Composição florístico-estrutural e imagens RGB obtidas por drone na avaliação de florestas ripárias urbanas

Abstract

Riparian forests play a crucial role in maintaining biodiversity and ecosystem services, especially in regions of the Brazilian Cerrado. However, these areas face anthropogenic pressures that compromise their structure and function. This study compared the quality of riparian forests in urban and rural areas of the city of Ipameri-GO, using phytosociological data and remote sensing techniques with drones, aiming to evaluate the potential of these technologies for environmental monitoring. As a methodology, 23 plots were sampled in 6 urban streams and 13 plots in conserved rural areas. Parameters such as richness, density, basal area, and floristic diversity were analyzed. Aerial images captured by drones were processed using Agisoft Metashape and QGIS software, using vegetation indices (ExG, TGI) to quantify vegetation cover. Statistical analyses included non-parametric tests, similarity dendrograms, and linear regressions. The results showed that urban riparian forests presented lower richness, density, and basal area compared to rural areas, in addition to the significant presence of exotic species such as *Mangifera indica* and *Leucaena leucocephala*, among others. Floristic similarity analysis revealed two distinct groups: one formed by degraded urban plots and the other by conserved rural areas. Vegetation indices, especially ExG, demonstrated effectiveness in quantifying canopy cover, with significant correlations with phytosociological parameters, although limitations were observed in areas with grasses and invasive shrub species. The degradation of urban riparian forests is associated with fragmentation, the introduction of exotic species, and the reduction of ecological connectivity. The presence of invasive species such as *Leucaena leucocephala* exacerbates the loss of native biodiversity. The use of drones and vegetation indices proved promising for monitoring, but RGB technology has limitations in heterogeneous environments. Therefore, this study showed that most of the streams in IpameriGO are highly anthropized, requiring restoration and adaptive management strategies, including the replacement of invasive species with native ones. The effectiveness of high-resolution remote sensing was proven as a complementary tool to assess the quality of riparian forests, highlighting the urgency of public policies for the preservation and recovery of these ecosystems in urban areas.