Weed-related yield losses are challenging to measure and predict but are estimated to be up to 34% in temperate environments and over 50% in tropical regions. Weed impact depends on their abundance but also on their trophic resource acquisition strategies. This study aims to: (i) assess the response of sugarcane yield to different weed communities shaped by three weed-management strategies, (ii) classify weed species as a function resource competition strategy (defined from their functional traits), and (iii) link these strategies to their effects on sugarcane yields. We performed two controlled sugarcane field trials in La Réunion island from 2021 to 2023. In these trials, we artificially created different weed communities using weed control interventions (chemical and manual), and then measured the sugarcane growth obtained on these plots. We also measured various aboveground and root functional traits linked to competition for light and soil resources on major weed species. Linear mixed models were used to assess the effects of treatments on biomass and cane yield components, and a Principal Component Analysis followed by hierarchical classification were used to group species according to their traits. Finally, an RLQ and fourth-corner analysis explored the links between weed traits, species and sugarcane yield components. Sugar cane growth was significantly influenced by the weed communities in both trials. The impact of weed cover was clearly visible, but some differences could be attributed to the weed resource acquisition strategies. Creepers were mostly associated with acquisitive aboveground strategies and conservative belowground ones. Some dicotyledonous weeds were more likely to be associated with conservative aboveground strategies level and acquisitive belowground ones. However, these strategies did not have the same impact on the sugarcane growth depending environmental conditions. Under optimal fertilization and irrigation, sugarcane growth was mainl