| dc.description.abstract | Appendix for: Species-level forecasts of distributional potential and likely distributional shifts in the face of changing climates have become popular in the literature in the past 20 years. Many refinements have been made to the methodology over the years, and the result has been an approach that considers multiple sources of variation and how that variation translates into signal and uncertainty in the signal in the model outputs. Although numerous previous reviews and overviews of this field have pointed out a series of basic assumptions and caveats with the methodology, three aspects of the methodology have impacts that are nothing less than overwhelming, and that have not been treated previously in such assessments. Here, we assess (1) the effects of niche truncation on niche model transfers to future climate conditions, (2) the effects of model selection procedures in a future-transfer challenge for ecological niche models, and (3) the relative contributions of different factors (samples of point data, general circulation models, representative concentration pathways, alternative model parameterizations) to overall variance in model outcomes. Overall, the view is one of caution… this paradigm has pitfalls and potholes that may bias and limit confidence in niche model outputs as regards the implications of climate change. | en_US |
