Because hummingbirds are small and have an expensive mode of locomotion, they have constrained energy budgets. Torpor is used to buffer against these energetic challenges, but its frequency and duration vary. We measured lipid content, metabolic rates, and torpor use in two species of migrating hummingbirds, Calliope (Selasphorus calliope) and Rufous hummingbirds (Selasphors rufus) at a stopover site. We constructed a mass-balance model to predict lipid thresholds for torpor entry, torpor duration, and minimum morning lipid reserves. Hummingbirds entered torpor if their lipid contents were below a sharply defined threshold. Torpor duration increased as initial lipid content decreased, and birds that entered torpor had relatively constant morning lipid reserves. We propose a Minimum Morning Reserve Hypothesis that identifies torpor lipid thresholds and predicts frequency and duration. Several hypotheses were proposed previously to explain torpor’s ultimate function, which can be derived as special cases that result from modifying our mass balance model’s parameters. Torpor entails a balance between energy savings and the non-energetic risks of torpor, such as predation and physiological stress. We assessed energy equivalents of the non-energetic costs of torpor by accounting for the energetic costs and benefits of torpor, and by documenting its occurrence and length.

The data used in this research was collected through capture of two species of migrating hummingbirds (S. rufous and S. calliope). Body lipid content was measured using a Quantitative Magnetic Resonance Scanner, and metabolic rates were determined using Flow-through Respirometry. This dataset was used to develop a model to predict torpor thresholds and durations, and minimum morning lipid reserves in migrating hummingbirds.