °C and held isothermally for 3 minutes, then further heated to 650 °C at a rate of 30 °C per minute. During this phase, hydrocarbon compound emissions were measured with a flame ionization detector (FID), while CO and CO₂ emissions were measured with an infrared detector. Following pyrolysis, samples underwent an oxidation phase in a CO₂-free air atmosphere. They were held at 300 °C for 1 minute, then heated to 850 °C at a rate of 20 °C per minute and finally held at 850 °C for 5 minutes. Emissions of CO and CO₂ during this phase were measured with an infrared detector. All emissions were measured every second during both the pyrolysis and oxidation phases. Each RE measurement provided five thermograms: hydrocarbons (HC_PYR), CO (CO_PYR), and CO₂ (CO2_PYR) from the pyrolysis phase, and CO (CO_OXI) and CO₂ (CO2_OXI) from the oxidation phase. A series of parameters were automatically calculated from the thermograms. From those, we particularly used: Pyrolyzed Organic Carbon (PC), Residual Organic Carbon (RC), both expressed as % of OC, Hydrogen Index (HI), expressed as mg HC g^-¹ OC, and Oxygen Index (OIRE6), expressed as mg O₂ g^-¹ OC. Additional parameters obtained with RE are shown in Supplementary Table S1. For detailed information on these parameters, see (14).

Pyrolysis gas chromatography/mass spectrometry

The molecular composition of SOM was investigated through Py-GC/MS. The procedure involved loading the samples in quartz tubes and pyrolyzing to 550 °C for 0.15 seconds, with a 30-second hold, in a CDS Pyroprobe 5000 Series. This temperature was selected because, according to the CH_PYR from RE analysis (Supplementary Figure S1), almost no additional organic matter was released above 550 °C. This approach minimizes extensive thermal decomposition, which could otherwise lead to secondary reactions such as cracking and condensation of organic compounds, thereby altering the chemical composition of the pyrolyzates (15).

The pyrolysis products were then transferred to an injector (Hewlett Packard HP-6890 gas chromatograph) set to 280 °C and operated in splitless mode. Separation of the pyrolysis products was achieved using a 60 m fused silica capillary Sol Gel Wax column (SGE, 0.32 mm i.d., film thickness 0.5 μm) with helium as the carrier gas (1 ml/min). The GC oven temperature was programmed to increase from 40 to 280 °C at a rate of 8 °C/min, with a final hold at 280 °C for 5 minutes. Following GC separation, the pyrolysis products were detected using an HP-5973 electron ionization mass spectrometer in scan mode (scan range m/z: 50–700, 1.2 scan/s, 70 eV).

The raw pyrograms were smoothed using a Savitzky–Golay filter (second-order polynomial over 21 points), and baseline corrections were performed using Statistics-Sensitive Non-Linear Iterative Peak-Clipping (SNIP). These preprocessing steps were conducted with OpenChrom^® software (Community Edition 0.8.0 Dempster). The smoothed and baseline-corrected pyrograms were then normalized using the Standard Normal Variate (SNV) process, where each pyrogram was centered and scaled by dividing by its standard deviation using Origin 2017 software (OriginLab Corporation, Northampton, Massachusetts, USA). The normalized pyrograms were aligned using the FFT/peak matching combined method with SpecAlign software (SpecAlign ver. 2.4.1, University of Oxford). Alignment is essential due to random fluctuations in mobile-phase pressure, oven temperature, flow rate, and stationary phase degradation, which can cause small shifts in retention times for the same compound across different analyses. The pyrograms then underwent statistical analysis as described below.

Scanning electron microscopy

Another soil sample aliquot underwent physical fractionation to isolate the particulate organic matter (POM), which was then analyzed by scanning electron microscopy. Briefly, 10 g of equivalent dry mass soil sample were mixed with 80 mL sodium polytungstate solution adjusted to a density of 1.60 g cm-3 and dispersed by ultrasonication at an energy of 400 J mL-1 (16). The mixture was then centrifuged at 10,000 g for 10 minutes to separate the POM. The supernatant containing the floating material (i.e., POM) was vacuum filtered over a Whatman GF/D glass filter (2.7 µm pore size) and rinsed with deionized water to eliminate any residual salt until the electrical conductivity was below 5 μS cm -1. The obtained POM was then freeze-dried.

The morphology of POM was investigated using scanning electron microscopy (SEM). POM particles were affixed to a tape, coated with a thin layer of gold to enhance electrical conductivity, and then analyzed using a field emission gun scanning electron microscope (FEG-SEM, Ultra 55 Zeiss).

Statistical analyses

We used permutational multivariate analysis of variance (PERMANOVA) to assess the differences in variables responses (SOC, C/N ratio, δ¹³C, and δ¹⁵N), across sampling sites (ManP, ManS and Shrimp; see Supplementary Table S2). PERMANOVA is a non-parametric method that does not rely on the assumption of normality and is suitable for ecological data, which often do not meet the assumptions of traditional parametric tests and is known to be robust to unequal sample sizes (17). For each variable, we performed a PERMANOVA based on a Euclidean distance matrix, treating groups as fixed factors and depth as a nested effect within each group. The analysis involved 999 permutations. We conducted pairwise comparisons between all pairs of groups, adjusting the p-values using the Bonferroni correction method (Supplementary Table S3). The analysis was performed using the adonis2 function from the vegan package in R (version 4.0.5). We further verified the robustness of PERMANOVA by testing the homogeneity of multivariate dispersions (PERMDISP, functions betadisper() and permutest() in vegan), since differences in within-group variability can bias the interpretation of multivariate tests. At the Group level (ManP, ManS, Shrimp), dispersion was homogeneous for δ¹⁵N, TOC and C/N (p > 0.05), while some heterogeneity was detected for δ¹³C (p < 0.05). To refine this, we also explored PERMDISP within each depth stratum, given the nested sampling. Results indicated that heterogeneity in dispersion was limited to specific depths for δ¹³C (e.g., 10–20 cm, p = 0.001; 30–40 cm, p = 0.049) but was not systematic across the profile. The dispersion of all other variables remained homogeneous across depths.

An exploratory approach was applied to the pyrolysis-GC/MS chromatogram data based on previous studies (18–20). Non-informative variables (i.e., almost constant among the samples) from the treated pyrograms were removed using a interquartile range criteria, eliminating 40% of the dataset (i.e., 2761 x,y-pairs). One-way ANOVA (p-value < 0.01) was then performed on the remaining peaks (Supplementary Figure S2). Data processing and analyses were conducted using MetaboAnalyst (21); see https://www.metaboanalyst.ca/). Significant retention times identified by ANOVA were used to reconstruct peak features (see the detailed zoom-in of the cyan box in Supplementary Figure S3), highlighting distinguished molecular features across sampling sites (ManP, ManS and Shrimp). Then, the original pyrogram was revisited to isolate peaks identified by ANOVA, and their respective mass to charge ratio (m/z) spectra were collected.

The final identification of the main organic compounds was achieved by comparing the collected m/z spectra with published data and the digital NIST 11 (Maryland, USA) libraries and published Py-GC/MS literature (22–26). This processing workflow allows the identification of major organic compounds (i.e., molecular fingerprints) and pprovidesa semi-quantitative assessment of relative abundances of similar molecules across sampling sites (ManP, ManS and Shrimp).

PCA was conducted using RE (PC, RC, HI, and OIRE6) and elemental analysis (SOC and C/N ratio) data to identify key variables that explain the variance in the dataset and the differences among groups. Missing data (i.e., HI values for four Shrimp samples as SOC was too low for accurate HI calculations) were imputed using the missMDA package, which uses the Regularized Iterative PCA (RIPCA) method (27). Then, the PCA was performed using the FactoMineR package in R (version 4.0.5). Group centroids (mean scores per treatment) and 95% confidence ellipses were added to visualize between-group differences. To validate the robustness of the ordination, we also conducted a non-metric multidimensional scaling (NMDS) using Euclidean distances (metaMDS, vegan package). The NMDS stress value was 0.06, indicating an excellent fit. The NMDS plot is provided as Supplementary Figure S4. While PCA emphasizes variance structure, NMDS preserves pairwise distances; the agreement between both approaches reinforces the robustness of the results.

Results

SOC, C/N ratio, and stable isotopes

Pristine mangroves (ManP) exhibited SOC concentrations ranging from 25 ± 5 to 32 ± 7 g.kg^-1, with a slight increase in SOC content observed with depth (Figure 2a). The C/N ratio in ManP soils varied between 15.6 ± 0.1 and 20.9 ± 2.8 across depths (Figure 2b). Both δ¹³C and δ¹⁵N remained stable with depth, with δ¹³C ranging from -27.5 to -27.0 ‰ and δ¹⁵N from 3.6 to 3.3 ‰. Mangroves adjacent to shrimp ponds (ManS) showed greater variability in these parameters compared to ManP, but no significant differences were detected between ManP and ManS, except for δ¹³C (p-value = 0.003; Supplementary Table S3), which was slightly higher in ManS, ranging from -27.1 to -26.6 ‰ (Figure 2c).

Figure 2

Figure 2. SOC (a), C/N ratio (b), δ¹³C (c), δ¹⁵N (d) across different soil depths and the relationship between δ¹³C and δ¹⁵N for all samples (e) in pristine mangroves (ManP), mangroves adjacent to shrimp ponds (ManS), and shrimp ponds (Shrimp). Whiskers represent the standard deviation. ManP (n =2), ManS (n=3) and Shrimp (n=3).

In shrimp ponds, SOC was consistently below 10 g.kg^-1 across all depths, and the C/N ratio decreased to approximately 12, despite being highly variable (Figure 2b). Soils converted to shrimp ponds were enriched in both δ¹³C and δ¹⁵N, with δ¹³C ranging from -24.0 ± 3.4 ‰ to -25.5 ± 3.8 ‰ and δ¹⁵N from 4.2 ± 0.7 to 6.2 ± 1.1 ‰ (Figures 2c–e). Significant differences (p-value < 0.05) were observed between shrimp ponds and the other areas for all variables, except for δ¹³C in comparison to ManP (p-value = 0.08; Supplementary Table S3). PERMANOVA results indicated that depth did not significantly influence the observed differences between the areas studied (Supplementary Table S3).

SOM chemistry and thermal stability

The Py-GC/MS analysis showed contrasting SOM composition between shrimp ponds and mangrove sites, with no significant differences between ManS and ManP (Table 1). Mangrove soils were more enriched in pyrolysis products typical of carbohydrates, lignin, and other polyaromatic compounds (Table 1). In contrast, soil converted to shrimp ponds had a greater contribution of pyrolysis products derived from nitrogen-containing, sulfur-containing compound and especially aliphatic compounds (i.e., alkanes, alkenes and lipid) (Table 1).

Table 1

Table 1. List of the main pyrolysis products and possible biochemical sources.

The RE analysis revealed contrasting SOC stoichiometry and resistance to thermal decomposition in shrimp ponds in relation to ManP and ManS. The shrimp pond site presented higher HI (232 ± 47) and lower OIRE6 (126 ± 1) in the 0–10 cm depth (Supplementary Table S1). The shrimp ponds also exhibited a greater proportion of pyrolizable carbon (PC) in relation to residual carbon (RC), on average twice as high as in ManP and ManS (Supplementary Table S1), indicating more thermally labile SOM. The PCA (Figure 3) further underscored the differences between shrimp ponds and mangrove soils, particularly with respect to the strong correlation between HI and PC associated with the shrimp pond site. Conversely, higher SOC, C/N ratio, RC, and OIRE6 were closely associated with soils from both mangrove sites, highlighting the impact of land use change on SOM chemistry and thermal stability.

Figure 3

Figure 3. Principal Component Analysis (PCA) of soil samples from pristine mangroves (ManP), mangroves adjacent to shrimp ponds (ManS), and shrimp ponds (Shrimp) including SOC, C/N ratio and RE data (HI, OIRE6, PC and RC). For each site use, three depths (0–10, 40–50, and 80–100 cm) were included, resulting in 24 samples in total. Larger symbols represent the average position of all data points within each group (centroids), and the ellipses represent the 95% confidence intervals around the centroids.

POM morphology

The SEM images revealed differences in POM morphology when comparing ManP and Shrimp pond soils (Figure 4). In ManP, the POM consisted of highly fragmented, irregularly shaped particles in various stages of decomposition, found throughout all depths (Figures 4a–c). The presence of woody and honeycomb-like structures, indicative of lignin-rich material, is particularly notable. In soils converted to shrimp ponds a more advanced degree of decomposition is evident with few POM fragments displaying the woody and honeycomb-like structures reminiscent of mangrove-derived organic matter (Figure 4d). Additionally, structures resembling shrimp cuticles also appear (Figures 4e, f), possibly chitin-rich tissues characterized by smooth and fibrous surfaces. However, this identification is based solely on visual evidence from SEM images, without complementary chemical analyses (e.g., chitin biomarkers), and should therefore be regarded as speculative.

Figure 4

Figure 4. Scanning Electron Microscopy (SEM) images of the particulate organic matter (POM) fraction in pristine mangroves (ManP) at various depths: 0–10 cm (a), 40–50 cm (b), and 80–100 cm (c). Images referring to shrimp pond (Shrimp) at the 0–10 cm depth show plant material (d) and structures resembling shrimp cuticles (e, f).

Discussion

Impact of mangrove conversion to shrimp ponds on SOM

Our work reveals a significant decrease in soil organic matter content and composition following the conversion of mangrove forests to shrimp farms. These alterations are likely to have cascading effects on adjacent mangrove ecosystems near active aquaculture operations. Few variations in SOC concentrations, δ¹³C and δ¹⁵N, and C/N ratios in the ManP and ManS soils indicate a steady input of organic matter with depth, which is attributed to root decay. This root-based C enrichment is supported by POM morphology (Figures 4a–c) that is possible visualize root structures and by the lignin-enriched SOM composition highlighted by Py-GC/MS analysis (Table 1). In fact, fine roots are the primary source of organic matter in these soils, potentially contributing to over 90% of the organic inputs (28). The preservation of lignin-rich fragments, even at greater depths (Figures 4a–c), underscores the significant role of constant fine root inputs to the SOM in natural mangrove forests. In addition to lignin, the Py-GC/MS results from mangrove soils reveal a diverse range of pyrolysis products, with a significant contribution from carbohydrate-derived and to lignin compounds derivatives (Table 1). These results align with previous work, where lignin and carbohydrates were identified as major components of SOM in mangrove soils, accounting for up to 40% and 30%, respectively, of total quantified peak area (29).

In contrast, the shrimp pond site were characterized by consistently low SOC concentrations, lower C/N ratios, and elevated δ¹³C and δ¹⁵N values. The decline in SOC is primarily driven by significant alterations to the natural inundation patterns of these areas, because while shrimp ponds may experience periods of anoxia when filled with water (30), they are frequently drained during the farming cycle (31). This redox oscillation creates oxidizing conditions that are more intense than those naturally occurring in mangrove forests (7) accelerating oxidation and consequently the decomposition of SOM, ending up for the lower SOC concentrations observed in shrimp ponds.

Additionally, there were notable changes in the remaining SOM. Although we did not have access to the specific composition of the shrimp feed, the enrichment in δ¹³C values is likely related to the incorporation of carbon from shrimp feed, which is often derived from marine fishmeal and C4 plants (32) and typically has higher δ¹³C values compared to the plant-derived organic matter in pristine mangroves (33, 34), which primarily originates from C3 plants (35). Regarding δ¹⁵N, shrimp feed frequently contains animal-derived proteins, which are naturally enriched in δ¹⁵N compared to the plant-derived nitrogen sources typical of mangrove ecosystems (32). The metabolic processes of shrimp, along with the decomposition of their waste, contribute to higher δ¹⁵N values observed (36). Combined with the lower SOC concentrations and C/N ratios (Figures 2a, b), these findings indicate a decline in plant-derived organic matter and an increase in animal-derived and microbial-derived nitrogen-rich inputs, as highlighted by pyrolysis-GC/MS data (Table 1). Moreover, there was visual evidence of animal-derived SOM as the observed structures were consistent with chitin-rich shrimp cuticles previously described in the literature (e.g., 37) supporting our interpretation that shrimp-derived residues contribute to SOM in converted ponds.

Interestingly, OIRE6 values were lower in shrimp pond soils than in mangrove soils, contrary to the expectation that SOM would be more oxidized due to the prolonged exposure to the atmosphere from pond drainage. This discrepancy arises from the distinct organic matter inputs in these areas rather than the sharp change in geochemical conditions upon conversion. Shrimp pond soils were enriched in branched and n-alkane/alkene aliphatic compounds commonly found in shrimp tissues (38). These molecules enrichment in shrimp pond sites corresponds with the increased HI and PC values from Rock-Eval analysis (Table 1, Figure 3) as fatty acids have high H/C ratios and are easily degraded under pyrolysis conditions. These findings suggest that, in addition to the loss of mangrove-derived organic matter and the input of limited new organic matter the SOM in shrimp ponds is thermally more labile than in mangroves.

Microbial communities in shrimp ponds are likely to specialize in decomposing the few available organic compounds due to limited diversity of organic matter input in this sites (39). Under the aerobic conditions typical of shrimp ponds, the decomposition of lipids likely promotes faster SOM turnover (40), further diminishing the capacity for long-term carbon storage. This rapid decomposition, combined with the low structural complexity of lipid-rich organic matter, likely reduces the potential for stabilization in soils under shrimp production.

Implications for carbon cycling and mangrove resilience

The slightly higher δ¹³C values observed in ManS may reflect subtle shifts in organic matter sources due to proximity to shrimp ponds. While previous studies have reported notable changes in SOC concentrations and chemistry in mangroves adjacent to shrimp ponds (6, 41, 42), the effects observed in our study were less pronounced. This could be attributed to high initial resilience of these mangrove forest and the relatively recent history of shrimp farming in this region, as well as in the Amazon (10). Over time, however, it is likely that not only the soils from former mangroves converted to shrimp ponds, but also the adjacent mangroves, will experience shifts in SOM composition due to shrimp farming.

Shrimp pond effluents rich in lipids and nitrogen compounds are expected to disrupt mangrove soil processes related to SOM cycling and possibly increase emissions of CO₂ and non-CO₂ (i.e., CH₄ and N₂O) greenhouse gases (43, 44). Lipids, which are found to be more resistant to microbial decomposition than lignin in forest soils (45), could potentially influence the microbial community structure, favoring a more rapid consumption of lignin in mangrove soils. In fact, it has been shown that additions of palmitic acid to soil can increase SOC mineralization rates due to a positive priming effect, with very low incorporation into microbial biomass (46). This could lead to a decrease in SOC stocks in these mangrove forests adjacent to ponds over time. It is important to note, however, that GHG fluxes were not directly measured in this study. Thus, our statements regarding GHG emissions are based on indirect evidence from SOM concentration and composition and should be interpreted as indicative only.

Excessive N inputs from shrimp pond effluents can significantly impact biogeochemical processes associated with SOC cycling. Thus, mangrove soils, which are naturally low in N, may experience stimulated microbial activity and faster decomposition of SOM as N levels rise (47, 48). Moreover, increased N from shrimp farming effluents can enhance nitrate reduction rates over sulfate reduction (5). Since the nitrate reduction pathway yields more energy per oxidized carbon molecule than sulfate reduction, higher rates of carbon mineralization are expected (49).

In addition to the well-known eutrophication risks, increasing nitrogen inputs can induce mangrove mortality (50). Although nitrogen enrichment may initially enhance mangrove growth, particularly in aboveground parts, it also causes mangroves to invest less in their root systems (51). As previously noted, since most organic matter inputs in mangroves are derived from roots (2, 28), this shift could result in significant decline in organic matter inputs and, consequently, SOC stocks. Moreover, this reduction in root biomass severely diminishes the resilience of mangroves to environmental stressors, such as increased salinity and reduced rainfall, which are anticipated consequences of climate change (52–54).

Conclusions

The conversion of mangroves to shrimp ponds significantly disrupts SOM dynamics. Mangrove soils converted to shrimp ponds exhibit substantial changes in SOM quality, indicating a shift from predominantly plant-derived to animal-derived organic inputs. The introduction of more labile organic matter further reduces the soil’s capacity for long-term C sequestration. Although the immediate impacts on adjacent mangrove soils were less pronounced, our data suggests that the discharge of shrimp farm effluents may, over time, significantly compromise the SOC sequestration capacity and weaken the overall resilience of neighboring mangrove ecosystems.

These findings underscore the broader environmental consequences of mangrove conversion for aquaculture. Mangroves are critical global carbon sinks, and their degradation not only releases stored carbon but also reduces their ability to sequester future carbon, exacerbating climate change. The loss of mangrove resilience further threatens coastal protection, biodiversity, and the livelihoods of communities dependent on these ecosystems. Therefore, sustainable management practices and policies are urgently needed to balance aquaculture development with the conservation of mangrove ecosystems.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://zenodo.org/records/16420149.

Author contributions

FR: Conceptualization, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft. TF: Investigation, Project administration, Resources, Supervision, Visualization, Writing – review & editing. MS: Formal Analysis, Investigation, Methodology, Visualization, Writing – review & editing. CR: Investigation, Methodology, Supervision, Validation, Visualization, Writing – review & editing. MD: Investigation, Methodology, Validation, Visualization, Writing – review & editing. FB: Investigation, Methodology, Visualization, Writing – review & editing. XO: Investigation, Methodology, Visualization, Writing – review & editing. AB: Conceptualization, Funding acquisition, Investigation, Project administration, Visualization, Writing – review & editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. We gratefully acknowledge the support of the RCGI—Research Center for Greenhouse Gas Innovation (23.1.8493.1.9), hosted by the University of São Paulo (USP) and sponsored by FAPESP—São Paulo Research Foundation (20/15230-5), and PETRONAS Petróleo Brasil Ltda. The authors acknowledge the strategic importance of the support given by ANP (Brazil’s National Oil, Natural Gas and Biofuels Agency) through the R&DI levy regulation (ANP–Project #23.702-4; BlueShore). We appreciate the support from the Center for Carbon Research in Tropical Agriculture (CCARBON) at the University of São Paulo, sponsored by the São Paulo Research Foundation (FAPESP) under grant 21/10573-4. This study was financed, in part, by the São Paulo Research Foundation (FAPESP), Brasil. Process Number #2023/06841-9 to F.R. Also, grant 305013/2022–0, National Council for Scientific and Technology Development (CNPq) to T.O.F. Additionally, A.F.B. was supported by the National Geographic Society and Rolex Perpetual Planet Amazon Expedition (PFA-21-PP031), C.R. was supported by the Fair Carbon project TROPECOS (ANR-22-PEXF-012) and X.L.O. was supported by the Consellería de Educación, Universidade e Formación Profesional-Xunta de Galicia (Axudas á consolidación e estruturación de unidades de investigación competitivas do SUG do Plan Galego IDT, Ambiosol Group ref. ED431C; 2022/40.

Acknowledgments

We gratefully acknowledge the support of the RCGI—Research Center for Greenhouse Gas Innovation (23.1.8493.1.9), hosted by the University of São Paulo (USP) and sponsored by FAPESP—São Paulo Research Foundation (20/15230-5), and PETRONAS Petróleo Brasil Ltda. The authors acknowledge the strategic importance of the support given by ANP (Brazil’s National Oil, Natural Gas and Biofuels Agency) through the R&DI levy regulation (ANP–Project #23.702-4; BlueShore). We appreciate the support from the Center for Carbon Research in Tropical Agriculture (CCARBON) at the University of São Paulo, sponsored by the São Paulo Research Foundation (FAPESP) under grant 21/10573-4. This study was financed, in part, by the São Paulo Research Foundation (FAPESP), Brasil. Process Number #2023/06841–9 to F.R. Also, grant 305013/2022–0, National Council for Scientific and Technology Development (CNPq) to T.O.F. Additionally, A.F.B. was supported by the National Geographic Society and Rolex Perpetual Planet Amazon Expedition (PFA-21-PP031), C.R. was supported by the Fair Carbon project TROPECOS (ANR-22-PEXF-012) and X.L.O. was supported by the Consellería de Educación, Universidade e Formación Profesional-Xunta de Galicia (Axudas á consolidación e estruturación de unidades de investigación competitivas do SUG do Plan Galego IDT, Ambiosol Group ref. ED431C; 2022/40.

Conflict of interest

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

The author CR declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

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

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsoil.2025.1675689/full#supplementary-material

References

1. Kauffman JB, Adame MF, Arifanti VB, SChile-Beers LM, Bernardino AF, Bhomia RK, et al. Total ecosystem carbon stocks of mangroves across broad global environmental and physical gradients. Ecol Monogr. (2020) 90:e01405. doi: 10.1002/ecm.1405

Crossref Full Text | Google Scholar

2. Adame MF, Cormier N, Taillardat P, Iram N, Rovai A, Sloey TM, et al. Deconstructing the mangrove carbon cycle: Gains, transformation, and losses. Ecosphere. (2024) 15:e4806. doi: 10.1002/ecs2.4806

Crossref Full Text | Google Scholar

3. Kauffman JB, Bernardino AF, Ferreira TO, Giovannoni LR, de O. Gomes LE, Romero DJ, et al. Carbon stocks of mangroves and salt marshes of the Amazon region, Brazil. Biol Lett. (2018) 14:20180208. doi: 10.1098/rsbl.2018.0208

PubMed Abstract | Crossref Full Text | Google Scholar

4. FAO. The world’s mangroves 2000–2020. Rome: FAO (2023).

Google Scholar

5. Nóbrega GN, Ferreira TO, Romero RE, Marques AGB, and Otero XL. Iron and sulfur geochemistry in semi-arid mangrove soils (Ceará, Brazil) in relation to seasonal changes and shrimp farming effluents. Environ Monit Assess. (2013) 185:7393–407. doi: 10.1007/s10661-013-3108-4

PubMed Abstract | Crossref Full Text | Google Scholar

6. Suárez-Abelenda M, Ferreira TO, Camps-Arbestain M, Rivera-Monroy VH, Macías F, Nóbrega GN, et al. The effect of nutrient-rich effluents from shrimp farming on mangrove soil carbon storage and geochemistry under semi-arid climate conditions in northern Brazil. Geoderma. (2014) 213:551–9. doi: 10.1016/j.geoderma.2013.08.007

Crossref Full Text | Google Scholar

7. Ruiz F, Bernardino A, Queiroz H, Otero XL, Rumpel C, and Ferreira T. Iron’s role in soil organic carbon (de) stabilization in mangroves under land use change. Nature communications. (2024) 15, 10433. doi: 10.21203/rs.3.rs-4650556/v1

PubMed Abstract | Crossref Full Text | Google Scholar

8. Adame MF, Connolly RM, Turschwell MP, Lovelock CE, Fatoyinbo T, Lagomasino D, et al. Future carbon emissions from global mangrove forest loss. Glob Chang Biol. (2021) 27:2856–66. doi: 10.1111/gcb.15571

PubMed Abstract | Crossref Full Text | Google Scholar

9. Bernardino AF, Queiroz HM, Nobrega GN, Coppo GC, Sanders CJ, Silva AEB, et al. Soil greenhouse gas fluxes partially reduce the net gains in carbon sequestration in mangroves of the Brazilian Amazon. Environ Res. (2024) 263:120102. doi: 10.1016/j.envres.2024.120102

PubMed Abstract | Crossref Full Text | Google Scholar

10. Bernardino AF, Mazzuco ACA, Costa RF, Souza F, Owuor MA, Nobrega GN, et al. The inclusion of Amazon mangroves in Brazil’s REDD+ program. Nat Commun. (2024) 15:1549. doi: 10.1038/s41467-024-45459-w

PubMed Abstract | Crossref Full Text | Google Scholar

11. Baudin F, Disnar J-R, Aboussou A, and Savignac F. Guidelines for Rock–Eval analysis of recent marine sediments. Org Geochem. (2015) 86:71–80. doi: 10.1016/j.orggeochem.2015.06.009

Crossref Full Text | Google Scholar

12. Baudin F, Bouton N, Wattripont A, and Carrier X. Carbonates thermal decomposition kinetics and their implications in using Rock-Eval® analysis for carbonates identification and quantification. Sci Technol Energy Transition. (2023) 78, 38. doi: 10.2516/stet/2023038

Crossref Full Text | Google Scholar

13. Wattripont A, Baudin F, de Rafelis M, and Deconinck J-F. Specifications for carbonate content quantification in recent marine sediments using Rock-Eval pyrolysis. J Anal Appl Pyrolysis. (2019) 140:393–403. doi: 10.1016/j.jaap.2019.04.019

Crossref Full Text | Google Scholar

14. Lafargue E, Marquis F, and Daniel P. Rock-eval 6 applications in hydrocarbon exploration, production, and soil contamination studies. Oil Gas Sci Technology-revue L Institut Francais Du Petrole - Oil GAS Sci Technol. (1998) 53:421–37. doi: 10.2516/ogst:1998036

Crossref Full Text | Google Scholar

15. Faure P, Schlepp L, Mansuy-Huault L, Elie M, Jardé E, and Pelletier M. Aromatization of organic matter induced by the presence of clays during flash pyrolysis-gas chromatography–mass spectrometry (PyGC–MS): A major analytical artifact. J Anal Appl Pyrolysis. (2006) 75:1–10. doi: 10.1016/j.jaap.2005.02.004

Crossref Full Text | Google Scholar

16. Schmidt MWI, Rumpel C, and Kögel-Knabner I. Evaluation of an ultrasonic dispersion procedure to isolate primary organomineral complexes from soils. Eur J Soil Sci. (1999) 50:87–94. doi: 10.1046/j.1365-2389.1999.00211.x

Crossref Full Text | Google Scholar

17. Anderson MJ. Permutational Multivariate Analysis of Variance (PERMANOVA). In: Wiley StatsRef: Statistics Reference Online (2017). p. 1–15. doi: 10.1002/9781118445112.stat07841

Crossref Full Text | Google Scholar

18. Jarde E, Vilmin F, Mansuy L, and Faure P. PY-GC/AED and chemometric correlation to characterize sewage sludges of different origins. J Anal Appl Pyrolysis. (2004) 71:553–67. doi: 10.1016/j.jaap.2003.08.008

Crossref Full Text | Google Scholar

19. Ribeiro JS, Augusto F, Salva TJG, Thomaziello RA, and Ferreira MMC. Prediction of sensory properties of Brazilian Arabica roasted coffees by headspace solid phase microextraction-gas chromatography and partial least squares. Anal Chim Acta. (2009) 634:172–9. doi: 10.1016/j.aca.2008.12.028

PubMed Abstract | Crossref Full Text | Google Scholar

20. Jacobson D, Monforte AR, and Ferreira ACS. Untangling the chemistry of port wine aging with the use of GC-FID, multivariate statistics, and network reconstruction. J Agric Food Chem. (2013) 61:2513–21. doi: 10.1021/jf3046544

PubMed Abstract | Crossref Full Text | Google Scholar

21. Pang Z, Chong J, Zhou G, de Lima Morais DA, Chang L, Barrette M, et al. MetaboAnalyst 5.0: narrowing the gap between raw spectra and functional insights. Nucleic Acids Res. (2021) 49:W388–96. doi: 10.1093/nar/gkab382

PubMed Abstract | Crossref Full Text | Google Scholar

22. Nierop KGJ, van Bergen PF, Buurman P, and van Lagen B. NaOH and Na4P2O7 extractable organic matter in two allophanic volcanic ash soils of the Azores Islands—a pyrolysis GC/MS study. Geoderma. (2005) 127:36–51. doi: 10.1016/j.geoderma.2004.11.003

Crossref Full Text | Google Scholar

23. Rumpel C, Chabbi A, Nunan N, and Dignac M-F. Impact of landuse change on the molecular composition of soil organic matter. J Anal Appl Pyrolysis. (2009) 85:431–4. doi: 10.1016/j.jaap.2008.10.011

Crossref Full Text | Google Scholar

24. Vancampenhout K, Wouters K, De Vos B, Buurman P, Swennen R, and Deckers J. Differences in chemical composition of soil organic matter in natural ecosystems from different climatic regions – A pyrolysis–GC/MS study. Soil Biol Biochem. (2009) 41:568–79. doi: 10.1016/j.soilbio.2008.12.023

Crossref Full Text | Google Scholar

25. Barreto MSC, Schellekens J, Ramlogan M, Rouff AA, Elzinga EJ, Vidal-Torrado P, et al. Effects of horticulture on soil organic matter properties in highly weathered tropical soils. Soil Tillage Res. (2021) 213:105156. doi: 10.1016/j.still.2021.105156

Crossref Full Text | Google Scholar

26. Ruiz F, Rumpel C, Dignac M-F, Baudin F, and Ferreira TO. Combining thermal analyses and wet-chemical extractions to assess the stability of mixed-nature soil organic matter. Soil Biol Biochem. (2023) 187:109216. doi: 10.1016/j.soilbio.2023.109216

Crossref Full Text | Google Scholar

27. Josse J and Husson F. missMDA: a package for handling missing values in multivariate data analysis. J Stat Softw. (2016) 70:1–31. doi: 10.18637/jss.v070.i01

Crossref Full Text | Google Scholar

28. Arnaud M, Krause S, Norby RJ, Dang TH, Acil N, Kettridge N, et al. Global mangrove root production, its controls and roles in the blue carbon budget of mangroves. Glob Chang Biol. (2023) 29:3256–70. doi: 10.1111/gcb.16701

PubMed Abstract | Crossref Full Text | Google Scholar

29. Kaal J, Martínez Cortizas A, Mateo M-Á, and Serrano O. Deciphering organic matter sources and ecological shifts in blue carbon ecosystems based on molecular fingerprinting. Science of The Total Environment. (2020) 742:140554. doi: j.scitotenv.2020.140554

Google Scholar

30. Torun F, Hostins B, Teske J, De Schryver P, Boon N, and De Vrieze J. Nitrate amendment to control sulphide accumulation in shrimp ponds. Aquaculture. (2020) 521:735010. doi: 10.1016/j.aquaculture.2020.735010

Crossref Full Text | Google Scholar

31. Boyd CE. Bottom soil and water quality management in shrimp ponds. J Appl Aquaculture. (2003) 13:11–33. doi: 10.1300/J028v13n01_02

Crossref Full Text | Google Scholar

32. Gamboa-Delgado J, Rojas-Casas MG, Nieto-López MG, and Cruz-Suárez LE. Simultaneous estimation of the nutritional contribution of fish meal, soy protein isolate and corn gluten to the growth of Pacific white shrimp (Litopenaeus vannamei) using dual stable isotope analysis. Aquaculture. (2013) 380–383:33–40. doi: 10.1016/j.aquaculture.2012.11.028

Crossref Full Text | Google Scholar

33. Tian Y, Chen G, Lu H, Zhu H, and Ye Y. Effects of shrimp pond effluents on stocks of organic carbon, nitrogen and phosphorus in soils of Kandelia obovata forests along Jiulong River Estuary. Mar pollut Bull. (2019) 149:110657. doi: 10.1016/j.marpolbul.2019.110657

Crossref Full Text | Google Scholar

34. Vivier B, David F, Marchand C, Thanh-Nho N, and Meziane T. Fatty acids, C and N dynamics and stable isotope ratios during experimental degradation of shrimp pond effluents in mangrove water. Mar Environ Res. (2019) 150:104751. doi: 10.1016/j.marenvres.2019.104751

PubMed Abstract | Crossref Full Text | Google Scholar

35. Muzuka ANN and Shunula JP. Stable isotope compositions of organic carbon and nitrogen of two mangrove stands along the Tanzanian coastal zone. Estuarine, Coastal and Shelf Science. (2006) 66, 447–458. doi: 10.1016/j.ecss.2005.10.007

Crossref Full Text | Google Scholar

36. Serrano-Grijalva L, Sánchez-Carrillo S, Angeler DG, Sánchez-Andrés R, and Álvarez-Cobelas M. Effects of shrimp-farm effluents on the food web structure in subtropical coastal lagoons. J Exp Mar Biol Ecol. (2011) 402:65–74. doi: 10.1016/j.jembe.2011.03.019

Crossref Full Text | Google Scholar

37. Mohan K, Muralisankar T, Jayakumar R, and Rajeevgandhi C. A study on structural comparisons of α-chitin extracted from marine crustacean shell waste. Carbohydr Polymer Technol Appl. (2021) 2:100037. doi: 10.1016/j.carpta.2021.100037

Crossref Full Text | Google Scholar

38. Yerlikaya P, Topuz OK, Buyukbenli HA, and Gokoglu N. Fatty acid profiles of different shrimp species: Effects of depth of catching. J Aquat Food Product Technol. (2013) 22:290–7. doi: 10.1080/10498850.2011.646388

Crossref Full Text | Google Scholar

39. Lehmann J, Hansel CM, Kaiser C, Kleber M, Maher K, Manzoni S, et al. Persistence of soil organic carbon caused by functional complexity. Nat Geosci. (2020) 13:529–34. doi: 10.1038/s41561-020-0612-3

Crossref Full Text | Google Scholar

40. Yang S, Jansen B, Absalah S, Kalbitz K, Chunga Castro FO, and Cammeraat ELH. Soil organic carbon content and mineralization controlled by the composition, origin and molecular diversity of organic matter: A study in tropical alpine grasslands. Soil Tillage Res. (2022) 215:105203. doi: 10.1016/j.still.2021.105203

Crossref Full Text | Google Scholar

41. Queiroz HM, Artur AG, Taniguchi CAK, da Silveira MRS, do Nascimento JC, Nobrega GN, et al. Hidden contribution of shrimp farming effluents to greenhouse gas emissions from mangrove soils. Estuar Coast Shelf Sci. (2019) 221:8–14. doi: 10.1016/j.ecss.2019.03.011

Crossref Full Text | Google Scholar

42. Santos-Andrade M, Hatje V, Arias-Ortiz A, Patire VF, and da Silva LA. Human disturbance drives loss of soil organic matter and changes its stability and sources in mangroves. Environ Res. (2021) 202:111663. doi: 10.1016/j.envres.2021.111663

PubMed Abstract | Crossref Full Text | Google Scholar

43. Mao F, Ullah S, Gorelick SM, Hannah DM, and Krause S. Increasing nutrient inputs risk a surge of nitrous oxide emissions from global mangrove ecosystems. One Earth. (2021) 4:742–8. doi: 10.1016/j.oneear.2021.04.007

Crossref Full Text | Google Scholar

44. Barroso GC, Abril G, MaChado W, Abuchacra RC, Peixoto RB, Bernardes M, et al. Linking eutrophication to carbon dioxide and methane emissions from exposed mangrove soils along an urban gradient. Sci Total Environ. (2022) 850:157988. doi: 10.1016/j.scitotenv.2022.157988

PubMed Abstract | Crossref Full Text | Google Scholar

45. Dai G, Zhu S, Cai Y, Zhu E, Jia Y, Ji C, et al. Plant-derived lipids play a crucial role in forest soil carbon accumulation. Soil Biol Biochem. (2022) 168:108645. doi: 10.1016/j.soilbio.2022.108645

Crossref Full Text | Google Scholar

46. Herre M, Heinze S, Heitkötter J, and Marschner B. Different factors control organic matter degradation in bulk and rhizosphere soil from the top- and subsoils of three forest stands. Soil Biol Biochem. (2022) 172:108775. doi: 10.1016/j.soilbio.2022.108775

Crossref Full Text | Google Scholar

47. Weiss C, Weiss J, Boy J, Iskandar I, Mikutta R, and Guggenberger G. Soil organic carbon stocks in estuarine and marine mangrove ecosystems are driven by nutrient colimitation of P and N. Ecol Evol. (2016) 6:5043–56. doi: 10.1002/ece3.2258

PubMed Abstract | Crossref Full Text | Google Scholar

48. Luo L, Meng H, Wu R, and Gu J-D. Impact of nitrogen pollution/deposition on extracellular enzyme activity, microbial abundance and carbon storage in coastal mangrove sediment. Chemosphere. (2017) 177:275–83. doi: 10.1016/j.chemosphere.2017.03.027

PubMed Abstract | Crossref Full Text | Google Scholar

49. Kristensen E and Holmer M. Decomposition of plant materials in marine sediment exposed to different electron acceptors (O2, NO3–, and SO42–), with emphasis on substrate origin, degradation kinetics, and the role of bioturbation. Geochim Cosmochim Acta. (2001) 65:419–33. doi: 10.1016/S0016-7037(00)00532-9

Crossref Full Text | Google Scholar

50. Lovelock CE, Ball MC, Martin KC, and Feller I. Nutrient enrichment increases mortality of mangroves. PloS One. (2009) 4:e5600. doi: 10.1371/journal.pone.0005600

PubMed Abstract | Crossref Full Text | Google Scholar

51. Wei L, Hong H, Bee MY, Wu Y, Ndayambaje P, Yan C, et al. Different adaptive strategies of three mangrove species to nutrient enrichment. Plant Ecol. (2022) 223:1093–102. doi: 10.1007/s11258-022-01260-7

Crossref Full Text | Google Scholar

52. Stagg CL, Schoolmaster DR, Krauss KW, Cormier N, and Conner WH. Causal mechanisms of soil organic matter decomposition: deconstructing salinity and flooding impacts in coastal wetlands. (2017). doi: 10.1002/ecy.1890

PubMed Abstract | Crossref Full Text | Google Scholar

53. Sippo JZ, Santos IR, Sanders CJ, Gadd P, Hua Q, Lovelock CE, et al. Reconstructing extreme climatic and geochemical conditions during the largest natural mangrove dieback on record. Biogeosciences. (2020) 17:4707–26. doi: 10.5194/bg-17-4707-2020

Crossref Full Text | Google Scholar

54. Gomes LE, de O, Sanders CJ, Nobrega GN, Vescovi LC, Queiroz HM, et al. Ecosystem carbon losses following a climate-induced mangrove mortality in Brazil. J Environ Manage. (2021) 297:113381. doi: 10.1016/j.jenvman.2021.113381

PubMed Abstract | Crossref Full Text | Google Scholar