Morphometric and soil chemical properties govern spatiotemporal variations of porewater carbon in coastal mangroves

Abstract

Understanding the spatiotemporal variations in porewater carbon (C) exchange in mangroves is essential for uncovering their roles in lateral C transfer to marine ecosystems and deepening our understanding of global C sequestration. This importance is particularly evident in coastal regions like the Pearl River Delta, where diverse mangroves face increasing anthropogenic pressures. In this study, we investigated how porewater dissolved inorganic C (pDIC) and dissolved organic C (pDOC) varied and interacted in mangroves of the Pearl River Delta, focusing on the key influencing components, including tidal hydrology, water chemistry, soil properties, and morphometric characteristics in shaping C dynamics. Our results revealed a strong consistency (r = 0.870) between the spatiotemporal variability of pDIC and pDOC, exhibiting significant negative trends related to tidal flow hours (pDIC: R² = 0.918, pDOC: R² = 0.905). Among the influencing components, soil properties had the most substantial influence on pDOC concentrations, accounting for 39 % of its variance, while morphometric characteristics and pDOC concentration had similarly high impacts on pDIC, each accounting for approximately 30 % of its variance. Specifically, soil bulk density (SBD) and soil pH (SpH) contributed negatively, whereas soluble dissolved organic C (SDOC), soil total C (STC), and soil total nitrogen (STN) contributed positively to the overall soil property composition. Within the morphometric component, roundness and perimeter made the strongest positive contributions. Furthermore, this study also demonstrated the feasibility of utilizing several key metrics—especially easily accessible ones like morphometric characteristics and tidal flow hours—to simulate pDIC and pDOC variability in mangroves. This approach offers a straightforward approach for evaluating C exchange dynamics in coastal mangroves and has significant implications for large-scale C flux estimation.