The impact of non-native insects on North-American trees can be widespread and significant. A long-standing goal of urban and forest managers and invasion ecology is to predict which newly arriving invader may cause significant damage. Using available data pertaining to currently established non-native insects in forest systems, we evaluated potential drivers that could explain why some invading insects cause high impacts while others are benign. These drivers included 1) insect traits, 2) host tree traits, especially those related to defenses, 3) evolutionary divergence time between the insect’s native and novel (North American) host trees, and 4) the phylogenetic relatedness of the invader to native insects coevolved with the novel host. Contributions of these drivers were quantified using information theoretic inference, and significant driver sub-models were combined for overall predictive models. In addition, all collected data was compiled into a Traits and Factors Catalog (TRAFAC) as a resource for future invasion biology research questions.
Results/Conclusions
This study offers quantitative insights into the role of evolutionary history in predicting the impact of non-native insects on host trees, particularly for insects with specialized host ranges. We discovered the evolutionary histories of the novel host and its native-insect community are more important determinants of impact than traits of the invading insect. As the invader is the typical focus of invasion frameworks, our findings are a major step in advancing understanding of the interactions between non-native insects and novel host plants. Our composite models can be used to predict the risk that non-native insects not yet present in North America will cause tree mortality should they successfully establish. For example, the risk that sap-feeding invaders will cause high impact was greatest when the insect was attacking a novel conifer that was shade tolerant and drought intolerant, that diverged from the insect’s coevolved (native) host 12-17 million years ago, and there was no congeneric insect already utilizing the novel host. The results from our predictive models and the TRAFAC databases, can be useful tools for regulatory and pest management efforts focused on preventing and mitigating insect invasions in North America.