Resilience of mangroves to climate and land-use changes in the Pacific Islands

• We model the ecogeomorphological response of a mangrove wetland in Fiji. • We use remote sensing data to set up the model and validate results. • We analyse scenarios of climate and catchment land use changes in the catchment. • All scenarios predict considerable wetland loss for high sea-level rise conditions. • Resilience is limited by natural topography and manmade tidal flow restrictions. Mangrove wetlands in Pacific Islands are vital ecosystems that provide important services such as habitat for marine species, flood protection and carbon storage. They occupy low-lying areas that are at risk of sea level rise (SLR), which, combined with anthropogenic pressures exerted on adjacent zones can lead to destabilising effects on these ecosystems. Their resilience to climate and land use changes is closely related to sediment availability, as mangroves can vertically adjust their soil surface (accretion) by trapping sediment and building root mass, offsetting SLR. In this contribution, an ecogeomorphological model is applied to a mangrove wetland site in Vanua Levu Island, Fiji, to predict its evolution over the next 100 years and to assess its resilience under five different scenarios of climate and land use change. Scenarios consider the SSP5-8.5 pathway SLR and include a scenario with current conditions, two scenarios with different levels of deforestation in the catchment, one scenario with increases in temperature and rainfall intensity, and a scenario with conservation practices to reduce sediment from the catchment. Changes in the catchment are assessed using a hydro-sedimentological model previously calibrated, and produce increases in sediment supply to the wetland for all scenarios except the conservation scenario. Changes in sediment supply are incorporated into the ecogeomorphological model, which is able to quantify improvements in the resilience of the wetland due to increases in sediment. Improvements in resilience are not enough to prevent substantial wetland losses (30 % to 60 %) because mangroves cannot colonise new areas due to topographic constraints and a manmade embankment that prevents tidal incursion. The methodology uses numerical models that are set up and verified with regional information and remote sensing derived data, so it has enormous potential for the assessment of wetland vulnerability in other mangrove wetlands of the world with limited ground information.