TRANSIENT HOVERING PERFORMANCE OF HUMMINGBIRDS UNDER CONDITIONS OF MAXIMAL LOADING

Maximal load-lifting capacities of six ruby-throated hummingbirds (Arc hilochus colubris) were determined under conditions of burst performan ce. Mechanical power output under maximal loading was then compared wi th maximal hovering performance in hypodense gas mixtures of normodens e air and heliox. The maximal load lifted was similar at air temperatu res of 5 and 25 degrees C, and averaged 80% of body mass. The duration of load-lifting was brief, of the order of 1 s, and was probably sust ained via phosphagen substrates. Under maximal loading, estimates of m uscle mass-specific mechanical power output assuming perfect elastic e nergy storage averaged 206 W kg(-1), compared with 94 W kg(-1) during free hovering without loading. Under conditions of limiting performanc e in hypodense mixtures, maximal mechanical power output was much lowe r (131 W kg(-1), five birds) but was sustained for longer (4s), demons trating an inverse relationship between the magnitude and duration of maximum power output. In free hovering flight, stroke amplitude and wi ngbeat frequency varied in inverse proportion between 5 and 25 degrees C, suggesting thermoregulatory contributions by the flight muscles. S troke amplitude under conditions of maximal loading reached a geometri cal limit at slightly greater than 180 degrees. Previous studies of ma ximum performance in flying animals have estimated mechanical power ou tput using a simplified actuator disk model without a detailed knowled ge of wingbeat frequency and stroke amplitude. The present load-liftin g results, together with actuator disc estimates of induced power deri ved from hypodense heliox experiments, are congruent with previous loa d-lifting studies of maximum flight performance. For ruby-throated hum mingbirds, the inclusion of wingbeat frequency and stroke amplitude in a more detailed aerodynamic model of hovering yields values of mechan ical power output 34% higher than previous estimates. More generally, the study of performance limits in flying animals necessitates careful specification of behavioral context as well as quantitative determina tion of wing and body kinematics.