Abstract
Background: Coastal mangrove wetlands have great carbon sequestration potential, but assessments lack precision due to limited CO2 flux monitoring stations. This study aims to clarify the variation patterns and mechanisms of carbon sinks in typical mangrove ecosystems and propose a viable framework to address sparse CO2 flux station coverage. Methods: This study systematically investigated the carbon sequestration potential, temporal dynamics, and driving mechanisms of mangroves in the Futian Mangrove Nature Reserve, Shenzhen, via eddy covariance (EC) observations and machine learning-based modeling, while conducting multi-scale comparative analyses with regional and global mangrove ecosystems. Results: Results revealed that the Shenzhen mangrove ecosystem maintained a robust net carbon sink capacity during the study period. Monthly net ecosystem production (NEP) was lower in August and higher in December, a pattern primarily driven by attenuated nighttime respiratory activity in the dry season. Further comparative analyses highlighted the Shenzhen Futian mangroves as a high-efficiency carbon sink at both regional and global scales, with the annual NEP of 1265.5 g C m⁻2 a⁻1, which is higher most of the investigated sites in this study. This high NEP in Shenzhen likely arises from the synergistic interplay of a favorable subtropical marine monsoon climate, the resilience of native species, and advanced conservation management practices. Furthermore, we developed a data-driven RF regression-based NEP prediction model integrating multi-source environmental variables, which achieved good predictive performance for the Shenzhen observations and showed preliminary transferability to two congeneric Kandelia-dominated mangrove sites. Conclusions: This study provides a novel framework to address the critical gap of sparse EC station coverage in global mangrove ecosystems. These findings may help refine mangrove blue-carbon accounting and provide methodological support for future Monitoring, Reporting, and Verification (MRV) systems and coastal wetland management.
| Original language | English |
|---|---|
| Article number | 84 |
| Journal | Ecological Processes |
| Volume | 15 |
| Issue number | 1 |
| DOIs | |
| State | Published - Dec 2026 |
| Externally published | Yes |
UN SDGs
This output contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 13 Climate Action
-
SDG 14 Life Below Water
Keywords
- Eddy covariance
- Environmental driver
- Machine learning-based modeling
- Mangrove ecosystem
- Net ecosystem production
Fingerprint
Dive into the research topics of 'Exceptional carbon sink potential of Shenzhen mangroves and a machine-learning framework for addressing sparse global observations'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver