Does adding multi-scale climatic variability improve our capacity to explain niche transferability in invasive species?

Our understanding of how species will respond to global change is still limited. Reasons hindering our ability to forecast species range shifts and expansions are the mismatch between realized climate niches in species' native and invasive ranges, and the lack of available climatic datasets offering multiple scales of climatic variability (e.g., monthly and inter-annual climatic variability). Here we present a multi-taxon analysis of invasive species niche transferability that considers multi-scale climatic variability using ten noxious terrestrial invasive species. We compare native versus invasive ranges in geographic space as estimated using the species distribution modeling (SDM) algorithm Maxent, and the comparative index Hellinger's I with three possible climatic layer configurations representing natural climatic variability: (1) inter-annual, (2) monthly and (3) a combination of inter-annual and monthly climatic variability. We assess model performance using the area under the receiver characteristic curve (AUC). Results show that combined scales of climatic variability improved performance of the models for 60% of the species in the native range and 70% of the species in the invaded range. Contrasting native and invaded range SDM performance showed that the same climate layer configuration produced the best models only in 70% of the species. For 90% of the species the most similar niches were obtained based on monthly climatic variability. The divergence in our findings between higher performing SDMs and the most transferable SDMs, suggest some species range limits might be constrained by one scale of climatic variability in the native range and a different one in the invaded range. Where sufficient occurrence data in the invaded range is available, the inclusion of an additional scale of climatic variability can enhance ecological understanding of the invasion events. However, when invaded range occurrence data is not available, the most conservative approach would use only monthly climatic variability. If these finding are extrapolated to niche transferability in time, we suggest that historical collection records should be analyzed to understand species' response to multiple scales of climate variability in the past, thereby informing the selection of appropriate scales of climate variability in the future. (C) 2012 Elsevier B.V. All rights reserved.