Flight at Low Ambient Humidity Increases Protein Catabolism in Migratory Birds

See allHide authors and affiliations

Science  09 Sep 2011:
Vol. 333, Issue 6048, pp. 1434-1436
DOI: 10.1126/science.1210449


Although fat is the primary fuel for migratory flight in birds, protein is also used. Catabolism of tissue protein yields five times as much water per kilojoule as fat, and so one proposed function of protein catabolism is to maintain water balance during nonstop flights. To test the protein-for-water hypothesis, we flew Swainson’s thrushes (Catharus ustulatus) in a climatic wind tunnel under high- and low-humidity conditions at 18°C for up to 5 hours. Flight under dry conditions increased the rates of lean mass loss and endogenous water production and also increased plasma uric acid concentration. These data demonstrate that atmospheric humidity influences fuel composition in flight and suggest that protein deposition and catabolism during migration are, in part, a metabolic strategy to maintain osmotic homeostasis during flight.

During migration, birds may travel thousands of kilometers between breeding and wintering grounds, stopping periodically to replenish fuel stores. The energy for flight is derived primarily from the oxidation of fatty acids stored in subcutaneous, abdominal, and intramuscular fat depots (1). Lean mass (mainly protein) is also catabolized, even while substantial fat stores remain, which results in reductions in the sizes of muscles and organs during flight (2). This use of lean mass can be explained by a variety of factors, including (i) a beneficial reduction in mass in order to minimize energy costs and to increase flight range, (ii) the requirement for gluconeogenesis and anaplerosis of Kreb’s cycle intermediates, (iii) endogenous protein turnover, and (iv) the production and liberation of endogenous water for the maintenance of water balance (16).

During flight, influx of water arises solely from endogenous sources (1, 7). High ventilation rates result in elevated rates of respiratory water loss (8, 9), and it has been suggested that dehydration, not fuel supply, may limit flight range under some conditions (7, 10). Nonstop flights last upwards of 8 hours in songbirds and can last for several days in some shorebirds (11, 12), yet birds show no signs of postflight dehydration in the wild (13, 14). Because the catabolism of fat yields little endogenous water per unit energy (0.029 g H2O kJ–1) relative to protein catabolism (0.155 g H2O kJ–1), protein may be a crucial source of water for long-distance flights (1, 7). Therefore, increasing the relative contribution of protein to the fuel mixture would represent a functional metabolic strategy whereby substrates are catabolized to avoid water stress, rather than simply to meet energy requirements.

Here, we test the protein-for-water hypothesis in a migratory bird during flight. Swainson’s thrushes (Catharus ustulatus), nearctic-neotropical migrants, were flown in a temperature- and humidity-controlled wind tunnel under conditions of high evaporative water loss [HEWL: 2 g H2O m–3; 13% relative humidity (RH)] and low evaporative water loss (LEWL: 12 g H2O m–3; 80% RH) at 18°C and 10 m s–1 true wind speed. These conditions were chosen because they are ecologically relevant for this species and resulted in a minimum water vapor density deficit of 13.2 g m–3 for the HEWL flights and 3.2 g m–3 for LEWL flights (the underlying assumption is that breath was cooled to ambient temperature upon exhalation and was saturated to 15.2 g m–3). The HEWL treatment represents moderately dry conditions comparable to what birds may encounter during desert crossings (15).

Five individuals flew a total of 20 flights, where each individual was flown two to eight times in paired flights of matching durations. Each pair of flights consisted of one flight under each humidity treatment, and the initial humidity conditions were determined randomly. Total body mass changes were measured with a balance (0.001 g). Dry fat, wet lean, and total body water masses were measured immediately before and after flights using quantitative magnetic resonance (QMR) body-composition analysis (16). Blood samples were taken after the final QMR scan, within 5 min of the end of the flight. General linear mixed models (Proc Mixed; SAS 9.2) were used for comparisons between humidity treatments, with flight duration taken into account and repeated measures on individuals. For further description of the wind tunnel, detailed experimental methods, and statistical analyses, see the supporting online material (17).

Initial voluntary flight durations ranged from 30.3 min to 315.8 min (mean ± SEM: 158.9 ± 19.6 min). Each flight under the initial humidity condition was matched in duration (within 29 ± 7 s) under the alternate humidity condition. Flight costs, as estimated by changes in body composition, if we assume 39.6 kJ/g for dry fat and 5.3 kJ/g for wet lean mass (1), were 4.1 ± 0.2 W (HEWL) and 4.2 ± 0.2 W (LEWL), which represents about a ninefold increase in metabolic rate over resting, if we assume a respiratory quotient of 0.71 at rest (18, 19). These flight costs were not significantly different between treatments (F1,8 = 0.09, P = 0.777), and are similar to those estimated in wild free-flying Swainson’s thrushes by the doubly labeled water technique (4.3 W) (20). However, a direct comparison should be made cautiously because of differences in experimental conditions between studies.

During flight, the rate of total mass loss differed between humidity treatments (flight duration by treatment interaction: F1,7 = 13.11, P = 0.0085) (Fig. 1A). Under HEWL conditions, birds lost 10.23 ± 0.81 mg min–1 [95% confidence interval (CI) 8.36 to 12.10], whereas birds lost 7.27 ± 1.4 mg min–1 (95% CI 4.04 to 10.50) under LEWL conditions. The rate of lean mass loss also differed between humidity treatments (flight duration by treatment interaction: F1,7 = 7.40, P = 0.0297) (Fig. 1B), when flight under HEWL conditions resulted in a 3.55 ± 0.91 mg min–1 rate of lean mass loss (95% CI 1.45 to 5.65). There was no significant relation between amount of lean mass lost and flight duration under LEWL conditions (95% CI –2.00 to 4.11). The rate of fat mass loss did not differ between humidity treatments (flight duration by treatment interaction F1,7 = 0.31, P = 0.597) (Fig. 1C) and was 5.7 ± 0.31 mg min–1 (95% CI 4.98 to 6.41) for both treatments combined. Lean mass contributed on average 16.8 ± 3.2% of the energy under HEWL conditions and 10.6 ± 3.4% under LEWL conditions, yet owing to the low energy density of lean mass, this had little effect on total energy expenditure. All birds had substantial fat stores (preflight fat was 16.8 ±1.1% of initial mass), and there were no differences in initial fat between treatments (F1,9 = 0.27, P = 0.614). Therefore, it is unlikely that differences in lean mass catabolism between humidity treatments were due to differences in fat availability, flight costs, or energy demand (6).

Fig. 1

(A) Mass loss and (B) wet lean mass loss were significantly higher during flights at HEWL conditions (closed symbols and solid lines) relative to LEWL flights (open symbols and dashed lines). (C) Fat mass losses were similar between treatments. Symbol shapes are specific to individual birds. See text for detailed statistics.

There were no significant differences between humidity treatments in relative body-water content postflight (relative to total mass: 54.7 ± 0.01%, F1,9 = 1.09, P = 0.325; relative to lean mass: 78.7 ± 0.002%, F1,9 = 0.22, P = 0.647). Plasma osmolality was not affected by humidity treatment (F1,5 = 1.88, P = 0.229) (Fig. 2A) or flight duration (F1,5 = 0.22, P = 0.656), which indicated that the birds were not dehydrated (21), even after prolonged flight under HEWL conditions. Plasma uric acid concentration was not significantly related to flight duration (F1,6 = 3.76, P = 0.10), but was higher after HEWL flights (F1,6 = 6.23 P = 0.046), which indicated elevated protein catabolism (Fig. 2B). The additional lean mass catabolism during HEWL flights resulted in a 21.7 ± 4.9% increase in endogenous water production relative to energy expenditure (F1,8 = 23.09, P = 0.001) (Fig. 2C), which offset evaporative water losses experienced during flight.

Fig. 2

Plasma osmolality of Swainson’s thrushes was unaffected by ambient conditions experienced during flight in a wind tunnel (A), yet plasma uric acid was higher in birds flown under HEWL conditions (B). Greater lean mass catabolism during HEWL flights resulted in significantly higher endogenous water production relative to energy expenditure (C). Lines connect paired flights within individuals.

We have previously shown a protein-for-water metabolic strategy in water-restricted house sparows (Passer domesticus) at rest but not when metabolic rate was elevated by shivering (22). Therefore, it was not known if this strategy would be evident in flight because of high water gains from fat catabolism alone. Our findings show that under moderately dry conditions, water loss in flight is sufficient to induce water production through increased protein catabolism. Moreover, the strong influence of the rate of water loss on the relative use of fat and protein indicates that the fuel composition in fight is influenced by factors other than energy demand and that a physiological mechanism must exist by which lean mass catabolism is responsive to ambient humidity and/or water stress.

A basal rate of protein catabolism likely occurs to produce metabolites necessary for sustained fatty acid oxidation and gluconeogenesis, while also reducing body mass and producing some water. High rates of water loss result in an elevation of lean mass catabolism above this basal level. An added benefit to elevated lean mass loss may be a concomitant reduction in body mass, which over the course of a long flight could result in energy savings (5), although this was not evident in our experiment. The benefits of lean mass catabolism in flight must be weighed against potential costs of reduced size and functional capacity of muscles and organs. Most important, nutrient-processing tissue lost from the alimentary tract, liver, and other organs must be rebuilt upon arrival at stopover sites before high rates of refueling are possible (23). Therefore, protein sparing should be most important for species that need to refuel quickly during frequent, short refueling bouts, such as most migratory songbirds.

Many studies have documented lean mass deposition and catabolism during the migratory period (2, 2427), yet the significance of these processes for the maintenance of physiological homeostasis has only previously been explored theoretically (1). Thrushes in our experiment were able to maintain osmotic homeostasis even in the face of a fourfold difference in water vapor–density deficit. Humidity treatment did not affect plasma osmolality or relative body water content, which suggests that cellular water homeostasis is maintained at the expense of protein. By maintaining cellular hydration status, perturbations of ion gradients important to cellular metabolism, and in particular muscle contraction, may be avoided.

The protein-for-water strategy has clear functional significance for bird migration. The maintenance of water balance is an immediate necessity during migratory flight, and the use of protein to this end, within limits, allows birds to complete migratory flights in the face of unfavorable environmental conditions. For example, recent observations indicate that trans-Saharan migrants will fly under conditions unfavorable to water balance in order to take advantage of favorable winds (28). It may be that these birds are able to optimize flight range despite high rates of evaporative water loss by increasing lean mass catabolism. When compared with the conditions of this experiment, trans-Saharan passerine migrants seem to be exposed to much more extreme water deficits [estimated minimum water vapor–density deficit of 22.6 g m–3 in (15)], which may account for the dramatic reductions in lean mass that have been recorded in trans-Saharan migrants (24, 29)

When considered from the perspective of water balance, the oft-reported hypertrophy of guts and muscles during migration seasons could in part reflect the only mechanism available to birds to provision water for flight. Flying for long periods under dehydrating conditions may require storage of water in a low–energy density, high–water content form, such as glycogen or protein, which would cause additional increases in the sizes of muscles and organs beyond those associated with powering flight at higher body mass or processing more food. We hypothesize, therefore, that birds preparing to cross major ecological barriers, particularly those with low atmospheric humidity (deserts), should store excess lean mass in preparation for departure. Moreover, the protein-for-water strategy may be unique to uricotelic animals, which dispose of nitrogenous wastes with minimal associated water losses. Migrating bats, for example, may show very different responses to water stress than birds.

Our findings will be useful for the modification of existing bird flight range models by incorporating the effects of atmospheric conditions on fuel mixture, which provides new insight regarding the limits to migratory flight. We have shown that ambient conditions experienced aloft can affect the composition of metabolic fuels used in flight, which may then influence flight range or overall pace of migration. Higher average temperatures in flight could potentially increase water losses and require birds to stop prematurely or to catabolize more protein. Additional time at stopover could then be required to replenish excess protein catabolized in flight.

Supporting Online Material

Materials and Methods

Tables S1 and S2

References (3033)

References and Notes

  1. For further description of the wind tunnel and detailed experimental methods and statistical analysis, see the supporting online material on Science Online.
  2. Acknowledgments: We thank W. Bezner-Kerr for technical assistance with wind tunnel operations; M. Rebuli, T. Farrell, B. Kriengwatana, and A. Macmillan for assistance with animal care; Y. Morbey and L. McGuire for expert statistical advice; A. Boyle, T. Crewe, L. Kennedy, B. McCabe, S. Nebel, E. Price, B. Thurber, and M. Gerson for assistance or advice during this project; and P. Taylor, J. McCracken, S. Mackenzie, M. Burrell, and the volunteers at the Long Point Bird Observatory for their assistance. A.R.G. was supported by an Natural Sciences and Engineering Research Council of Canada (NSERC) Alexander Graham Bell Canada Graduate Scholarship. This study was funded by an NSERC Discovery Grant and by Leaders Opportunity and New Initiatives grants from the Canada Foundation for Innovation and Ontario Research Fund to C.G.G. and colleagues.
View Abstract

Stay Connected to Science

Navigate This Article