Introduction

Tidal forested wetlands are situated between terrestrial and aquatic ecosystems. Although the global area is unknown, tidal forested wetlands occur in large areas along the U.S. Gulf and Atlantic, Europe, Central America, and the Amazon (Adame et al. 2024). Within the United States, tidal forested wetlands occupy > 400 km² in South Carolina, > 230 km² in Georgia, and > 828 km² in Louisiana (Shaffer et al. 2016; Krauss et al. 2018); actual tidal forested wetland area is likely to be much higher than currently realized. Tidal forested wetlands play an important role in sequestering and storing “blue carbon,” like mangroves, saltmarshes, and seagrasses (Krauss et al. 2018; Adame et al. 2024), thereby helping to mitigate negative impacts of climate warming. Nevertheless, the location and low-lying position of tidal forested wetlands make them especially vulnerable to saltwater intrusion from climate change-induced sea-level rise (SLR), drought, flooding and increased storm surges (Neubauer et al. 2013; Noe et al. 2013; Pierfelice et al. 2015; Kirwan and Gedan 2019).

The Maurepas Swamp is the second largest contiguous coastal swamp forest in Louisiana, containing 776 km² of freshwater forested wetlands, and about 52 km² of fresh and oligohaline marshes (Shaffer et al. 2016). It was once naturally connected to the Mississippi River (MR) and received freshwater input from the MR with high nutrients and sediment loads. Historically, the Maurepas forests were 90% baldcypress–water tupelo (Taxodium distichum–Nyssa aquatica) tidal swamps (Shaffer et al. 2016). The Lake Maurepas swamp forests are productive with the highest above-ground net primary productivity (ANPP) reaching ~932 g C m^-2 yr^-1 in the area near the MR with an average of ~485 g C m^-2 yr^-1 and range of 195 - 932 g C m^-2 yr^-1 in the entire swamp (Shaffer et al. 2003; Hoeppner et al. 2008; Shaffer et al. 2016). This highest ANPP is comparable to mangrove forests (mean ± SE: 1,320 ± 170 g C m^-2 yr^-1, median: 900 g C m^-2 yr^-1, and higher than that of saltmarshes (mean ± SE: 500 ± 30 g C m^-2 yr^-1, median: 370 g C m^-2 yr^-1) (Alongi 2020). However, over the last century, this swamp has become highly degraded with high tree mortality and conversion to freshwater marshes. The swamp degradation was attributed to subsidence, permanent flooding, lack of mineral sediment and nutrient input, nutria herbivory, and saltwater intrusion from Lake Pontchartrain as well as canal dredging, specifically a large shipping channel, which exacerbates surge, high tide salinity pulses, and drought (Shaffer et al. 2003; Effler and Goyer 2006; Hoeppner et al. 2008; Day et al. 2009a, b; Shaffer et al. 2009a, b; Middleton et al. 2015; Hunter et al. 2016; Shaffer et al. 2016).

The impacts of saltwater intrusion range from stress and mortality of individual trees to total tree mortality and conversion of swamp forests to tidal marshes and open water (Krauss et al. 2018). Swamp forest basal area was reduced from 44 – 87 m² ha^-1 without salinity impacts to 23 – 36 m² ha^-1 with salinity levels of 1.3 – 3.0 parts per thousand (ppt) in South Carolina and Louisiana (Krauss et al. 2007), on average a 13.4 m²/ha decrease per ppt salinity specifically for Louisiana (Krauss et al. 2017). These salinity impacts have continued over time and led to up to 75% reduction in live tree aboveground carbon stock over a 7-year period (Shipway et al. 2024). Moreover, the magnitude and direction of changes in gross primary productivity (GPP), net primary productivity (NPP), ecosystem respiration (ER), methane (CH₄) and nitrous oxide (N₂O) emissions, soil carbon sequestration, and carbon storage in tidal forested wetlands along the Waccamaw River and the Savannah River varied greatly with site-specific soil salinity and water level conditions, implying that there are large uncertainties in blue carbon accounting and budget estimation from tidal forested wetlands (Wang et al. 2022, 2023a). For salinity impacted tidal forests along the Savannah River, drought caused soil salinity to increase from 3.46 ± 2.12 to 7.23 ± 0.72 ppt, resulting in a -64% (-106 g C m^-2 yr^-1, from 167 to 61 g C m^-2 yr^-1) reduction in NPP compared to NPP versus pre-drought, and, soil carbon dioxide (CO₂) flux increased by 140 g C m^-2 yr^-1 from 116 to 256 g C m^-2 yr^-1, equating to a 120.7% change (Wang et al. 2022). Under the same drought condition, CH₄ fluxes switched from emission of 0.81 g C m^-2 yr^-1 to uptake of 0.91 g C m^-2 yr^-1, whereas N₂O emissions increased from 0.19 to 1.09 kg N ha^-1 yr^-1 (Wang et al. 2023a). In contrast, for the upper tidal river (less salinity impacted) swamp forest site along the same river, NPP increased from 240 g C m^-2 yr^-1 pre-drought to 362 g C m^-2 yr^-1 (51% increase) under drought, while soil CO₂ flux remained relatively constant (178 vs. 176 g C m^-2 yr^-1) (Wang et al. 2022), CH₄ emissions increased from 56 to 144 kg C ha^-1 yr^-1, and N₂O declined slightly from 0.51 to 0.47 kg N ha^-1 yr^-1 (Wang et al. 2023a). Although carbon fluxes and greenhouse gas (GHG) emissions are critical components of carbon budget estimation in tidal forested wetlands, data and understanding of carbon dioxide exchange between the atmosphere and forested wetlands remain limited. There are very few measurements of net ecosystem exchange (NEE) of CO₂ from tidal forested wetlands that can be used to determine if the forested wetlands are carbon sources or sinks of atmospheric CO₂ under different ecosystem-based management scenarios, such as large-scale river re-introduction projects (aka, diversions).

To combat the degradation of the Lake Maurepas swamp forests, Louisiana Coastal Protection and Restoration Authority (CPRA) are building the River Reintroduction into Maurepas Swamp Project (CPRA 2023). A Mississippi River reintroduction into the Maurepas Swamp is designed to restore and sustain the natural hydrology of one of the largest remaining forested wetlands along the U.S. Gulf coast (e.g. Day et al. 2009a). Construction of the diversion project began in December 2024 and is expected to be completed in 2029. However, although the growth and mortality of tree species in Lake Maurepas swamp have been intensively studied over the past three decades (Shaffer et al. 2003; Hoeppner et al. 2008; Shaffer et al. 2009a, b; Shaffer et al. 2016), our understanding of projected changes in carbon fluxes and GHG emissions in the swamp forests with hydrologic regime change is unknown. The objectives of this present study are to 1) validate a process-based wetland biogeochemistry model for the Lake Maurepas swamp forests in the Mississippi River Deltaic Plain (MRDP); 2) use the calibrated model to determine the spatial variability in NEE, NPP, ER, CH₄, and N₂O emissions; and 3) examine the impacts of climate change induced drought, flooding, and SLR, as well as a river diversion on these carbon fluxes and GHG emissions. Results of this modeling study could help monitoring and adaptive management of Lake Maurepas forested wetlands as a major restoration tool to combat the impacts of SLR and salinity intrusion, understand MR diversion potential, and improve understanding of the potential effects of river reintroduction as a coastal management approach.

Methods

Results

Discussion

Our first discovery is that Lake Maurepas swamp forests have high net carbon uptake capacity, reflected through large net ecosystem exchange (NEE) of CO₂ and moderate CH₄ fluxes. NEE in the Lake Maurepas swamp forests ranged from -1,143 to -1,650 g C m^-2 yr^-1 across the twelve CRMS sites under average hydrologic conditions. NPP in the Lake Maurepas swamp forests from this study were at the high end of a previously reported range of 886.7 to 1,779.9 g C m^-2 yr^-1 for cypress and tupelo dominated swamp forests of Louisiana (Rivera-Monroy et al. 2019). These large and varying carbon fluxes indicate that Lake Maurepas swamp forests are hotspots of carbon cycling, our modeling revealed that biogeochemical processes among sites drive quite a lot of variability among specific sites. In order to quantify carbon processes properly with traditional study plots, many would be required, offering modeling as perhaps the most reasonable approach.

Measurements of NEE for tidal forested wetlands are uncommon. Modeled NEE was in the range of -630 to -1,929 g C m^-2 yr^-1 for tidal freshwater forests (salinity < 0.5 ppt) along the floodplains of the Waccamaw River, South Carolina and the Savannah River, Georgia (Wang et al. 2022). Simulated NEE values in Lake Maurepas swamp forests were in the middle to high end of this NEE range. Our NEE values were even higher than the measured NEE values (-900 to -1,000 g C m^-2 yr^-1) of the cypress swamp in the Big Cypress National Preserve in southern Florida (Shoemaker et al. 2015). Thus, Lake Maurepas swamp forests are potentially among the larger sinks of atmospheric CO₂ in the southeastern United States. Furthermore, NEE in Lake Maurepas swamp forests were higher than that of other tidal wetland types (e.g. NEE from -45.6 to -337.2 g C m^-2 yr^-1 for tidal marshes, Windham-Myers et al. 2018), suggesting that the swamp forests play a more critical role than other tidal wetland types in taking up atmospheric CO₂ and storing large amounts of carbon in vegetation and soils along the Gulf coast. A large area of tidal wetlands along the Gulf coast has been lost or degraded due to climate change, SLR, land use change, and disruption of freshwater flow and sediment supply, resulting in losses of sequestered and stored blue carbon and degradation of what remains (e.g. Lane et al. 2016; Wang et al. 2024). Because of this degradation, coastal wetland restoration is positioned to facilitate uptake of CO₂ on broad scales.

Our second discovery is that the high net carbon uptake capability in the Lake Maurepas swamp forests could be significantly affected by climate change induced drought, flooding, and SLR depending on the direction and magnitude of hydrologic regime and salinity changes. Simulated NEE declined under drought conditions for all sites as well as under high SLR conditions for the degraded and relict sites. However, the capacity to take up CO₂ increased under the wet conditions for all sites and under both SLR scenarios for the throughput sites because of the higher elevation at the throughput sies to handle more SLR. Although both drought and high SLR resulted in the decreases in NEE, the major driving factors could be different. Under the drought condition, the reduction in NEE is mainly attributed to the increased ER rather than the decrease in GPP and NPP. Soil respiration with drought increased by an average of 232% (range: 45 to 1,259%) whereas plant respiration decreased slightly by -8.4% (range: -4.8 to -14%) compared to the average condition. The increase in soil respiration reflected the increase in soil organic matter decomposition and mineralization from the lower soil water table depth under the drought condition (Wang et al. 2022). Wang et al. (2023b) found that the water table depth threshold for CO₂ emissions was about 5 cm above soil surface for tidal forested wetlands and CO₂ emissions declined dramatically once that threshold was exceeded.

Water table depths at the swamp forest sites under the drought condition were reduced from 11 to 45 cm above soil surface to < 5 cm, except CRMS0090, CRMS0089, and CRMS0063. In contrast, the decrease in NEE under the high SLR for the degraded and relict sites is due to the larger reduction in GPP and NPP than in ER. The decline in NPP under high SLR is due to the salinity stress from the increased soil salinity (mean soil salinity as high as 5.0 ppt at the degraded sites). Wood growth and litterfall were found to decline in Lake Maurepas swamp forests during drought years (2000 – 2002) when soil salinity increased from < 0.5, < 1.0, and < 2.5 ppt under non-drought years to 1 – 2, 2 – 3, and 3 – 5 ppt during drought year 2000 for the throughput, relict, and degraded forests, respectively (Shaffer et al. 2003; 2016). Declines in wood growth and litter fall were also observed in tidal forested wetlands along the Waccamaw and Savannah rivers in South Carolina and Georgia (Cormier et al. 2013; Pierfelice et al. 2015). The precipitous decrease in NPP in forest sites with soil salinity in exceedance of the 2 – 3 ppt thresholds (Wang et al. 2023b) can be explained by reduction in photosynthesis as reflected by reduction of sapflow under salinity stress (Krauss and Duberstein 2010) and nutrient stress particularly of P, resulting in inhibition of nutrient uptake (e.g. Zhai et al. 2018).

On the other hand, NEE increased under the wet year for all sites and under high SLR at the throughput sites. Such increases in NEE can be explained by the decreased ER and increased NPP. Under the high SLR, ER decreased by -13.7% (range: -8.6 to -19.3%) while NPP increased slightly by approximately 2% (range: 0.2 to 3.1%) for the throughput sites. Tidal wetlands respond to rising sea levels by increasing belowground productivity and altering root characteristics to influence the elevation of the soil surface, thus sequestering additional carbon consequently (Rogers et al. 2019). Throughput sites had a greater capacity for this response with higher SLR than degraded and relict sites with higher NEE from increased NPP especially belowground productivity, suggesting that throughout sites may be more responsive to potential management (e.g. river diversion) under high SLR. However, NPP in tidal forested wetlands is impacted more by soil salinity than water table depth, confounding hydrology-only conclusions, and NPP is reduced when salinity is in exceedance of the 2 – 3 ppt threshold (Wang et al. 2023b). Under the wet condition, NPP increased by approximately 16.4% (range: 12.3 to 23.1%) while ER decreased by -4.5% (range: -2.5 to -8%). Soil salinity under the wet condition was less than 0.5 ppt, well under the 2 – 3 ppt threshold. Under the high SLR condition, soil salinity at the throughput sites was around 1.5 ppt, still less than the 2 – 3 ppt threshold (Wang et al. 2023b), thus still in favor of plant growth for carbon sequestration over soil respiration. Under the wet condition, tidal swamp forests could experience frequent inundation, resulting in waterlogged soils with low oxygen availability. Nevertheless, baldcypress and water tupelo are species tolerant of flooding (Hook 1984), and nutrients may even partially offset flood stress by increasing growth potential, as documented for water tupelo (Effler and Goyer 2006). With reductions in soil oxygen, the efficiency of aerobic microbial activity will be suppressed in favor of a greater capacity to lose carbon through methanogenesis.

The Lake Maurepas swamp forests are net sources of CH₄ and N₂O emissions at all sites under all hydrologic conditions (except CRMS3913 under the dry year), but carbon release from methane emissions is small compared to release of CO₂ into the atmosphere. CH₄ and N₂O emissions in Lake Maurepas swamp forest were relatively small compared to NEE. For example, CH₄ emissions ranged from 10 to 30 g C m^-2 yr^-1 excluding CRMS3913 and N₂O emissions ranged from 0.05 to 0.3 g N m^-2 yr^-1 under the average condition. CH₄ was less than 2.8% of simulated NEE, ranging from -2,154 to -3,769 g C m^-2 yr^-1. Therefore, CH₄ emissions played a less important role in the simulated carbon budgeting within Maurepas’ swamp forests. However, when converting CH₄ and N₂O fluxes into CO₂ equivalents for understanding global warming potential (GWP), CH₄ and N₂O were amplified 25 to 298 times based on a 100-year period from 2007 IPCC Fourth Assessment Report. In this way, we discovered that flux of both molecules equal approximately 5 to 29% of net ecosystem exchange of CO₂, which is not insignificant to climate change. This study confirmed the observation that CH₄ emissions declined along an increasing soil salinity gradient (up to 30 ppt) in tidal wetlands (e.g. Poffenbarger et al. 2011), which we also applied to Lake Maurepas swamp forests under the high SLR condition. Most CH₄ emissions would be suppressed when soil salinity exceeds 3 ppt for Lake Maurepas swamp forests (Fig. 7). Reduction of CH₄ emissions under the high SLR scenario may be attributed to the inhibition of methanogenesis due to the increased activities of sulfate-reducing bacteria with saltwater intrusion (Wang et al. 2023a), creating a trade-off between greater CH₄ emissions with intensified flooding and reduced CH₄ emissions with salinity in swamp forests (Krauss and Whitbeck 2012). The presence of sulfate in soils allow sulfate-reducing bacteria to outcompete methanogens for energy, resulting in inhibition of methane production by sulfate-reducing bacteria (Poffenbarger et al. 2011; Holm et al. 2016; Krauss et al. 2016). Furthermore, soil water level was a critical limiting factor for N₂O emissions. Increased N₂O emissions in the Lake Maurepas swamp forests under the drought condition reflected the larger role of soil water level than soil salinity on redox potential and on nitrification and denitrification (Fig. 7). Soil redox potential increased under the drought conditions and nitrification was dramatically enhanced, leading to the large increase in N₂O production because nitrification can also produce N₂O during the NH₄⁺ oxidation (Murray et al. 2015). Nitrification of NH₄⁺ is a relatively unstudied pathway for N₂O production in swamp forests, possibly even surpassing denitrification as an N₂O and NO source in some settings (Krauss and Whitbeck 2012). In this study, the incomplete nitrification and incomplete denitrification could be attributed to increased N₂O production under drought conditions with insufficient oxygen supply (Rassamee et al. 2011). Increased N₂O emissions under drought conditions may also be attributed to stimulated organic carbon decomposition because labile organic carbon is significantly correlated with N₂O emission rates because denitrification, in lieu of just nitrification, is favored to occur with high availability of organic matter (Swarzenski et al. 2008).

Our third discovery is that the response of net carbon uptake capacity of the Lake Maurepas swamp forests to a MR diversion is bi-directional, triggering both increases and decreases in net carbon uptake capacity related to site characteristics. However, uptake of carbon from CO₂ was consistent with diversion scenario, suggesting that a MR diversion into the Lake Maurepas swamp forests may be beneficial to carbon sequestration and storage. Specific site characteristics control the magnitude of enhanced carbon uptake. In this study, NEE increased by an average of ~9% (range: 4 to 19%) at 9 out of the 12 swamp sites under MR diversion compared to the average condition. Due to the decrease in NEE at three sites (average: -16%, range: -6 to -26%), the site-averaged NEE across all the swamp forests under MR diversion only increased about 3% compared to the average condition. The increase in NEE at the 9 sites is the result of the larger increase in NPP (14%), particularly from throughput sites, than the increase in ER (6%). There are no field measurements on tree productivity under river diversions per se, but Middleton et al. (2015) found that aboveground growth of baldcypress was enhanced in a year with higher river water loading into Louisiana’s coastal wetlands as a mitigation tool to prevent oiling after the Deepwater Horizon oil spill. Furthermore, the MR diversion at Caernarvon, south of New Orleans, promoted the growth of marshes by enabling high belowground biomass production (Day et al. 2009b). For the Lake Maurepas swamp forests, woody and herbaceous vegetation are nutrient limited (Lane et al. 2003; Shaffer et al. 2003). Therefore, diversion of freshwater associated with mineral sediments and nutrients will likely stimulate wetland productivity (Martin 2002; Hoeppner et al. 2008; Day et al. 2009b). For example, Day et al. (2009b) found that the high marsh productivity at the site in the diversion outfall area was likely the result of increased nutrient availability, decreased porewater salinity, higher redox potential, and lower porewater sulfide concentrations. Soil salinity in the Lake Maurepas swamp forests under the MR diversion scenarios is still less than 0.5 ppt, so the increase in NPP is unlikely due to the lowering of salinity, which would directly suppress forest productivity through osmotic stress and ion toxicity (e.g. DeLaune et al. 2003; Swarzenski et al. 2008; Wang et al. 2022). Thus, the increase in NPP under river diversion is most likely attributed to increased soil carbon and nutrient (e.g. N and P) availability from diverted river water. In fact, soil organic carbon and soil nitrogen fluxes increased under river diversion compared to the average condition (Wang et al. 2024). On the other hand, the decreased NEE with MR diversion at the three sites (CRMS3913, CRMS0047, CRMS0061) could be explained by the relatively larger increase in ER (>18%) than the increase in NPP (~15%). The large increase in ER under river diversion could be attributed to increased soil organic matter decomposition under diversion similar to the findings of Swarzenski et al. (2008) where chronic riverine input led to enhanced soil organic matter decomposition in highly organic freshwater marshes. The three sites (CRMS3913, CRMS0047, CRMS0061) with decreased NEE under the diversion scenario are located far away from diversion outfall area (Fig. 1) and at relatively high elevation (0.24 – 0.32 m NAVD88) compared to other sites (0.05 – 0.14 m NAVD88) (CRMS 2025). With the 0.08 m increase to their water level under the average condition from the diversion design, their water table depths ranged 0.19 to 0.23 m, resulting in higher ER, CH₄ and N₂O emissions (Figs. 7d, f, h), thus lower NEE, compared to other sites. The lowest NEE at CRMS0061 (-1,143 g C m^-2 yr^-1, a throughput site, Table 3) under the average condition and reduction in NEE (net carbon uptake capacity) could be attributed to the high ER that would offset the contribution of high NPP and also be attributed to the lower soil organic matter content (18%, CRMS data) with insufficient nutrient availability for vegetation growth. The implication of this study for MR diversion is that delivery of diverted freshwater to forest stands at high elevation and/or far away from diversion outfall area via increased discharge or length of channel can help to increase nutrient availability and mitigate drought stress.

CH₄ emissions varied greatly by site and were reduced substantially by diversion, especially at the throughput sites. These results are consistent with previous diversion studies. For example, Holm et al. (2016) found that CH₄ emissions in receiving marshes was inversely related with MR diversion discharge at Davis Pond although the relationship was weak because the marsh location was outside of the dominant flow paths. Furthermore, Holm et al. (2016) pointed out that near-field, localized areas and/or higher discharge conditions would be required to stimulate a detectable effect on methane efflux. Methane emissions would be suppressed by increased river discharge, given the potential for increased loading of river-borne alternate electron acceptors, such as nitrate, sulfate, and iron-oxides (Holm et al. 2016). There tend to be critical inundation levels for methane emissions due to the interplay between methanogenesis, methanotrophy, and plant-mediated transport (Calabrese et al. 2021; Wang et al. 2023b). This global critical water table level is about 0.5 m above soil surface for wetlands including swamp forests (Calabrese et al. 2021). Higher than the threshold inundation level, methanogenesis could be decreased due to the declined vegetation productivity and diluted organic substrate concentration. Below this threshold, CH₄ oxidation by methanotrophic bacteria occurs in unsaturated zones of the soil column and the saturated zone and the water column (Calabrese et al. 2021). For this study, the critical water table depth is about 0.35 m (Fig. 7f), providing evidence for the observation of Calabrese et al. (2021) that “the specific location of the critical water level above the soil surface may differ depending on wetland characteristics”. Additionally, N₂O emissions were reduced under diversion in the Lake Maurepas swamp forests. This reduction in N₂O emissions could be attributed to the increased uptake of ${\text{NO}}_3^{-}$ by plant growth, leading to reduced ${\text{NO}}_3^{-}$, therefore, less N₂O emissions by denitrification (e.g. White et al. 2019). Overall, our modeling results demonstrate that MR diversion could promote plant productivity more than increased ecosystem respiration (e.g. soil respiration from organic matter decomposition), inhibit CH₄ at most of the sites and N₂O emissions at all sites, leading to an increase of the net carbon uptake capacity. This finding can have management implications. River diversions into degraded tidal swamp forests that have stagnant, standing water and mineral sediment and nutrient deprivation can be an effective solution to enhancing carbon storage and restoring important energy transformations, perhaps even serving as a tool for helping to mitigate climate change if implemented over large scales. This approach can not only promote land building and vertical accretion through enhanced sediment deposition and SOC production (e.g., Couvillion et al. 2013; Wang et al. 2017) but also facilitate the net carbon uptake capacity via increasing plant productivity to outpace soil organic matter decomposition, and as we show here, reduce CH₄ and N₂O emissions.

In this study, only wood growth and litterfall data were available and field data on NEE, ER, CH₄ flux, and N₂O flux were not available, so we were unable to validate NEE, ER, CH₄ and N₂O emissions for the Lake Maurepas swamp forests. Nevertheless, the WCAT-DNDC model was previously validated for CH₄ and N₂O emissions at the Savannah River swamp forest sites along the Georgia Coast using static chamber methods (Wang et al. 2023a) and for NEE, GPP, ER and CH₄ at the brackish marsh sites in Terrebonne Basin along the Louisiana Coast using the eddy covariance technique (Wang et al. 2024). When it was applied to Lake Maurepas swamp forest sites, the model structure (e.g., impact of soil salinity on litterfall and wood growth, SOM decomposition, methane emissions, Wang et al. 2022) was not changed. The model tended to overestimate CH₄ emissions (bias = 0.004, RMSE = 0.026 kg C ha^-1 day^-1) whereas it tended to underestimate fluxes of N₂O (bias = -0.84, RMSE = 1.74 g N ha^-1 day^-1) in swamp forests (Figs 3 and 4 in Wang et al. 2023a). The model tended to underestimate NEE (bias = −0.03 g C m⁻² day⁻¹), and slightly overestimate GPP, ER and CH₄ fluxes (biases = 0.08 g C m⁻² day⁻¹, 0.23 g C m⁻² day⁻¹, and 0.03 kg C ha⁻¹ day⁻¹, respectively) in brackish marshes (Wang et al. 2024). Continuous and high-frequency field monitoring and measurements of NEE, ER, CH₄ flux, and N₂O flux could be beneficial for model improvement to reduce uncertainty in simulating the biogeochemical processes in the Lake Maurepas swamp forests. In addition, field studies on carbon exchange between swamp forests and river tributaries (mostly as carbon sink), Lake Maurepas, Lake Pontchartrain, as well as downstream estuaries (as carbon source) in the form of lateral flux of dissolved organic carbon (DOC), dissolved inorganic carbon (DIC), and particulate organic carbon (POC) could enable the calibration and validation of WCAT-DNDC model. In lieu of field studies, the model has been shown to be capable of simulating these fluxes for producing new insight from future modeling exercises. Lateral carbon fluxes are the most understudied component of tidal wetland carbon fluxes and can be large in magnitude but difficult to measure directly (Maher et al. 2018). Furthermore, soil salinity amplification by a factor of 3 in the dry year was used for the high SLR scenario based on the observed large increase in soil salinity at Lake Maurepas swamp under a severe drought condition discovered back in 2000. This assumption is one of the sources of uncertainty on simulation results. A more accurate magnitude and distribution of soil salinity under the low and high SLR scenarios would require hydrodynamic modeling with incorporation of the mixing of freshwater and saltwater across the landscape from sea to swamp, while incorporating the interactions of geomorphology, soil, vegetation, wind, and wave conditions for improvement. Finally, it has been argued that a large diversion (e.g. 1,416 m³ s^-1of river flow) is needed to deliver sufficient sediments to achieve high accretion rates and stimulate organic soil formation in Lake Maurepas swamp forests (Shaffer et al. 2016). This study simulated a river diversion currently under construction with a much lower discharge rate (up to 71 m³ s^-1) under specific operational timing and durations of water delivery from a very specific location along the Mississippi River. Changes in carbon fluxes and GHG emissions in the Lake Maurepas swamp forests under a larger river diversion could be further studied with hydrological and biogeochemistry monitoring, but models such as WCAT-DNDC could support river diversion management both in advance of project design and as adaptive management once a diversion project is implemented. Future plans include monitoring the Maurepas Swamp forest response after the current 71 m³ s^-1 diversion is completed, using new empirical data to improve the WCAT-DNDC model, and applying the model to larger diversion scenarios.