Abstract
During migration, long-distance migratory songbirds may fly nonstop for days, whereas shorter-distance migrants complete flights of 6 to 10 h. Fat is the primary fuel source, but protein is also assumed to provide a low, consistent amount of energy for flight. However, little is known about how the use of these fuel sources differs among bird species and in response to flight duration. Current models predict that birds can fly until fat stores are exhausted, with little consideration of protein's limits on flight range or duration. We captured two related migratory species-ultra long-distance blackpoll warblers (Setophaga striata) and short-distance yellow-rumped warblers (Setophaga coronata)-during fall migration and flew them in a wind tunnel to examine differences in energy expenditure, overall fuel use, and fuel mixture. We measured fat and fat-free body mass before and after flight using quantitative magnetic resonance and calculated energy expenditure from body composition changes and doubly labeled water. Three blackpolls flew voluntarily for up to 28 h-the longest wind tunnel flight to date-and ended flights with substantial fat reserves but concave flight muscle, indicating that protein loss, rather than fat, may actually limit flight duration. Interestingly, while blackpolls had significantly lower mass-specific metabolic power in flight than that of yellow-rumped warblers and fuel use was remarkably similar in both species, with consistent fat use but exceptionally high rates of protein loss at the start of flight that declined exponentially over time. This suggests that protein may be a critical, dynamic, and often overlooked fuel for long-distance migratory birds.