Résumé
Monitoring carbon budget is a major challenge to improve coral reefs management (protection, conservation or restoration), including predictions about their fates by 2100. It requires sensors deployment, a collection of seawater samples and consideration of local factors such as tide, wind and benthic cover. This study assessed the benthic metabolic processes (organic production, respiration, calcification and calcium carbonate -CaCO3- dissolution) at three sites on the main shallow fringing reef of La Reunion Island. Autonomous sensors (current speed, temperature, salinity, PAR, dissolved oxygen -DO- and pH) were deployed during the warm and cold seasons between 2020 and 2022 and were complemented by discrete sampling of seawater for DO, pH and total alkalinity (TA) analysis. Metabolic budgets were calculated using a one-dimensional Eulerian approach of flow respirometry (study of changes in seawater chemistry along an axis, between two points, using fixed moored sensors). Due to the reef geomorphology, the variation in current orientation and its unidirectionality, spatio-temporal variability of metabolic budgets could not be related to benthic cover (Pleiades satellite images). Those factors strongly influenced variations of seawater chemistry and our ability to estimate the seawater residence time. We will show here, for instance, that reef geomorphology creates conditions for a flow re-entrainment of TA-depleted seawater (due to reef calcification activity) exiting the reef. Such TA-depleted seawater is immediately redirected towards the reef flat leading to an overestimation of daily reef net calcification rates (Gnet). At one site (33% of coral cover), Gnet was estimated at 283 mmolCaCO3.m-2.d-1. Such rate characterizes reefs with a coral cover up to 60%. Gnet accounted for 40% of reef gross primary production GPP (GPP = 726 mmolC.m-2.d-1, similar to the Indo-Pacific median) while calcification usually corresponds to 15-20% of GPP for typical coral/algal reef flats. The Eulerian method that was used in our study is a punctual approach and can be a powerful tool for monitoring coral reefs metabolism, mainly through the use of autonomous sensors. However, accurate estimations on a large spatio-temporal scale need long-term surveys and may be limited due by the difficulty of integrating the specific hydrodynamic conditions of each coral reef studied.