Summary

As climate change alters temperature and rainfall patterns, the size and intensity of wildfires are projected to increase. Invasive plants, particularly in the tropics and subtropics, contribute significantly to wildfire fuel. Identifying high-risk invasive plants through a literature-based screening system can help inform management and prevent wildfires. This study develops and tests a screening system to assess the wildfire risk posed by invasive plants in Hawai′i, using a combination of expert surveys and machine learning techniques. 49 invasive species in Hawai′i were used to train the system to predict expert fire risk scores based on 21 plant traits. The screening system had high accuracy for identifying plant species that pose higher fire risk.

Take-home points:

  • The most important predictors of fire risk are whether a plant is reported as flammable, if it is a grass (graminoid), if it has flammable relatives, and if it is a fire-promoting invader elsewhere. whether it has been reported as flammable somewhere else was the most important variable for predicting wildfire risk.

  • The literature-based assessment scores used in the system had a high predictive ability, able to correctly identify 90% of plants considered to be a high fire risk.

  • The assessment questions are not location-specific and the screening system can be applied with little or no modification in other fire-prone regions of the world.

Management Implications:

  • Implementing the screening system can help resource managers and policymakers identify and manage high-risk invasive plants before they become widespread.
  • Combining the screening results with local expertise and historical fire data can enhance the effectiveness of wildfire management strategies.
  • The screening system used in this study may be useful to land managers and decision makers for identifying plant management priorities in any area where wildfire is a concern.