Data-driven framework for eco-sensitive planning and design: a game theory-enhanced LUCIS approach to human–elephant conflict in Sri Lanka

Abstract

Human–elephant conflict (HEC) represents one of the most pressing socio-ecological challenges in Sri Lanka, where expanding agriculture and human settlements increasingly overlap with elephant habitats. Conventional mitigation strategies—such as electric fencing, translocation, and compensation schemes—have largely remained reactive and sectoral, addressing symptoms rather than underlying spatial drivers of conflict. This study proposes a data-driven framework for eco-sensitive regional planning and design that integrates urban informatics, the Land Use Conflict Identification Strategy (LUCIS) and game theory to proactively manage human–elephant interactions through spatial planning. Using a 7, 500 km2 landscape in Sri Lanka's Dry Zone as a case study, multi-criteria suitability models were developed for elephant habitat conservation, agriculture, and human settlements using geospatial datasets and stakeholder-informed criteria. Spatial overlay analysis revealed that 27.6% of the landscape exhibits high suitability overlap, indicating structural competition among land uses. Spatial correspondence analysis using 1, 754 recorded HEC incidents showed that 79.8% of conflicts occurred within these overlap zones, confirming the spatial drivers of conflict. To address this, a game-theoretic negotiation model was applied to reallocate contested grid cells while balancing stakeholder utilities. Under modeled conditions, the resulting allocation scenario reduced the mean Conflict Index from 0.32 to 0.10 and decreased high suitability overlap areas by 70%, while increasing elephant habitat utility by 15% and maintaining agricultural and settlement capacity. These results represent scenario-based outcomes derived from the model and should be interpreted as indicative spatial planning potentials rather than empirically validated reductions in conflict. The resulting land-use structure delineates Elephant Conservation Zones, Agricultural Development Zones, Settlement Zones, Buffer Zones, and Coexistence Zones, creating a conflict-optimized landscape configuration. By integrating geospatial AI, participatory decision tools, and simulation-based planning, the framework demonstrates how urban informatics can transform reactive wildlife conflict management into proactive spatial governance. The methodology offers a scalable and transferable planning support system for managing human–wildlife conflicts in rapidly transforming socio-ecological landscapes. By integrating geospatial data analytics, participatory decision-support tools, and spatial simulation, the framework demonstrates how urban informatics can support sustainable development by enabling data-driven planning and design decisions that foster resilient landscapes and promote sustainable coexistence between human settlements and ecological systems.