Cold-induced fever and peak metabolic rate in the nine-banded armadillo (Dasypus novemcinctus).

Abstract

In response to cold exposure, some mammals, including the nine-banded armadillo (Dasypus novemcinctus), exhibit an increase in core temperature. This response, which can be qualified as a cold-induced fever, could increase cold tolerance by increasing peak metabolic rates because of the Q(10) effects. This hypothesis, however, is not compatible with the observation that peak core temperature can occur up to 100 min before peak metabolic rate in nine-banded armadillos during acute exposure to cold heliox (79% He; 21% O(2)). This temporal separation between the timing of peak metabolic rate and core temperature could be the result of regional heterothermy, of the confounding effects of activity, or of using heliox as a respiratory gas. We tested these potential sources of error by exposing nine-banded armadillos to cold air while simultaneously monitoring behavior, metabolic rates (V dot o2 and V dot co2), and four core temperatures. Cold air exposure resulted in a smaller but significant temporal separation, with peak core temperature occurring on average 10 min before peak metabolic rate. Animals exhibited low activity levels, and the four core temperatures changed according to the same temporal pattern, thus eliminating the possibility that activity or regional heterothermy caused the temporal separation. Using a conceptual model, we propose that the temporal separation resulted from a rate of cooling that was too fast.