The development of alternative management practices with less reliance on herbicides is a critical challenge for sugarcane production in La Réunion Island. Long-term effects of management practices on seedbank dynamics and weed communities must be considered. However, field experiments to address these questions are time-consuming and expensive, delaying actionable insights. The FLORSYS model, initially designed for arable cropping systems in temperate regions, is a promising tool for overcoming this deadlock. The aim of this study is to assess the performance of the FLORSYS model under tropical conditions and its ability to account for the characteristics of semi-perennial sugarcane cropping systems and tropical weeds. So far, we have parameterized one sugarcane variety and 15 dominant tropical weed species using a combination of previously measured functional traits, existing literature, and expert knowledge. To refine key parameters, we set up dedicated experiments to characterize the early growth phase of sugarcane and to assess its morphological plasticity under light competition, a critical factor in weed-crop interactions. The model, along with our parameterization, is currently being evaluated against historical experimental datasets to test its ability to accurately simulate weed competition effects on sugarcane yield. In preparation for the congress, we compare different weed parameterization approaches and their impact on yield prediction: (i) using average values to represent the weed flora without species discrimination, (ii) clustering weeds into broad functional types (e.g., tall grasses, small dicotyledons, perennial weeds), and (iii) explicitly parameterizing individual weed species to reflect weed-flora diversity. This work will provide a sugarcane–weed model parameterized and evaluated for La Réunion Island. In the future, the model will be used to assess various weed management strategies in relation to the objectives and constraints of different