Challenges in Identifying Climatically Matched Sites for Animal Invaders
Author Information
Author(s): Gordon H. Rodda, Catherine S. Jarnevich, Robert N. Reed
Primary Institution: U.S. Geological Survey Fort Collins Science Center
Hypothesis
Can MaxEnt models effectively identify climatically suitable areas for invasive species like the Indian Python?
Conclusion
MaxEnt models, when used with default settings, may not accurately identify all sites at risk of colonization due to issues like overfitting and data dredging.
Supporting Evidence
- MaxEnt models showed high sensitivity to the selection of input localities.
- Inclusion of erroneous localities significantly affected model outcomes.
- Default settings in MaxEnt led to overfitting and instability in model predictions.
- Models failed to accurately capture the fundamental climate space necessary for predicting invasions.
Takeaway
This study looks at how well a computer program can predict where invasive snakes might live based on climate, and it finds that the program can make mistakes.
Methodology
The study evaluated MaxEnt models for the Indian Python, focusing on the sensitivity of the models to input choices and the potential for overfitting.
Potential Biases
Potential biases arise from using collection localities that may not represent the true distribution of the species.
Limitations
The models were sensitive to the choice of input localities and background regions, which could lead to inaccurate predictions.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website