National-scale risk mapping and modeling for invasive forest pests

PARTNERS: North Carolina State University (NCSU) Department of Forestry and Environmental Resources; USDA Forest Service Forest Health Technology Enterprise Team (FHTET)

SUMMARY: Eastern Threat Center researchers have been working with FHTET, the Canadian Forest Service, and other scientists to develop regional-, national-, and continental-scale risk map products for non-native invasive forest pests, such as the sirex woodwasp (Sirex noctilio), Asian longhorned beetle (Anoplophora glabripennis) and Phytophthora disease of alders (P. alni). Much of the work has focused on addressing gaps in current knowledge about fundamental characteristics of biological invasions, such as pathways--particularly human-mediated pathways--of long-distance pest spread and pests’ subsequent access to host species in both developed and undeveloped landscapes. Knowledge gaps are being addressed through exploration and derivation of spatial and statistical data sets on international trade, pest interceptions, commodity and freight analysis, transportation corridors, landscape connectivity, disturbance, and other environmental factors, many of which are designed for different purposes and so must be adapted to a forest pest context. Work with FHTET has also extended to development and implementation of methodologies for tasks related to risk assessment, such as the design of national pest detection surveys and spatio-temporal modeling of pest range expansion.

Researchers are examining the representation of uncertainty in forest pest risk maps, with a particular emphasis on how to take typically non-spatial techniques for uncertainty management and place them in a spatial context. Current research topics include the representation of uncertainty in risk map outputs, sensitivity analysis/uncertainty analysis techniques, and multi-model approaches to inference that are intended to minimize uncertainty. Special interest exists in info-gap decision theory and similar conceptual approaches that consider whether the degree or level of uncertainty present in a risk assessment would be significant enough to change decisions made based on the primary results of the assessment. Related project: Characterizing and quantifying uncertainty in forest pest risk analyses

EFETAC's ROLE: This project is supported by Eastern Threat Center funding and collaborative research.

STATUS: Ongoing

PROGRESS: EFETAC research ecologist Frank Koch has collaborated with FHTET and scientists from other Forest Service units as well as other agencies and institutions to develop national-scale and/or regional-scale risk map products (such as maps of introduction, establishment, and susceptibility potential, as well as survey sample area polygons derived from the risk maps) for a number of forest pests:

Koch has also been working with FHTET staff on the development of methodologies and associated data for the current (2012) and future iterations of the National Insect and Disease Risk Map (NIDRM). In particular, he has been focusing on potential options for the representation and management of uncertainty in the NIDRM, as well as the use of dynamic, spatio-temporal modeling to estimate risks associated with new or recently discovered forest pests that have not reached equilibrium in terms of their potential distributions. In both cases, Koch is working on ways to adapt both methods and data appropriately for a national-scale risk mapping effort that involves a large group of participants and stakeholders. He has also done extensive work on the spatio-temporal analysis of drought as a stress agent that predisposes trees to pest attack.

With FHTET and Canadian Forest Service, the predicted spread of sirex woodwasp (Sirex noctilio) in eastern North America over the next 30 years has been modeled.

With John Coulston (FIA-SRS), Bill Smith (EFETAC), and Frank Sapio (FHTET), Koch published a paper in the Canadian Journal of Forest Research that describes how to determine optimum sample size as well as the distribution of survey hexes/points in order to stipulate freedom from a particular insect or disease based on a risk map for the pest:

Coulston, J.W.; Koch, F.H.; Smith, W.D.; Sapio, F.J. 2008. Invasive forest pest surveillance: survey development and reliability. Canadian Journal of Forest Research 38: 2422-2433. (PDF)

With Denys Yemshanov from the Canadian Forest Service and FHTET staff, Koch has been exploring quantitative alternatives to multi-criteria/weighted-average approaches for combining multiple spatial data layers into a single map of species risk. One approach, Pareto ranking, has a particular advantage in that it does not require knowledge regarding the relative importance of individual risk components, rather it ranks map locations objectively based on their relative positions in a k-dimensional multivariate space (where k is the number of spatial data layers that are being combined). The collaborators are currently testing the Pareto ranking approach using spatial data layers for the oak splendor beetle (Agrilus biguttatus). Moreover, Koch worked with Yemshanov to develop corresponding spatial data layers for Canada, in order to expand the project to a cross-border geographic scope. A paper describing this work was published in the journal Risk Analysis:

Yemshanov, D.; Koch, F.H.; Ben-Haim, Y.; Downing, M.; Sapio, F.; Siltanen, M. 2013. A new multicriteria risk mapping approach based on a multiattribute frontier concept. Risk Analysis 33: 1694-1709. (PDF)

Other work with FHTET personnel and Yemshanov includes an article, published in the Journal of Environmental Management, which describes a method for determining when and where preemptive quarantine measures might best slow the spread of an expanding invasive species, and thereby reduce the invasion risk for high-priority locations:

Withrow, J.R.; Smith, E.L.; Koch, F.H.; Yemshanov, D. 2015. Managing outbreaks of invasive species – A new method to prioritize preemptive quarantine efforts across large geographic regions. Journal of Environmental Management 150: 367-377. (PDF)

This project has resulted in other related work and publications.

LINKS:

NCSU APHIS Plant Pest Forecasting System (NAPPFAST)

MetroInvasive