Estimativa do estoque de carbono em ecossistema de manguezal por sensoriamento remoto.

Abstract

Mangrove forests make up the coastal landscape of tropical, subtropical and temperate regions. This ecosystem offers a wide range of services, including carbon sequestration and storage, which, in the face of climate change and global warming, have global relevance. Therefore, the objective was to estimate carbon stocks, in biomass and soil, in a mangrove ecosystem, from the structural analysis of vegetation, species variability, abiotic characteristics of sediments, in the Gramame River estuary, located on the South Coast. from the state of Paraíba, Brazil. To this end, a literature review was carried out covering 113 scientific articles, of which 21 studied Brazilian mangrove areas. Subsequently, carbon stocks in the biomass were measured using a forest-phytosociology inventory. Soil carbon was measured through composite samples, collected in the first meter of depth, and the soil properties analyzed, such as density, organic matter, nutrients, granulometry, pH and salinity. As for the estimates by remote sensing, the data collected in the field measurement was used and, in the sequence, the Google Earth Engine (GEE) cloud processing platform was used to calculate the NDVI, SAVI and EVI. To define the fit curves, linear and non-linear regressions were used. The choice of model occurred considering the highest coefficients of determination (R²). The global stock estimated by the literature varies between 4.19 Pg C and 6.6 Pg C, so that the average value calculated corresponds to 5.34 Pg C, which is present in specific regions, such as Indonesia, Brazil and Australia. The values recorded in Brazilian mangroves vary between 358.79 Mg C/ha and 1,851 Mg C/ha, so that the average stock corresponds to 709.36 Mg C/ha. Analogous to what happens on a global scale, it can be seen that the soil concentrates greater amounts of carbon when compared to biomass. The carbon stocks in biomass and soil, determined from field measurements, corresponded to 44.89 ± 7.41 Mg C/ha and 813.08 ± 20.87 Mg C/ha in the Gramame river mangrove, from so that the total stock in the ecosystem was 857.98 ± 20.57 Mg C/ha, equivalent to 1,562.27 Gg of CO2. It was found that soil and biomass account for 94.77% and 5.23%, respectively. When analyzing the distribution of carbon stock, considering the variability of the species, it was noticed that Rhizophora mangle was the dominant species, a fact that contributed to the formation of larger stocks (323 Mg C/ha – 65.41%), followed by Laguncularia racemosa (119.33 Mg C/ha – 24.16%) and Avicennia schaueriana (51.51 Mg C/ha – 10.43%). The total stocks estimated by remote sensing, for vegetation carbon, corresponded to 26.66Gg C and 27.76 Gg C, for Landsat 8 and Sentinel 2B. These values are above the field value by 16.45% and 19.78%, respectively. The differences were related to the structural heterogeneity of the vegetation, the spatial resolution of the products and the manipulations performed to adapt the data, especially for Sentinel 2B, given the incompatibility between plot size and pixel dimension. This study showed that optical products from Landsat 8 and Sentinel 2B allow the successful estimation of aboveground biomass and carbon stock in a mangrove ecosystem.