Characterizing and Mapping the Grassland Vegetation of the Colombian Orinoquia

The current distribution of grassland vegetation in the Colombian Orinoquia is presented and discussed through the application of machine learning techniques to multidimen-sional imagery and an empirical-statistical modeling framework. This model was devel-oped by integrating 281 vegetation surveys grouped into 18 phytosociological alliances, environmental complexes, a geomorphological gradient, bioclimatic factors, and a multi-sensor image composed of Landsat-8 bands from the visible and infrared portions of the electromagnetic spectrum, along with dual-polarization Sentinel-1 bands. Approximately 73.74% of the region still retains some form of natural vegetation, with 32.19% corre-sponding to forests and 41.55% to natural grasslands. The most prevalent alliances are Paspalo pectinati–Axonopodion aurei and Axonopodo aurei–Trachypogonion spicati, covering 6.02% and 4.37% of the area, respectively. Additionally, the ecogeographical distribution of the syntaxa and the dominance of C4 Poaceae species are discussed. Notably, Rhynchosporo barbatae–Andropogonion virgati and Sipaneo pratensis–Axonopodion purpusi stand out due to the marked dominance of C4 Poaceae among their characteristic species (50% and 60%, respectively). These findings underscore the relevance of integrating environmental complexes with bioclimatic, floristic, and physiological attributes to enhance the spatial interpretation of ecogeographical vegetation patterns at a regional scale.