Air densities and associated oxygen levels decline by around 10% for each 1000 m rise in altitude, which roughly translates to a 27% decline at 3 km, a 40% decline at 5 km and a 65% decline at 10 km, with minor temporal and regional variations. The decline in air density means that, in theoretical still air, birds have to work harder to keep aloft at high altitude, but meet less resistance to forward flight. The maximum range speed increases by an estimated 5% for each 1000 m rise in altitude (Alerstam 1990a), but so does the power required to fly at this speed, which in turn entails a corresponding increase in energy and oxygen consumption. Although there may be no gain in overall range, it is still advantageous to fly high, because the increased cruising speed shortens the flight time. In addition, stronger winds and reduced turbulence at high altitudes may reduce the energy costs and flight times even further. Other advantages of high-altitude flight include: (a) a wider view of the ground (which may help birds to stay on course, and maintain a straighter track); (b) avoidance of obstacles and deflection by mountains; and (c) a reduced chance of predation (because birds get above the zone where falcons and other aerial predators normally hunt).
Compared with mammals, birds can breathe much more efficiently. Their various air-sacs, which lie within the thoracic and abdominal cavities, as well as between the integument and body walls and within the bones, are connected to the lungs, and enable birds to extract oxygen from thin air more effectively than could lungs alone (Chapter 1). Respiratory movements act mainly on the air-sacs, causing a continual stream of air to pass back and forth through the lungs. The lungs themselves do not hold a substantial residue of used air, as in mammals. In addition, the air within the air-sacs reduces the specific weight of the bird and, being warmer (and hence thinner) than ambient air, may help to provide lift. As flight speed increases with height, while oxygen availability declines, the optimal flying altitude might be regarded as that where the bird could extract just enough oxygen to maintain its maximum range speed. In practice, however, wind conditions seem to have much more influence on flight altitudes, often overriding other factors of potential importance.
One cost of high-altitude flight is the climb involved, which is a major effort for heavy species such as swans (see earlier), although some of this cost may be compensated towards the end of the flight, as the birds gain distance while losing height. Climbing capacity becomes critical to Brent Geese Branta bernicla during spring as they attempt to cross the 2500-m-high Greenland ice cap (Gudmundsson et al. 1995). These birds leave Iceland with enough fuel to reach their breeding areas in northeast Canada 2600-3500 km away. Five radio-marked individuals were tracked by satellites to west Greenland. Their climbing rates were extremely low, only 0.01-0.06 m per second, reflecting the severe limits to their flight power when carrying large body reserves. Their average movement up the ice slope was so slow (on average only 2.5-14.0 km per hour) that the geese must have stopped frequently on the way up. Their downward journey was of normal speed, 29.5-54.7 km per hour, indicating a more or less uninterrupted descending flight. Not all birds crossing high-altitude areas would have the option of resting en route, as suitable habitat may not be available. Climbing to high altitudes may be more costly for migrants making only short flights than for those making long flights, because a greater proportion of the journey is spent on the climb.
Air temperature falls by about 7°C for every 1000 m rise above sea level (or 2°C for every 1000 feet). Over much of the temperate zone, a typical night temperature at ground level in autumn might be 5°C, so that at 1-km altitude, a bird experiences an ambient temperature of about — 2°C, while at 3 km it experiences — 16°C and at 5 km about —30°C. To this must be added a chill-factor dependent on wind. To some extent, the heat generated by flight could compensate for the ambient heat loss caused by low air temperature, but at the high altitudes at which birds sometimes fly, heat generation could clearly impose an additional energy drain. In contrast, over hot deserts the heat generated by flight could lead to overheating which the bird could counter only by evaporative water loss through panting. In hot environments, therefore, high-altitude flight at lower temperatures could greatly reduce the dehydration risks, at least up to the point where the air becomes so thin that water loss is raised through increased respiratory panting (Chapter 6). Air temperatures are also lower at night than by day, and by flying at night, migrants realise at least the same temperature difference as they would experience between sea level and 1000 m during daytime.
The heat produced by working flight muscles is not trivial: only about one-fifth of the energy they generate is mechanical, while the remaining four-fifths is heat, the surplus of which must be dissipated by respiration or convection. The lower the ambient temperature, the more heat can be lost by convection, and the less water is required for cooling. In one early experiment, a budgerigar flying in a wind tunnel at an air temperature of 18-20°C dissipated by evaporation only about 15% of the waste heat generated in the flight muscles, whereas at 36-37°C some 47% of the heat was dissipated in this way, entailing much greater water loss (Tucker 1968). Probably most of the remaining heat at both temperatures was lost by convection from the thinly feathered under-surface of the wings.
Air humidity tends to decline with increase in altitude, especially above the cloud layer. The extreme cold at high altitudes could also enhance water loss, for cold air is relatively dry when it is breathed in, but saturated when it is exhaled. In addition, if oxygen extraction by the lungs is to remain unchanged at high altitudes, the ventilated volume of air must increase (Carmi et al. 1992). This need has been calculated to increase from sea level to 5 km altitude by as much as 175% per unit distance flown and by 254% per unit time flown in a swan (although swans are unlikely to normally reach this altitude, Klaassen et al 2004). For this reason, too, migration at very high altitudes brings not further reduction in dehydration risk, as in hot deserts, but increased dehydration risk.
Clearly, the costs and benefits of flight at different altitudes vary with circumstances, as well as between species, and the physiological constraints under which they operate. Taking all these external factors into consideration, along with the variations in body mass, shape and flight mode between species, it is not surprising that different types of birds seem to fly at different altitudes: sea-ducks and quail just above the waves, songbirds mostly up to 1.5 km, and shorebirds often at more than 3 km. Even on the same night, geese migrating over southern Sweden flew at 100-800 m, while shorebirds over the same site flew at up to 3.7 km above sea level (Green 2004).
Whatever the chosen altitude, individual migrants travelling by flapping flight have been found by height-finding radar to maintain remarkably constant heights on migration (Eastwood & Rider 1965). Although birds have no known sense organ which could detect altitude, this and other evidence suggests that they have some sort of pressure sense (Chapter 9).
The main advantages of low flying (by flapping flight) are that: (1) little energy is expended on climbing; and (2) the ground below is more clearly visible. The main disadvantages are the greater risks from overheating and dehydration, and from predation. Species that migrate by day often fly fairly low (within visual range), and react to the presence of topographical features, such as mountains or coastlines. They may form into streams, as they fly along shorelines or river valleys, or become funnelled through mountain passes. But species that migrate by night at high altitude often seem from their radar tracks to be little influenced by topography below; they usually fly on broad fronts and often cross mountains and coastlines without deviation. In such species, the need to minimise the expenditure of time or energy may have promoted the evolution of behaviour which led to birds taking the shortest routes across barriers, even at the cost of increased risk (such as severe weather). The compromise between time-saving or energy-saving direct routes on the one hand and long risk-reducing detours on the other may be drawn differently in different species, according to their flight capabilities and refuelling needs, the length of their journeys, and the habitat and weather expected en route.
In conclusion, there is little doubt that the altitude of migratory flight is related to prevailing atmospheric conditions, especially wind speed and direction, but also to cloud thickness and height, topography and other factors, as well as features of the birds themselves. We are in need of models which predict the optimal flight behaviour for a range of different species, taking all these variables into account.
Was this article helpful?