Samples of great knots (Calidris tenuirostris) were collected in an earlier project, before and after a 5420-km migration stage from Australia to China (believed to be flown non-stop) to determine the mass of fat consumed, and also the mass of protein withdrawn from the flight muscles and other organs. The flight was simulated by a “time-marching” computation, which calculated the fuel energy required, and allowed different hypotheses to be tried for the consumption of protein. The simulation predicted that the great knots would take about 4 days to cover the distance, in agreement with field estimates. Realistic predictions of the consumption of fat and protein were obtained by setting the conversion efficiency to 0.23 and the body drag coefficient to 0.10, withdrawing sufficient protein from the flight muscles to keep the specific work in the myofibrils constant throughout the flight, and taking enough additional protein from other tissues to bring the energy derived from oxidising protein to 5% of the total energy consumed.
The same computation was applied to published data on the pre-migration body composition of bar-tailed godwits (Limosa lapponica), which are said to migrate over 10 000 km from Alaska to New Zealand. The computed range for a sample killed by collision with an obstruction, while actually departing from Alaska, was sufficient to reach the South Pole. A second sample, shot before departure from New Zealand, would have run out of fat before reaching Alaska, but could easily have reached northern Australia, where these godwits stage on their northbound migration. The higher range estimate for the Alaskan birds was not due to higher fat mass (only 5% difference) but to a higher fat fraction, which they had achieved by reducing the mass of other organs before departure.
Some recent observations of high chemical power, observed in wind tunnel experiments, have been interpreted as being due to much lower conversion efficiency than the value of 0.23 assumed here, but this interpretation is flawed. Measurements of mechanical power from another wind tunnel project were also unexpectedly high, suggesting that unsteady flight by wind tunnel birds increases their power requirements, both mechanical and chemical, with no implications for efficiency. The calculated power is for “steady horizontal flight”, meaning that a valid test of predicted power requires birds to be trained to hold a constant position in the test section, while maintaining a steady wingbeat frequency and amplitude. This has not been achieved in recent experiments, and is hard to achieve when using physiological methods, because of the long periods of continuous flight needed. Measurements of mechanical rather than chemical power require shorter flight times, and offer better prospects for reliable power measurements.