Generalized “avatar” niche shifts for modeling the potential distributions of emerging data-poor invasive species

Background/Question/Methods

Pathways of introduction and regions of origin for invasive species change over time, and climate change will continue to shift the regions where invasive species come from and where they can establish. Anticipating the potential distributions of emerging invasive species is complicated by the tendency for species distribution models to perform best when both native and invasive range data are available for model development. In response, we propose documenting the extent of niche elasticity in important climate predictors for well-studied invasive species, and then transferring this elasticity as precautionary boundaries for scenarios of niche shifts in emerging invaders. We term these established invaders “avatars” because they are remote representatives of the full invasion process for emerging invaders in early stages of introduction and spread. We demonstrate the avatar invader concept by first characterizing niche shifts for several major invasive crayfishes between native and total (native and invasive combined) ranges using a minimal set of relevant climatic predictors and a measure of multivariate similarity (Mahalanobis distance). We then transfer the magnitude and character of these niche shifts to an emerging crayfish invader, and conclude with discussion on selection of avatar invaders and proposals for evaluating avatar model performance.

Results/Conclusions

We documented niche shifts of similar character for the invasive crayfishes Pacifastacus leniusculus and Procambarus clarkii, but a niche shift of larger magnitude for P. clarkii. The area predicted as suitable for P. leniusculus increased from 8.6% (native) to 8.8% (total), while the area predicted as suitable for P. clarkii increased from 5.4% to 13.5%. Both crayfish experienced increases in minimum temperature variance and reduction in covariance between minimum and maximum temperature from native to total range models, but only P. clarkii had a large increase in maximum temperature variance. When these avatar niche shifts were applied to the emerging crayfish invader Cherax quadricarinatus, the area of the globe anticipated as suitable increased from 7.1% (native) to 10.8% (P. leniusculus avatar) and 19.2% (P. clarkii avatar). We suggest researchers interested in implementing our methodology select avatar species based on ecological and/or phylogenetic similarity to the focal emerging invader and evaluate ensembles of multiple avatars. We also propose testing the validity of the avatar invader concept by comparing transferability of niche shifts between many well-established invasive species, simultaneously providing a framework for rigorous testing of the niche conservatism hypothesis in general.