Monitoring impact of recent Sargassum influxes on mangrove cover in the coastal bays of Bonaire, Caribbean Netherlands

Samenvatting

A change detection analysis utilizing Very High-Resolution (VHR) satellite imagery was performed to evaluate the changes in mangrove cover in two coastal bays (i.e. Lagun and Lac Bay) on Bonaire, attributable to recent coastal influxes of holopelagic Sargassum brown algae. A baseline Pleiades satellite image from 2014 (i.e. before the first Sargassum influx on Bonaire) and a Pleiades satellite image from 2020 were co-registered, pan-sharpened at a resolution of 50 cm to enable detection of small changes. Next, changes in the Normalized Difference Vegetation Index (NDVI) were used to quantify changes in mangrove cover between 2014 and 2020. In addition, Sentinel-2 (S2) satellite imagery in combination with a previously developed Random Forest model to detect floating Sargassum was used to map the mangrove areas within Lac Bay and Lagun that experienced direct impact of Sargassum influxes between 2017 and 2020. Results showed that in 2014, the total area coverage of mangroves in Lac Bay was 221 ha, of which 16 ha (7.2%) was lost in the period between 2014 and 2020. However, based on the maps that showed where Sargassum accumulated within Lac Bay, only 0.6 ha (3.7%) of the total mangrove area that was lost between 2014 and 2020 could most likely be attributed to the direct impact of Sargassum influxes (i.e. the outer edge of the mangrove forest in the West of Lac Bay), while the remaining mangrove loss of 15.4 ha (96.3%) could most likely be attributed to run-off related sedimentation in the backwaters of Lac Bay. The total area coverage of mangroves in Lagun was 2.6 ha in 2014, of which 1.2 ha (46.2%) has been lost in the period between 2014 and 2020. Of the mangrove area that was lost in Lagun, 0.4 ha (33.3%) could most likely be attributed to the direct impact of Sargassum influx, while the remaining mangrove loss of 0.8 ha (66.7%) was most likely the result of run-off related sedimentation. The approach presented in this study provides a quantifiable method to evaluate Sargassum impacts to mangrove vegetation using change detection of VHR satellite imagery and showed that, although limited, Sargassum influxes seem to have caused mangrove die-offs in specific areas within Lac Bay and Lagun. Obtained findings will help identify priority areas within the mangrove forests of Lac Bay and Lagoon where management of Sargassum influxes and run-off-induced sedimentation may be required.