Development of an Original Spatiotemporal Risk Model for Highly Pathogenic Avian Influenza Spread in Nigerian Live Bird Markets

This dissertation develops and validates an original spatiotemporal risk model for the spread of highly pathogenic avian influenza (HPAI) through Nigeria's live bird market (LBM) network, providing a theoretically novel and empirically grounded tool for targeted HPAI prevention and response. Live bird markets are the principal amplification and dissemination points for HPAI in Nigeria, but no spatially explicit, network-informed predictive model for HPAI spread through Nigerian LBMs exists, creating a critical gap in outbreak risk management. The dissertation employs a three-phase design: systematic characterisation of LBM network structure through ethnographic mapping, trader interview surveys, and GPS tracking of bird movement across 80 markets in the South West, South South, and North West geopolitical zones; cross-sectional virological surveillance at 40 markets over four seasonal periods; and spatiotemporal model development integrating network connectivity, market throughput, biosecurity scores, and environmental suitability. Negative binomial regression, network diffusion simulation using NetLogo, and Bayesian hierarchical spatial models are the primary analytical tools. The dissertation proposes the Nigerian LBM HPAI Spread Risk Index (NLHSRI) as an original modelling contribution, classifying 80 markets into four risk tiers with explicit epidemiological justification. Validation against 2022 and 2023 outbreak records shows 79 percent accuracy in predicting tier 1 markets as first-affected sites. Policy applications include a risk-tiered surveillance intensification protocol and a network-based ring vaccination strategy for the highest-centrality market nodes. Keywords: avian influenza, live bird markets, spatiotemporal model, network analysis, Nigeria.