Accounting for physical barriers in insect pest modeling: A spatially explicit simulation-based approach

Understanding pest dynamics beyond greenhouse boundaries is critical for anticipating outbreaks and guiding sustainable management. Despite the central role of insect pest ecology in Integrated Pest Management, landscape-scale research on external greenhouse environments is limited. This knowledge gap constrains the development of spatially informed early-warning systems and green infrastructure to intercept pest movement. We introduce a spatially explicit simulation framework designed to model pest abundance in the peri-greenhouse landscape by integrating high-resolution data on greenhouse density, landscape structure, and resistance to pest dispersal. We used Barrier models to assess the comparative performance of cluster versus simple random sampling across varying spatial scenarios and sample sizes. Our results demonstrate that greenhouse spatial arrangement significantly mediates sampling efficiency. Cluster sampling consistently outperformed simple random sampling in scarcely and densely fragmented landscapes, reflecting its effectiveness in capturing strong spatial continuity. Crucially, in moderately dense landscapes, both methods showed comparable performance, suggesting that intermediate fragmentation disrupts the necessary spatial aggregation for cluster sampling's efficiency. These findings highlight the necessity of matching sampling design to spatial landscape features and pest management goals. The proposed framework is a customizable and scalable tool for simulating pest dynamics and optimizing field monitoring strategies. By bridging geostatistical modeling and ecological simulation, it provides a transferable workflow that advances landscape-scale ecological modeling and supports the design of adaptive pest management strategies. It strengthens the integration of ecological informatics into decision-making by enabling scenario testing under diverse spatial configurations, offering practical insights for spatially informed early-warning systems in agricultural landscapes. • A spatially explicit simulation framework models pest spread in greenhouse landscapes. • The model includes greenhouse layout and dispersal resistance at high resolution. • Two sampling strategies are evaluated under diverse pest and landscape scenarios. • Cluster sampling outperforms when pest aggregation and spatial barriers align. • The model informs scalable pest control planning across complex landscapes.