Abstract
Mangrove restoration is a vital nature-based solution for climate change mitigation and biodiversity conservation. However, large-scale mangrove restoration success is predominantly evaluated using area-based metrics, which fail to capture the complex temporal trajectories of forest structural and functional recovery. Stand age directly determines structural heterogeneity and carbon stocks, making it a key indicator for monitoring recovery trajectories and estimating carbon stocks. We developed a spatial age-based framework to map established mangrove stands and quantify the existing carbon stock and remaining potential. To this end, we combined an optimized LandTrendr change-detection algorithm with age-stratified logistic growth models using Landsat time-series data. We generated a 30-m resolution age map of all established mangroves from 2000 to 2023 across tropical Hainan Island, China. Our results identified 337 ha of established mangroves as of 2023. Notably, >50% (173 ha) of the detected stands are ⩽7 years old yet have already attained 43% of mature forest carbon stocks, whereas only 4% of them have reached ⩾15 years. Under continued protection, these forests are expected to store an estimated 33 668 Mg C (27 971–38 902 Mg C), with the greatest potential concentrated in the Dongzhaigang area. Conversely, the limited extent of older forests (4% of the established area) already secures significant recovered carbon stocks, warranting urgent protection. This contrast underscores the necessity of differentiated management strategies calibrated to both restoration scale and stand development stage. Ultimately, this spatially explicit framework informs evidence-based management to improve restoration outcomes and provides actionable baselines for national carbon credit markets.
Funding Information
  • Fundamental Research Funds for the Central Universities (QUTPRA)
  • National Natural Science Foundation of China (72394404)
  • National Key R&D Program of China grant (2022YFF0803100)
  • QUT Postgraduate Research Award Scholarship
  • MEL Internal Program (MELRI2501)

This publication has 78 references indexed in Scilit: