A Review of Integrating Geospatial Technologies and Artificial Intelligence for Habitat Connectivity Assessment in Fragmented Landscapes
Abstract
Breaking down natural habitats presents a major threat to biodiversity through disrupted ecological connections, mutated species paths, and enhanced separation of territorial areas. Most of these difficulties require advanced analytical tools that combine geographic information technology with Artificial intelligence (AI) capabilities to do better habitat connectivity assessments. This review investigates the modern development of remote sensing technologies alongside Geographic Information Systems (GIS) and AI-powered models for habitat fragmentation assessment and conservation planning designs. Monitoring habitat transformations through time becomes possible through high-resolution LiDAR-based satellite imagery and using Unmanned Aerial Vehicles-based (UAV-based) monitoring, which delivers extensive spatial data about habitat alterations. The analysis of ecological corridors and the assessment of fragmentation metrics become possible by implementing two GIS-based modeling techniques, consisting of least-cost path analysis and circuit theory modeling. Deep learning frameworks, including Convolutional Neural Networks (CNNs) and Object-Based Image Analysis (OBIA), have revolutionized land cover classification while simultaneously enabling automated connectivity assessments by enhancing accuracy levels. The application of AI in ecological assessments is hindered by challenges such as limited data availability, poor model generalization, and a lack of interpretability in predictive models. The study followed a systematic review approach to formulate the review article to address the research focus. The review research has been designed to achieve three objectives: to evaluate the integration of AI techniques for improving data access at fine scales, enhancing ecological connectivity metrics for sustainable habitat management, and to examine the challenges and opportunities of using AI and GIS in habitat connectivity, focusing on data access, model interpretability, and classification consistency.
