Remember your roots (if you’re an invasive plant, that is)

Pfadenhauer, W. G., Nelson, M. F., Laginhas, B. B., & Bradley, B. A. (2023). Remember your roots: Biogeographic properties of plants' native habitats can inform invasive plant risk assessments. Diversity and Distributions, 29(1): 4–18.

Summary

Incorporating climate change into invasive plant policy and management requires that we be more proactive about identifying novel species that pose high risk. Weed risk assessments are a proactive tool for evaluating risk from novel species. Weed risk assessments often include lots of different criteria for predicting invasion risk, but it’s unclear which of these criteria are actually effective. Pfadenhauer et al. (2023) tested how well environmental characteristics of plants’ native habitats predict invasiveness elsewhere. They found that heterogenous native ranges (in other words, native ranges with widely varied habitats) were significantly more common for invasive plants when compared to non-invasive plants. In other words, if many different habitats are part of a species’ native range, the species will likely be able to survive and invade in many different habitats outside of its native range. In contrast, resource availability, disturbance, and plant richness in the native range were not effective at predicting invasiveness. Pfadenhauer et al. conclude that breadth of environments encompassed by a species’ native range is an important criterion for assessing invasion risk associated with novel plant species. 

Take home points

  • It is common for invasive plants to have large native ranges that include lots of habitat variation. 

  • Native ranges with widely varied soil characteristics (soil texture, soil pH) are strong indicators of invasiveness outside of the native range.

Management implications

  • Management tools that aim to predict which plants will invade (for example, weed risk assessments) should include ‘native range size’ and/or ‘breadth of soil textures in the native range’ as predictive variables. 

  • Data for these “biogeographic properties” are commonly available online, making them particularly useful, especially when compared to other harder-to-compile variables like physical traits.

Keywords

Management efficacy, Invasive plants, Weed risk assessment, Predicting invasive species, Machine learning