Forest ThreatNet

Scientists' risk assessment tools important for forest management

By Bridget O'Hara, NEMAC

EFETAC biometrician Bill Smith and North Carolina State University cooperating researcher Frank Koch spend a significant amount of time compiling enormous amounts of data – understanding how forest pests spread – and have consequently developed a unique working relationship. Their 5-year partnership includes an unusual way to approach, and solve, a problem. Koch lends insight into the ultimate compromise…. “First, we think about possible solutions, and Bill suggests something completely outlandish. I tell him it won’t work, or will take way too much time or effort to implement. After a couple of days, I ultimately see a way to do what he suggested.”

Smith, who began his career computer programming with wire boards in the 1960s, sees patterns in numbers, and based on a lifetime of experience in numerous disciplines, can create unconventional yet workable concepts. Koch uses his expertise in spatio-temporal analysis, modeling, and data mining to execute a plan and, according to Smith, “can aggregate 15 sets of data in a beautiful way.” This partnership results in award-winning pest risk maps, model outputs, and methodologies for improving these outputs that provide forest managers a jump start on controlling pests nationally and internationally.

Pest risk maps offer a way to prioritize responses to emerging pest threats based on risk while incorporating uncertainty. To create these tools, Smith and Koch gather data sets describing pest risk parameters. Data include how healthy and affected host trees are spread across the landscape; environmental factors such as temperature, precipitation and altitude that may weaken or strengthen pests; and dispersal pathways that allow pests to travel from one forest to another.

They are particularly interested in dispersal pathways because, although invasive species are sometime spread unintentionally with human assistance, it has not been factored into most risk assessment efforts. “We have applied commodity flow and transportation network data, regional agricultural statistics, and similar data sources to identify and represent important human-mediated pathways for forest pest dispersal,” says Koch. “Natural resource managers use risk assessment maps to place detection traps, release biological control agents, or simply decide where devoting resources might be most effective.” Smith adds, “It helps that pest risk maps are also a visual aid, well suited to explaining the nature and degree of a particular threat for both managers and the general public.”

Their Redbay ambrosia beetle (Xyleborus glabratus) analysis demonstrates possible outcomes resulting from limited information and creative thinking. By assembling available data, Smith and Koch developed a fairly solid characterization of a recently detected pest and determined its likely range and expansion through time across the southeastern United States. The resulting journal article has been widely read and forest health specialists from several southeastern states have used these results to plan forest health activities. For additional information, please visit www.forestthreats.org/about/fhm.

Top photo: EFETAC North Carolina State University cooperating researcher Frank Koch (left) and EFETAC biometrician Bill Smith discuss field data collection methods.

Right: Smith and Koch develop risk maps useful to land managers and forest health professionals.

Researchers Receive Top Honors for Risk Analysis Paper

A collaborative effort among the USDA Forest Service, North Carolina State University, and Canadian Forest Service scientists received top honors from the Society for Risk Analysis (SRA). EFETAC collaborative researcher Frank Koch and biometrician Bill Smith coauthored the paper titled, “Evaluating Critical Uncertainty Thresholds in a Spatial Model of Forest Pest Invasion Risk.” The paper explores the role of increased uncertainties in pest risk mapping and was among five papers selected in the 2009 Best Paper Awards category