Résumé : Oil palm cultivation is continuously expanding in Colombia due to its increased demand worldwide, making Colombia the 4th largest producer of palm oil globally. Most oil palm plantations are located in the eastern plains (Colombian llanos), a highly biodiverse region characterized by savannahs, wetlands and high avian richness. Oil palm plantations may threaten biodiversity due to resulting loss and fragmentation of from land conversion to crops. In this study we used Generalized Linear Models and linear models to understand the impacts of patch and landscape characteristics and structure on bird species richness and functional richness, divergence, and evenness. We investigated scale differences in our responses as well as the associations between certain landscape types and different bird trophic levels. Our results show nearest neighbor distances and water cover are positively associated with both species and functional richness. Further, we found diverse levels of oil palm cover to maximize species richness, large patch areas to maximize functional richness and connectivity between patches to shape the functional space of the avian assemblages that inhabit it. Our findings suggest that certain levels of oil palm cover may be contributing to increased biodiversity but large forest patches and optimal connectivity levels between these are still essential to maintain the functionality of the ecosystem. Our results could inform decisions on palm oil cultivation and expansion that guarantee ecosystem health, biodiversity and functionality.