An Integrated Decision Support Framework for Sustainable Road Planning in Biodiversity Hotspots

Global infrastructure investment will reach $94 trillion by 2040, posing critical threats to biodiversity in rural regions newly exposed to road access. Traditional planning approaches have failed to sufficiently incorporate stakeholder preferences and account for uncertainty, leading to decisions that often compromise ecosystems. To overcome these limitations, we introduce an integrated decision-support framework that fuses Multi-Criteria Decision Analysis (MCDA) with Bayesian Belief Networks (BBN), allowing explicit incorporation of stakeholder-derived preferences and probabilistic modeling of ecological and socioeconomic outcomes. Applying this framework to four proposed roads in Tanzania's Greater Serengeti Ecosystem, we show that conservation-oriented options (Mbulu and Eyasi) offer 15–25% better sustainability performance than conventional connectivity-maximizing routes, while serving 70% of the population with 291% higher per-capita efficiency. Thirty-year projections confirm the long-term benefits of ecosystem-sensitive designs, and cross-validation between MCDA and BBN confirms methodological robustness (r, 0.89). This study provides a scalable, evidence-based planning framework for sustainable road development.