A ERSF-VIPA framework: scalable wildlife movement modelling for conflict mitigation.

1. Effective conservation planning and conflict mitigation can hinge on accurately modelling wildlife movement paths (WMPs), yet progress is hindered by both a shortage of reliable methods and limited data. The critical challenge, therefore, is to devise limited-data models that faithfully reproduce elusive species' movements and deliver actionable insights for human-wildlife conflict management. 2. We introduce the Enhanced Resource Selection Function-Vector-network Iterative Pathfinding Algorithm (ERSF-VIPA), a novel framework for simulating WMPs with limited data. Drawing on historical occurrence records of Asian elephants (Elephas maximus), we assume individuals make rational, goal-driven decisions based on local environmental knowledge. The ERSF employs a random forest on a hexagonal grid to estimate nonlinear resource-selection probabilities, while VIPA conducts an iterative, node-to-node search across that hexagonal vector network-scoring each candidate by combining selection probability with cubic distance coefficients to ensure ecological validity and energetic efficiency. 3. The model demonstrates high accuracy, with 90.3% of the 68 simulated paths approximating the observed paths with an average maximum deviation of 418 m. These findings underscore the model's robustness and its capacity to translate limited tracking data into actionable insights for conservation. 4. ERSF-VIPA operates using only coarse, non-continuous historical data that lack precise timestamps or spatial accuracy. By operating with minimal data requirements, ERSF-VIPA demonstrates exceptional extensibility and broad applicability for reconstructing movement paths of elusive wildlife species. Its proven accuracy in simulating Asian elephant paths further positions it as a potentially powerful decision-support framework for real-time animal monitoring and proactive human-wildlife conflict mitigation.