Résumé : In recent years the value of mangroves as blue carbon sinks has been acknowledged in the framework of nature based solutions for climate change. Therefore, it is increasingly important to understand the entity of the total carbon stocks in mangroves, by considering all major components of the ecosystem. This includes the study of vegetation biomass (in its above and belowground components), dead wood quantification and leaf litter production as well as the soil carbon pool. The soil in mangroves is known to contain the largest carbon pool with up to 90% of the ecosystem carbon stock, however soil sampling in previous research has always been limited in the depth variable with cores often only 1 m deep or less. Therefore, one of the aims of this PhD research was to investigate changes in soil carbon with soil depth in order to get a better understanding of the real carbon stock in mangroves by analysing for the first time soil cores up to 10 m depth.The present study was conducted in the Matang Mangrove Forest Reserve (MMFR) in peninsular Malaysia, which has been under silvicultural management for pole and charcoal production for over a century. The forest is composed mainly of Rhizophora apiculata trees and is managed with a thirty-year rotation cycle that includes two intermediate thinnings and a final clear-felling. One of the advantages of this study site is the known age of each forest compartment that allows us to interpret the changes in vegetation structure and carbon stock with forest age.We found that the MMFR had a considerable carbon stock, especially in the soil. However, the exploitation of mangroves for extraction of timber, notably releases carbon equivalents to the atmosphere, leading to the loss of important amounts of carbon, due to the extraction of the resources and to disturbance. In order to evaluate the losses of carbon from the pole and charcoal production, the emissions from all the activities involved in the silvicultural cycle were considered and duly compared with the carbon sequestration in a thirty-year cycle to assess the carbon budget of the forest. The results showed that the carbon budget was negative if considering only the forest zone used for wood extraction, but turned positive if the areas left for conservation (untouched forest) were considered. Therefore, the system can still be considered sustainable from a carbon point of view although the management should consider a reduction in the forest exploitation to obtain a more resilient mangrove forest and increase the sustainability of the silviculture.