Carbon dioxide gas is used as an insect anesthetic in many laboratories, despite recent studies which have shown that CO 2 can alter behavior and fitness. We examine the effects of CO 2 and anoxia (N 2 ) on cold tolerance, measuring the rapid cold-hardening (RCH) response and chill coma recovery in Drosophila melanogaster. Short exposures to CO 2 or N 2 do not significantly affect RCH, but 60 min of exposure negates RCH. Exposure to CO 2 anesthesia increases chill coma recovery time, but this effect disappears if the flies are given 90 min recovery in air before chill coma induction. Flies treated with N 2 show a similar pattern, but require significantly longer chill coma recovery times even after 90 min of recovery from anoxia. Our results suggest that CO 2 anesthesia is an acceptable way to manipulate flies before cold tolerance experiments (when using RCH or chill coma recovery as a measure), provided exposure duration is minimized and recovery is permitted before chill coma induction. However, we recommend that exposure to N 2 not be used as a method of anesthesia for chill coma studies.

Low temperature and desiccation stress are thought to be mechanistically similar in insects, and several studies indicate that there is a degree of cross-tolerance between them, such that increased cold tolerance results in greater desiccation tolerance and vice versa . This assertion is tested at an evolutionary scale by examining basal cold tolerance, rapid cold-hardening (RCH) and chill coma recovery in replicate populations of Drosophila melanogaster selected for desiccation resistance (with controls for both selection and concomitant starvation) for over 50 generations. All of the populations display a RCH response, and there is no effect of selection regime on RCH or basal cold tolerance, although there are differences in basal cold tolerance between sampling dates, apparently related to inter-individual variation in development time. Flies selected for desiccation tolerance recover from chill coma slightly, but significantly, faster than control and starvation-control flies. These findings provide little support for cross-tolerance between survival of nearlethal cold and desiccation stress in D. melanogaster .