Abstract
Despite evidence that conspecific lineages may display different climatic tolerances, most invasion risk assessment tools are calibrated without considering phylogeographic information. This study aims to investigate the existence of intraspecific niche divergence within a group of insect pests and to explore how the inclusion of phylogeographic information into species distribution models may alter the estimation of the potential distribution of a species. We studied North American bark beetles belonging to the genus Dendroctonus, a group of pests of conifers that are listed as quarantine species in numerous countries. Most Dendroctonus species exhibit strong genetic divergence that appears to be geographically structured and shaped by historical events and biotic factors. We modeled all lineage distributions within five species, using MaxEnt and Boosted Regression Trees, and compared the results with the models fitted at the species scale. Multivariate analysis and niche similarity and equivalency tests were additionally performed to investigate the existence and magnitude of climatic niche divergence between conspecific lineages. We also tested the ability of lineage-based models to predict the region invaded by D. valens in China. Conspecific lineages showed a climatic niche more similar than expected by chance, but displayed different climatic envelopes in their native range and, consequently, different estimates of potential distributions. We also observed that classical models calibrated using the entire range of the species could potentially under-or overestimate the potential range of the species when compared to a global prediction built by aggregating lineage-based projections. This study showed that the invasive phylogeographic lineage of D. valens has invaded regions characterized by climatic conditions highly similar to those encountered in its native range suggesting that preadaptations to environment might have played a role in this invasion. This study highlights how our perception of the invasion risk of pests may be altered when integrating phylogeographic information.