Performing Department
(N/A)
Non Technical Summary
Pesticide resistance remains a major concern for the global food supply, as pests cause considerable damage to crops if they are not managed effectively. Current resistance management models continue to fail to control or delay pesticide resistance. New precision tools are needed to improve predictions about the emergence of pesticide resistance phenotypes across entire growing regions. Specifically, predictive statistical tools are needed to model how pesticide resistance develops within and among pest populations, to predict how resistance can spread geographically, and to leverage ongoing monitoring data to refine management practices. The purpose of this project is to develop statistical tools and predictions of pesticide resistance evolution that can be used to generate a landscape level "risk-map" of resistance trait variation. This would benefit producers by targeting management towards key populations that have a high likelihood of resistance evolution. This project leverages a unique spatiotemporal dataset, comprising abundance, genomic, and environmental data, to make these predictions for the Colorado potato beetle, Leptinotarsa decemlineata. The results would directly address USDA program priorities in the area of "Pests and Beneficial Species in Agricultural Production Systems"to use molecular genetics to predict and manage insect pest outbreaks, as well as to investigate mechanisms of pest resistance to pesticides. Additionally, the project addresses the 2018 Farm Bill priorities to develop advanced technologies and improve crop health and production globally by reducing insect damage. The impact of this work provides a means to mitigate resistance, ensure food security, and increase agricultural sustainability.
Animal Health Component
40%
Research Effort Categories
Basic
40%
Applied
40%
Developmental
20%
Goals / Objectives
The long-term goal of this project is to build precision tools that can be used nationally to increase capacity to sustainably manage agricultural pests and ensure food security. Precision agriculture harnesses large datasets to direct management efforts for improved crop production.A major advance would be to predict the emergence of pesticide resistance phenotypes across the landscape of farms that encompass entire growing regions. Specifically, predictive statistical tools are needed to precisely model how pesticide resistance develops within and among field populations, predict how resistance can spread geographically, and to leverage ongoing monitoring data to update and refine management practices.We propose to develop a novel spatiotemporal statistical modeling approach that can work effectively at a landscape scale of management to deliver precision pest management.We will focus on the Colorado potato beetle,Leptinotarsa decemlineata, and the evolution of neonicotinoid resistance phenotypes in Wisconsin. We will leveragedatasets encompassinggenomic variation, population abundance, and associated phenotypic data on insecticide resistance, sampled in time and space. Our specific objectives are to: 1) conduct spatiotemporal analyses by building statistical tools for pest abundance and genomic data, 2) integrate these tools in a hierarchical Bayesian modeling framework to identify features that predict resistance phenotypes in spatiotemporal datasets, and 3) use these features in demographic simulations to forecast spatial "risk-maps" of resistance evolution, while updating and refining risk maps using additional temporal field sampling during the course of the project.
Project Methods
Objective 1: Develop spatiotemporal statistical tools to analyze pest abundance and genomic data.Effort: We will develop a Bayesian model framework to examine how spatiotemporal abundance and environmental data predicts the change in observed insecticide resistance phenotypes through time.The model willexamine the risk factors associated with insecticide resistance, which could include covariates such as temporally sampled population abundance,environmental data (climatic variation, acreage, and potato production intensity), and phenotypic data on neonicotinoid insecticide resistance. We will also developtemporal genomic data to identify the genetic basis of polygenic resistance evolution. Genomic data will be sampled from at least 16 populations repeatedly sampled in the years 2015, 2019, and 2023. We will use this data to measure allele frequency change across time. Gene regions that significantly exceed background levels of change will be identified as undergoingselection. Furthermore, we will test forpolygenic selection by examining the covariance of allele frequency change across the genome.Evaluation:We will evaluatethe relative predictive ability of our modeling approach to predict insecticide resistance in CPB over space and time. The developed statistical framework will be assessed for how it can be leveraged to improve precision management of pesticide resistance in Wisconsin. The temporal analysis of genomic data will be measured for its power to identify a set of causative genes involved in insecticide resistance in CPB. This approach will be assessed for its ability to identify novel pathways and processes (such as polygenic selection or gene flow) that contribute to pesticide resistance evolution.Objective 2: Integrate these tools in a single hierarchical Bayesian modeling framework to identify features that predict pesticide resistance phenotypes.Effort:To predict insecticide resistance phenotypes, we will extend our Bayesian hierarchical model to include the genomic data, as well as the abundance and environmental data, as different levels within a hierarchical model. The genomic data contains information on population genetic structure and gene flow, in addition to the causative alleles underlying insecticide resistance. The predictive features that best explain insecticide resistance phenotypes will be estimated from abundance data, genomic diversity at key molecular pathways, gene flow patterns, and environmental variables.Evaluation:Our hierarchical modeling approach will be evaluated based on its abilityto predict insecticide resistance insecticide resistance in CPB over space and time, ascomparedto the model from Objective 2. The developed statistical framework will be assessed for how it can be leveraged to improve precision management of pesticide resistance in Wisconsin.Objective 3: Forecast spatial "risk-maps" of resistance evolution.Effort:We will develop a spatially explicit, demographic simulation framework to simulate evolution of a quantitative trait (insecticide resistance) across the Wisconsin growing region. Forward-in-time, population-based simulations will be generated and then refinedby assessing its "fit" to an additional spatiotemporal dataset. This dataset will includephenotypic resistance, genomic diversity, abundance, and environmental variation from 2027. We will fit the simulated risk map data to actual observations of allele frequency change in 2027 using an approximate Bayesian computation (ABC) framework.Evaluation:With this model fitting exercise, our risk map prediction can be evaluated for accuracy using goodness of fit tests of the simulated and observed data. If the simulation framework is working well, this procedure can further refine the risk map to a narrow set of simulated conditions. We will assess how this risk map can be used for decision making about insecticide applications and toenhance regional management of CPB.