Early warning signals do not predict a warming-induced experimental epidemic.

Climate change can impact the rates at which parasites are transmitted between hosts, ultimately determining if and when an epidemic will emerge. As such, our ability to predict climate-mediated epidemic emergence will become increasingly important in our efforts to prepare for and mitigate the effects of disease outbreaks on ecological systems and global public health. Theory suggests that statistical signatures termed "early warning signals" (EWS), can function as predictors of epidemic emergence. While a number of works report post hoc detections of EWS of epidemic emergence, the theory has yet to be experimentally tested. Here, we analyse experimental and simulated time series of disease spread within populations of the model disease system Daphnia magna-Ordospora colligata for EWS of climate-mediated epidemic emergence. In this system, low temperatures prevent disease emergence, while sufficiently high temperatures force the system through a critical transition to an epidemic state. We found that EWS of epidemic emergence were nearly as likely to be detected in populations maintained at a sub-epidemic temperature as they were to be detected in populations subjected to a warming treatment that induced epidemic spread. Our findings suggest that the detection of false positives may limit the reliability and utility of EWS as predictors of climate-mediated epidemic emergence.