Spectral indices and ecosystem health in mangrove monitoring: a systematic review of optical remote sensing approaches

Abstract

Optical remote sensing has become a key tool for monitoring mangrove ecosystems and assessing their spatial and temporal dynamics. However, many studies evaluate spectral indices primarily from an applied perspective, often overlooking the methodological evolution and ecological questions that originally motivated their development. This study presents an interpretive systematic review structured following the general PRISMA 2020 framework (n = 40 articles) to reconstruct the historical and methodological trajectory of spectral metrics used to characterize mangrove ecosystem health. The results reveal a recent predominance of Sentinel-2 imagery (55% of the reviewed studies) and a persistent reliance on the Normalized Difference Vegetation Index (NDVI), which appears in approximately 85% of the literature analyzed. The bibliographic synthesis identified four dominant methodological transitions, ranging from early land-cover detection approaches to automated time-series analyses supported by cloud-computing platforms, particularly Google Earth Engine (GEE). The reviewed evidence suggests that indices based exclusively on canopy greenness have limitations for interpreting degradation processes in intertidal environments, where hydrological changes may precede detectable shifts in vegetation cover. Multi-index strategies that combine moisture-sensitive metrics (e.g., NDMI or NDWI) with vegetation-vigor or background-corrected metrics (e.g., EVI or SAVI) may provide a more informative framework. Their performance remains context dependent and should be interpreted in relation to tidal regime, canopy structure, spatial and temporal resolution, and ecological field validation. Finally, the study identifies key methodological gaps related to empirical validation, uncertainty quantification, and the interpretation of spectral signals in complex coastal environments. Addressing these limitations may improve the reliability of remote-sensing-based monitoring. The application to Estero Pargo, Jaina, and Atasta in the Campeche State, Mexico, is presented as a methodological translation of patterns identified in the reviewed literature rather than as direct empirical validation of spectral-index performance at those sites.

Keywords: Remote Sensing, mangroves, spectral indices, sentinel-2, ecosystem health, coastal ecosystems