Since 1980, the chytrid fungus, Batrachochytrium dendrobatidis (Bd) has putatively caused 67 extinctions out of just over 100 species in the toad genus Atelopus.  Some researchers have argued that Bd-related declines can be explained by climate change, whereas others have argued that climate is not necessary to explain Bd-related declines.  However, most of the evidence supporting climate change as a factor in declines comes from a positive, but temporally confounded, multi-decade correlation between air temperature and Atelopus extinctions.  If there is a true relationship between climate and extinctions, fluctuations around temperature and extinctions should also positively correlate, providing stronger evidence of a causal relationship.  We therefore used the Atelopus year of “extinction” database to test this hypothesis.  We also conducted a manipulative experiment to test two additional hypotheses: that the effects of climate on Bd-related declines could only be thoroughly understood by considering its effects on both Bd growth and amphibian susceptibility, and that temperature variability would increase amphibian susceptibility to Bd.  In this experiment, we first acclimated Cuban tree frogs and southern toads to either 15ºC or 25ºC and then, after half the animals in each treatment were challenged with Bd, we either maintained temperature or switched temperature treatments.  For each of these four temperature treatments, Bd growth was monitored in culture and on the frogs.

Results/Conclusions

After controlling for temporal trends in climatic variables (extrinsic factors) and presumed density-dependent spatiotemporal spread of Bd (intrinsic factor), it became apparent that Atelopus extinctions were most common in El Niño years and in years with greater temperature variability, variables that positively co-varied.  These results suggest that the availability of susceptible hosts is the primary factor influencing the spatiotemporal spread of Bd followed secondarily by climate.  In our manipulative experiment, a shift in temperature did not affect frog survival, suggesting that something other than temperature variability was driving the El Niño-related mortality.  Although Bd grew best in culture at 25ºC, it was most deadly to Cuban tree frogs at 15ºC, which might explain why Bd-related mortality often occurs at cool localities and in cool seasons.  These findings emphasize the importance of controlling for intrinsic factors and examining host-parasite interactions, rather than the host or parasite in isolation, to fully understand the effects of extrinsic factors on disease dynamics.