Global warming and sea-level rise are directly influencing the growth, distribution, and greenhouse gas emissions of mangrove forests. However, mangrove forests growing at their latitudinal limits are relatively susceptible to warming; nevertheless, few studies have focused on GHG emissions of latitudinal limits mangrove forests as related to global climate change. To address this knowledge gap, a multi-year in situ control experiment was established in a restored plantation at the northern distribution limit of Kandelia obovata, the most cold-tolerant mangrove species in China, to simulate both warming and sea-level rise. We investigated the growth patterns and sediment greenhouse gas emissions of a K. obovata population and identified the primary factors contributing to these changes. The results showed that warming and moderate sea-level rise enhanced biomass by more than 18%, indicating that warming stimulated plant growth while excessive sea-level rise inhibited it. The sediment greenhouse gas emissions ranged from 45.5 to 484.6 mgCO₂ m-2 h-1, 5.6 to 590.3 μgCH₄ m-2 h-1, and 11.4 to 385 μgN₂Om-2 h-1, which increased with warming while decreased with sea-level rise, acting as the net source of greenhouse gas emission. Our study predicted that sea-level rise, while directly changing sediment properties, had combined effects with warming on these studied mangrove forests that were predicted to emit more greenhouse gases from sediments in the future. These findings indicated that K. obovata plantations within their latitudinal limits tend to accumulate more CO₂ for biomass carbon storage under warming conditions, while stimulating sediment greenhouse gas emissions, which will offset their climate mitigating effect in future climatic scenarios.

Greenhouse gases sampling, flux measurements, and global warming potential analysis

In this experiment, sediment greenhouse gases were collected using a Closed Static Chamber. The chamber covered 0.013 m² of sediment surface with a headspace volume of 4.5 liters. The chamber was placed on the sediment surface with the edge inserted 2 cm into the sediment; the round hole channel at the top was plugged with a soft rubber plug, and a syringe needle was inserted into the rubber plug to maintain the air pressure balance inside and outside the chamber. The pressure was controlled in the chambers with a small vent on the chamber top. Chambers were sampled between 09:00 and 15:00 h after 2 h of an ebb tide. Air temperature (T_a) and sediment surface temperature at a depth of 0–5 cm (T_s) inside chambers at the beginning and end of the gas sampling process were measured using a HOBO MX2202 air temperature sensor and a sediment thermometer. For each gas sample, a 0.5# needle was connected to a 60 mL syringe through the air sampling port to collect each 20 mL of gas sample. The greenhouse gases were extracted from the chamber at 0-, 30-, and 60-min intervals, and then the gas samples in the syringes were transferred to a special gas collection bag immediately (Hede Technologies Ltd.). Each gas sample was injected into 50 mL air bags (Hede Technologies Ltd.) and analyzed within 24 h using a 7890B Greenhouse Gas Chromatograph (Agilent Technologies, Inc.). The flame ionization detector and electron capture detector were set at 25 °C and 30 °C, respectively, with N₂ as the carrier gas at a flow rate of 2 mL·min⁻¹, and mixed air and H₂ into the flame ionization detector and electron capture detector at flow rates of 450 and 50 mL·min⁻¹, respectively.

A set of standard gases (Linde China, http://www.linde-gas.com.cn/) was used to test the equipment after measuring every 10 samples. Standard curves were established between the concentrations of CO₂, CH₄, and N₂O according to the peak area of each set of standard gases.

Gas fluxes were calculated using Eq. 1:

Equation 1:

F = (M × P × V) / (R × (273 + T) × A) × (dC/dt)

Where:

• F is the gas flux (mg·m⁻²·h⁻¹ or μg·m⁻²·h⁻¹)
• M is the molar mass of each greenhouse gas (g·mol⁻¹)
• P is the atmospheric pressure (P = 1.013 × 10⁵ Pa)
• R is the universal gas constant (8.314 Pa·m³·mol⁻¹·K⁻¹)
• T is the average air temperature (°C)
• V is the volume of the sampling chamber (m³)
• A is the chamber surface area (m²)
• dC/dt is the emission rate of gas concentration (ppm·h⁻¹)

The global warming potential of the three greenhouse gases was calculated to evaluate the contribution of each greenhouse gas to global warming. According to the Intergovernmental Panel on Climate Change (IPCC AR6 WGI, 2021), we used values of 100-year global warming potential (GWP₁₀₀) measurements of 27.9 for CH₄ and 273 for N₂O. These indicate that 1 kg CH₄ and 1 kg N₂O are 27.9 and 273 times more potent than 1 kg of CO₂ in contributing to global warming over 100 years, respectively.

Carbon dioxide removal analysis

The carbon dioxide removal of mangrove forests in this study was calculated using Eq. 2. It represents the net carbon gain from biomass and greenhouse gas-equivalent CO₂, which reflects CO₂ fixation by mangrove forests. In this study, the carbon dioxide removal potential did not include sediment carbon increment.

Equation 2:

CDR = ΔBiomass × Cc × (44 / 12) − Σi=13GWPi

Where:

• ΔBiomass is the biomass increment per square meter
• C_c is the carbon content of mangrove biomass (default = 0.47)
• GWP_i is the global warming potential for each greenhouse gas

Statistical analyses

The relationship between the R_s rate and T_s was fitted using an exponential model (Eq. 3; Lloyd and Taylor 1994):

Equation 3:

R = ebTs

Where:

• R is the gas flux rate (mg·m⁻²·h⁻¹)
• T_s is the sediment temperature (°C)
• b is the temperature response coefficient

The responsiveness of the R_s rate to temperature is expressed by the Q₁₀ value (Eq. 4):

Equation 4:

Q10 = e10b

Where b is the same as in Equation 3.

All data distributions were tested for normality with the Shapiro–Wilk normality test, and Levene’s test was performed to determine the homogeneity of variances. The Kruskal–Wallis test was used to test differences between groups. Two-way ANOVA was used to examine the main and interaction effects of warming and sea-level rise on plant biomass, sediment properties, greenhouse gas fluxes, and global warming potential. Nonlinear regression was performed to explore the relationship between greenhouse gas fluxes and T_s. Correlation analysis was performed to detect the relationships between greenhouse gas emissions and driving factors such as warming, pH, and carbon content. The contribution of driving factors to R_s, CH₄, and N₂O was determined using the hierarchical partitioning method through the R package rdacca.hp (Lai et al. 2022). A structural equation model was conducted to analyze the linkage between driving factors and sediment greenhouse gas emissions using the piecewiseSEM package in R.