From Mexico to Argentina: Predicting the distribution of the invasive weed Tithonia tubaeformis beyond its native range using niche modelling

Abstract

1. Biological invasions are a major driver of global change, affecting biodiversity, ecosystem functioning and agricultural productivity. Tithonia tubaeformis (Asteraceae) is a weed native to Mexico that has become invasive in several regions worldwide, including Argentina, where it threatens crop systems. Here, we compared species distribution models trained with native- and invaded-range occurrences to estimate the potential distribution of T. tubaeformis in Argentina. We also developed an invasion risk map to support prevention and management strategies.2. Environmental predictors included 19 bioclimatic variables, elevation, and a Human Influence Index (HII). Models were calibrated using native- and invaded-range occurrences with MaxEnt, testing multiple regularisation multipliers and feature class combinations, with and without HII. The best-performing models were projected onto Argentina to estimate potential distribution and assess model transferability and environmental niche overlap between ranges.3. Models showed good performance, and including HII improved model accuracy and explained a large portion of the predicted distribution. The native-trained model projected a broader potential range than the invasive-trained model, suggesting that T. tubaeformis has not yet occupied all suitable habitats in Argentina. Most invasive occurrences fell within the environmental space predicted using the native-trained model, although some records indicated a slight niche expansion under previously unoccupied conditions. The risk map identified northern Argentina as highly suitable for invasion, highlighting priority areas for monitoring and prevention.4. Implications. Our findings demonstrate that native-range trained models can anticipate the spread of emerging weeds and that incorporating anthropogenic factors enhances predictive performance. These results contribute to understanding invasion processes and provide a practical tool for risk assessment and management planning in human-modified landscapes.