Fig. 5. Energetic costs of active arousal from torpor as a function of ambient temperature (T_a) for arousals from minimum T_a without passive heating (triangles) and for arousals after diurnal passive heating (filled circles). Passive heating provided a linear reduction in the average cost of active arousal [solid line: energy expenditure for active arousal (kJ)=0.84–0.026xT_a (°C); r²=0.87, P<0.001]. During torpor, passive heating resulted in a net increase in energy expenditure, which can be included as a component of the total energy expenditure for arousal [open circles; dashed line: total energy expenditure for arousal (kJ)=0.78–0.02T_a (°C); r²=0.82, P<0.001].