Predicting current and future biological invasions: both native and invaded ranges matter

Area of Science:

• Ecology
• Invasive Species Biology
• Biogeography

Background:

• Traditional species distribution models train on native ranges, but invasive species may adapt to new climates.
• Changes in realized climatic niche can limit the accuracy of predictions in invaded regions.

Purpose of the Study:

• To propose and evaluate an alternative modeling approach using pooled data from all known ranges.
• To improve predictions of invasive species' geographical extent, specifically for spotted knapweed (Centaurea maculosa).

Main Methods:

• Developed and applied a modeling approach that incorporates data from both native (European) and invaded (North American) ranges.
• Compared predictions from pooled data models with those trained solely on native or invaded ranges.
• Projected species distribution under a severe climate warming scenario for the year 2080.

Main Results:

• The pooled data approach significantly improved invasion extent predictions compared to models using only the native range.
• Pooled models performed comparably to models based solely on the invaded range.
• Projections indicated fewer areas of intermediate climatic suitability with pooled models, suggesting reduced uncertainty.

Conclusions:

• Pooling data from native and invaded ranges offers a more robust method for predicting invasive species spread.
• This approach enhances consensus among modeling techniques and decreases uncertainty in future distribution forecasts.
• The findings are crucial for effective invasive species management under changing climatic conditions.