RESPIRATORY EXCHANGE AND VENTILATION DURING NOCTURNAL TORPOR IN HUMMINGBIRDS

In mammals that undergo torpor and hibernation, a period of CO2 storag e is often a prelude to, and may be required for, the onset of a bout of torpor. Storage of CO2 has been hypothesized to induce an energy-co nserving metabolic suppression in torpid mammals. It is unclear whethe r CO2 storage also occurs in birds that undergo torpor. To study these questions, we measured rates of oxygen consumption (V over dot(O2)) a nd carbon dioxide production (V over dot(CO2)) during normothermy, ent rance into torpor, steady-state torpor, and spontaneous arousal in Sel asphorus hummingbirds. The respiratory exchange ratio (RER) declined f rom about 0.86 to about 0.71 in both normothermic and torpid individua ls, suggesting a shift from carbohydrate oxidation to lipid oxidation during the night. Torpor metabolism ranged from 5% to 42% of normother mic values. In most individuals that became torpid, a brief period of CO2 storage occurred during entrance into torpor, beginning after V ov er dotO(2) had fallen considerably below normothermic rates. The volum e of CO2 stored averaged 0.048 mL CO2 g(-1), somewhat less than report ed for mammalian hibernators. Some individuals entered torpor without measurable CO2 storage. There was no abrupt change in RER at the start of arousal, although mean arousal RER exceeded the RER during steady- state torpor. The change in metabolic rate between normothermia and to rpor had a Q(10) of 1.9-2.2 in birds that did not regulate body temper ature in torpor. We conclude that a normal Q(10) effect is sufficient to explain the metabolic rate of torpid hummingbirds and that transien t CO2 storage may occur during entrance into torpor but is not a prere quisite for entrance. Similarly, CO2 release is not a prerequisite for arousal.