Invasive species, including invasive pathogens and insect pests (hereafter referred to as pests), are an increasing threat to forest health across the U.S. and globally. Unfortunately, management options may be limited, especially when invasive pest populations are widespread because the cost and scale of management are prohibitive. Early detection and mitigation of invasive pests in forests, as well as in tree nurseries and other sources of seedlings for restoration initiatives, is essential to limiting invasive pest transmission.

Current methods for invasive pest detection rely on visual observation of symptoms associated with infection or infestation, followed by confirmation using PCR (polymerase chain reaction) tests in select cases. However, advances in imaging and sequencing technologies have expanded the tools available for early invasive pest detection. Two tools with great potential are field deployable, non-destructive spectral imaging and DNA barcoding.

Researchers will use a combined approach of spectral imaging and DNA barcoding to identify early symptoms of invasive pest infection in elm, beech and chestnut trees. Spectral imaging will be used to monitor symptom development over time in infected trees to identify early cues of invasive pest infection and to provide more quantitative assessments of variation in susceptibility to invasive pests. DNA barcoding will provide added sensitivity and specificity and will be used to confirm invasive pest presence. Researchers will engage multiple partners, from state, municipal, federal, and non-profit organizations in this project.

Objectives

Develop a tool for rapid, in-the-field identification of invasive pests and more accurate and quantitative symptom monitoring. Create a library of spectral images and DNA barcodes for detection of invasive forest pests. Develop a web-based portal for invasive pest identification. Organize training sessions for partners and stakeholders in use of new detection tools.

Expected Results

Outputs

Compilation of a library of tree spectral images, along with DNA barcodes, from target pests from each tree species of focus containing data from both healthy and infected trees. Models which can facilitate scaling up from leaf-level to remote applications in the future. An App for collecting color image data from trees (for symptom monitoring) and a web-based portal for data storage and processing. Scientists anticipate the tool being used to collect spectral/image data directly in the field and uploaded to the App/dashboard to aid in pest detection and quantifying symptom severity in near real-time.

Expected Outcomes

Development of more robust, early screening tools for identifying pest-infected trees, which in turn will facilitate surveys for invasive pests and resistance screening programs on forest lands. Deployment of tools by forest health monitoring and tree breeding programs. Reduction in the time, expense, space and variability associated with traditional pest detection and screening methods. Direct transferability of tools to projects already planned for screening elm trees for Dutch elm disease resistance. Trainings for partners and stakeholders in tool usage to ensure successful technology transfer.

Metrics of Success

Development of a spectral and image library including data from at least 100 individual trees from each species studied (elm, beech and chestnut). Development of models for predicting pest infections within each species. Development of a smartphone App for collecting color image data for pest symptom monitoring and a dashboard for data processing. Training sessions on tool usage organized for partners and stakeholders.

Geographies and High-Risk Landscapes to Be Addressed

National

This work will aid in the detection of forest pests in the Eastern U.S. with the potential to address global pest issues.