Network Systems Science and Advanced Computing | Resilient Societies and Interdependent Infrastructures

Simulation Tools for Pest Risk Analysis Accounting for Ecological and Anthropogenic Drivers (SPREAD)

Sponsor

USAID

Globalization has broken down geographical barriers to the movement of humans and their goods. This has caused unprecedented disruptions to native ecosystems by facilitating the spread of invasive alien species. In the context of an agricultural pest, the problem is further exacerbated because: the invaded region is usually devoid of its natural predators, and human-mediated pathways such as trade and travel aid in its rapid dispersal. It also leads to responses such as increased use of broad spectrum insecticides, insecticide resistance and trade restrictions.

In our work we apply novel AI methods to study a representative agro-pest: the South American tomato leafminer (Tuta absoluta).

Project Overview

This project develops computational models and simulation environments to study the spread and establishment of pests that affect major agricultural crops. Our approach not only considers the biological interactions of these organisms, but more importantly, the social and economic factors. It strives to provide a valuable tool for risk analysts, domain experts and policy makers for testing counterfactual scenarios, assessing and forecasting the extent of the damage, and designing targeted monitoring and control measures to mitigate the spread of leafminers. This project provides a unique opportunity to apply big data techniques, network science, simulation science, and machine learning techniques to a high-impact application.

Findings

We developed a framework to study the spatio-temporal spread of invasive species as a multi-scale propagation process. We account for climate, biology, seasonal production, trade and demographic information. Machine learning techniques were used in a novel manner to capture model variability and analyze parameter sensitivity.

We applied the framework to understand the spread of Tuta absoluta, in South and South-east Asia, a region at the frontier of the pest’s current range.

Analysis with respect to historical invasion records suggests that the pest can quickly expand its range through domestic city-to-city vegetable trade. Our models forecast that—without mitigation—within five to seven years Tuta absoluta will invade all major vegetable-growing areas of mainland Southeast Asia. Monitoring high-consumption areas can help in early detection, and targeted interventions at major production areas could effectively reduce the rate of spread. We also studied the long-term establishment potential of the pest and its economic impact on Nepal, which could range from $17–25 million in USD.

Team

Research Assistant Professor

Division Director

Distinguished Professor in Biocomplexity, Biocomplexity Institute

Professor of Computer Science, School of Engineering and Applied Science

Professor

Professor of Public Health Sciences, School of Medicine

Professor

Professor of Public Health Sciences, School of Medicine

Research Assistant Professor

Publications
Network Systems Science and Advanced Computing
Venkatramanan S; Wu S; Shi B; Marathe A; Marathe M; Eubank S; Sah L; Giri A; Colavito L; Nitin K . Crop Protection. Elsevier. 2020; 135:104736
Network Systems Science and Advanced Computing
McNitt J; Baek H; Mortveit H; Marathe M; Campos M; Desneux N; Brévault T; Muniappan R; Adiga Abhijin . Proceedings of the Royal Society B. The Royal Society. 2019; 286(1913):20191159
Network Systems Science and Advanced Computing
Nath M; Venkatramanan S; Kaperick B; Eubank S; Marathe M; Marathe A; Adiga Abhijin . International Conference on Complex Networks and their Applications. Springer, Cham. 2018; :524-535
Network Systems Science and Advanced Computing
Campos MR; Biondi A; Adiga A; Guedes RNC; Desneux N . Journal of Pest Science. 2017; 90(3):787-796
Network Systems Science and Advanced Computing
Venkatramanan S; Wu S; Shi B; Marathe A; Marathe M; Eubank S; Sridhar V . 2017 IEEE International Conference on Big Data (Big Data). 2017; :435-444