Amphibian Decline and Climate Envelope Modelling

A letter published in Nature two weeks ago expands nicely on two issues I explored in this blog. As I discussed, amphibians are one of the taxa most severely affected by climate change, and it has been suggested that their decline is not only caused by this single factor, but by the interaction of a number of threats including a fungal disease called chytridiomycosis. Holf and colleagues wanted to know how the three major suspected threats, namely climate change, chytridiomycosis, and land-use change may interact in the future to threaten amphibians, and what are the spatial distribution of these threats in relation to amphibian diversity. They used climate envelope modelling to project these threats to the year 2080, and examined weather the most affected areas overlap with species rich areas or with one-another.

Their findings are not too promising for amphibians.
 A general finding was that the most species diverse areas, namely Central America, the Northern Andes, Africa are amongst the most severely affected regions. A majority of species in all three amphibian orders are predicted to be 'climate losers' that is they are predicted to loose habitable climatic space. Moreover, the three threats may interact, although the geographic overlap between areas most affected by each of the three threats is relatively low, and also highly variable. For example in the range occupied by frogs, there's only 6.1% overlap between areas most affected by chytridiomycosis and climate change, so the interaction of these two may not be as significant in the future as previously proposed. For salamanders on the other hand, this figure is twice as high, so there is a significant area where populations will be strongly affected by both climate change and disease. Finally, there are a number of potentially harmful factors not considered by the model, such as direct exploitation for food, invasive species, pet trade and medicine.

The fact that  however does not necessarily mean extinction. As I explained earlier, these model do not account for dispersal ability, and therefore ignore the possibility that species may colonise new, climatically suitable habitats. Secondly, as it is evident from the map, this study was carried out on the continental scale. While disease dynamics or climate may be altered in the majority of an area, suitable conditions may persist on the micro or local scale, providing refugee for amphibians.