Spatio-temporal variations of the spectral response in mangroves of Havana, Cuba, by remote sensing. Introduction: Mangroves are one of the most important forest types because of their ecosystem services and ecological roles. They represent 5 % of the emerged land of Cuba. No previous studies are describing spectral vegetation indexes variations by remote sensing in Cuban mangroves, but these variables can be used as indicators of the conservation status of the ecosystem and sustain national wide assessments. Objective: in the current paper we describe spatial and temporal variations in two spectral vegetation indexes in four near-city mangroves at Havana, Cuba and compare them to a natural control site. Methods: Study was conducted in localities named: Bajo de Santa Ana, Cojímar, Rincón de Guanabo, and El Cobre – Itabo lagoon and as control site was selected northern Zapata swamp. By using Climate Engine platform we extract, from 2 460 Landsat satellite images, mean values of spectral indexes NDVI and EVI in 5 parcels per locality, from 1984 to 2019. Variables were statistically compared among localities and the global trend was described. Results: We detect geographic differences in both indexes, which can be related to structural properties and conservation degree of mangroves in each locality. Global trend of indexes was to increase, but differently among localities. Slighter changes appear in the control site and, among near city mangroves, in Rincon de Guanabo and Cojímar. The ordering of localities from spectral variables was consistent with the ranking in general conservation degree. Conclusions: Spectral responses describe uniquely each mangrove forest, in concordance to each ecological and conservation characteristics. There is a need for promoting studies using remote sensors at these forest types and to generate strong and reliable indicators that can sustain future researches and monitoring schemes in Caribbean mangroves.

Keywords: coastal wetlands, NDVI, EVI, ecological indicators, historical trends