Javascript verkar inte påslaget? - Vissa delar av Lunds universitets webbplats fungerar inte optimalt utan javascript, kontrollera din webbläsares inställningar.
Du är här

Feathers by day, membranes by night - Aerodynamic performance in bird and bat flight

  • Florian Muijres
Publiceringsår: 2011
Språk: Engelska
Dokumenttyp: Doktorsavhandling
Förlag: Department of Biology, Lund University


Popular Abstract in English

Animal flight has always intrigued people, and different aspects of animal flight have therefore widely been studied by biologists, physicists and engineers. At least to me, the aspect concerning the aerodynamics of animal flight in particular is very interesting because of the complex temporal varying wake dynamics as a result of continually wing morphing in flapping flight. This puts aerodynamics of animal flight among the most complex and intriguing phenomena in fluid dynamics.

Within the animal kingdom, there is a large diversity of flying animals, ranging from large soaring albatrosses to tiny hovering insects, which can be explained by two factors, namely ecological requirements and phylogenetic constraints.

The variation between the different taxa of flying animals (birds, bats and insects) is due to the fact that flight has evolved independently for each flying taxa. This has resulted in very different body plans of the flight apparatuses, for example birds have feathered wings and bats have membranous wings. These different body plans set different limitations to the flight dynamics for each taxa, and ultimately limits flight performance. These limitations to evolutionary change are called phylogenetic constraint.

The morphological differences between species within a certain taxa, for example between an albatross and a hummingbird, can primarily be explained by differences in ecological requirements. The ecological requirements for a certain species set the optimum flight performance for that species, and this is enforced by evolutionary selection pressure. In fact, the albatross in the example above is a result of nature’s attempt to make an efficiently gliding sailplane based on a bird body plan (Fig. 1), resulting in the well known landing and take-off problems. On the other hand, the hummingbird is nature’s attempt to make an insect out of a bird.

To summarize, optimum flight performance for a certain species is set by its ecological requirements, while the acquired maximum performance is limited by the phylogenetic constraints for that species. In this thesis I investigate how the relative flight performance in two bat species and one bird species differ and how these differences depend on their ecology and evolutionary history. For this I have developed methods for visualizing and quantifying the wake dynamics behind a flying animal in a wind tunnel, and which is used to quantify aerodynamic flight performance.

I have studied flight in two species of nectar feeding bats: the Pallas’ Long-tongued bat

(Glossophaga soricina) and the Lesser Long-nosed Bat (Leptonycteris yerbabuenae). Both species feed from flowers by hovering in front of them, so both may be regarded as hovering specialists. Next to this,

L. yerbabuenae is a highly mobile species, since it migrates annually up to 1600 km from Mexico to Arizona (USA), and it flies around 100 km every night between its roosting sites and different feeding patches. G. soricina, on the other hand, does not migrate and has an extremely small home-range where it both roosts and feeds from flowers (recapturing distance for wild G. soricina is only 200 m).

The studied bird is a typical small insectivorous passerine, the Pied flycatcher (Ficedula hypoleuca). It habitually hovers and flies slowly when catching insects on the wing, and it migrates annually from northern Europe to western Africa. Thus, flycatchers are expected to be well adapted to efficient slow flight as well as efficient cruising flight.

I studied the aerodynamics of flight for these species in a wind tunnel. In a wind tunnel the air moves instead of the animal, which enables me to study flight from up close in a controlled environment. The wake dynamics is studied using a technique called Particle Image Velocimetry (PIV). This technique allows visualization and measurement of air movements within a defined plane, by tracking fog particles illuminated by a laser sheet. I did two types of measurements: streamwise PIV measurements close to the wing surface, called on-wing PIV; and PIV measurements in a cross-stream plane closely behind the flying animal, called near-wake PIV.

For the PIV analysis, I developed a method for visualizing (fig 6 and 8) and quantifying the fluid dynamics around and behind a flying animal. The wake dynamics is quantified by converting the vortex strength and downwash in the wake into flight forces. These flight forces are then used to estimate the aerodynamic performance for flapping flight.

I have defined two aerodynamic performance measures, being the lift-to-drag ratio L/D and the span efficiency ei. L/D is a value for the efficiency of locomotion since the lift force L is the force required to stay up in the air and drag D is the force related to the cost of flight (Fig. 2). ei is a value for the efficiency of lift production, where a high ei means low power required to generate lift.

As explained above, both bat species studied are hovering specialist, but their movement ecology is very different: L. yerbabuenae flies much longer distances than G. soricina. This means that L. yerbabuenae should be more adapted to efficient flight at cruising flight speeds than G. soricina. In this thesis, I compare these differences in ecological requirements with differences in aerodynamic performance. It turns out that L/D is similar between the two bat species, but the speed at which L/D is maximum (highest efficiency) is close to cruising flight speed for L. yerbabuena , while this speed is much lower for G. soricina. Thus, both bats fly most efficient at the speed at which they fly most of the time.

Both the bird and bat generate a Leading Edge Vortex (LEV) (Fig. 6 and 7), which is an unsteady vortex structure commonly used by insects to enhance lift. In this thesis, I show for the first time that also vertebrates use LEVs. The strength of the LEV is very similar in birds and bats, adding 49% and 40% of the total lift produced, respectively, but the LEV structure is very different between bird and bats. In bats the LEV strength increases along the wing span from low close to the body to high near the wingtip. This LEV structure similar to that for insects. The flycatcher, on the other hand, produces a LEV that is strongest at mid wingspan, and weakest near the wingtip. The reduction in LEV near the wingtip is probably a result of upwards bending of feathers near the tip. This is probably a mechanism to stabilize the relatively strong LEV in birds.

The wake dynamics behind the flying birds and bats differ significantly (Fig. 8), which is a result of the fact that birds and bats have both a very different body shape and wingbeat upstroke.

Birds use a so called a feathered upstroke, where they retract the wings close to the body and spread the wing feathers. This enables air to flow through the gaps between the feathers, making their wings aerodynamically inactive during the upstroke (Fig. 8c).

Bats cannot do this since air cannot flow through the solid wing membrane, and if a bat would retracts its wings too much the membrane would go slack, resulting in uncontrolled fluttering. Therefore, bats have an active upstroke where they move the wing upwards at a negative angle-of-attack generating thrust and negative lift. This results in the production of a ‘reversed vortex loop’ in the wake behind each wing (Fig. 8a-b).

Throughout the complete wingbeat, the birds generate more body lift than the bats, which suggests that bird bodies are more streamlined than bat bodies. This could be partly due to bats having more blunt shaped bodies and protruding ears required for echolocation. Concave shaped structures such as bat ears are known to be among the most drag producing bluff bodies.

All these differences together result in both higher L/D and span efficiency for the birds than for the bats, suggesting that birds outperform bats in aerodynamic flight performance. Still it looks like the wake dynamics for both birds and bats are optimal for their relative flight performance regime (for the L/D regime they operate at). This suggests that evolution has adapted the wingbeat kinematics for birds as well as for bats to optimize their flight performance, although differences in their basic body plan (e.g. bird wings have feathers and bats wings consist of membranes) has resulted in a higher maximum performance in birds.

All in all, I have shown that aerodynamic flight performance depends strongly on the ecology and evolutionary history of a specific species. The currently used flight performance models, though, do not capture variation in flight performance due to these detailed characteristics. The here found results can be used as a starting points to improve these animal flight models. This would enable us to tailor flight performance estimates better for a specific species.
The efficiency and performance of a flying animal is directly related to the aerodynamics around its body and flapping wings. Here, I have developed methods for quantifying the wake dynamics around a flying animal. The results are used to estimate the aerodynamic performance of flapping flight. Using these methods, I have studied flight of the Pied Flycatcher (Ficedula hypoleuca), the Pallas’ Long-tongued bat (Glossophaga soricina) and the Lesser Long-nosed Bat (Leptonycteris yerbabuenae).

In paper I, the aerodynamics close to the wing surface of slow flying G. soricina bats was studied, showing that bats use a Leading Edge Vortex (LEV) to enhance lift with up to 40% of the total. LEVs are known to be used by insects, but here I have shown that also larger vertebrates can use LEVs. In paper II, the aerodynamics close to the wing surface of a slow flying Pied Flycatcher was studied. This results showed that Pied Flycatchers generate LEVs with similar strength as in G. soricina, but the LEV structure is significantly different from that of bats and insects. In paper III, a new high-speed stereoscopic Particle Image Velocimetry (PIV) system for studying animal flight was introduced. Using this system, the wakes of the two bat species were captured, and new methods for visualizing and analyzing wake data were introduced. In paper IV, the wake dynamics and aerodynamic performance of flapping flight for the two bat species was studied. Although the wake dynamics for the two species was similar, maximum aerodynamic performance was achieved at a significantly higher speed for the highly mobile and migratory L. yerbabuenae than for the non-migratory G. soricina. In paper V, I introduced an actuator disk model for analyzing time-resolved PIV data of flapping flight. Analysis of the wake data for the two bat species showed that the model can be used to compare flight efficiency of different animal species. In paper VI, the wake dynamics in flycatchers was studied. The results showed that the wake of slow flying flycatchers is more similar to that of fast flying passerines than to that of hummingbirds, and that flycatchers are probably aerodynamically more efficient than hummingbirds. In paper VII, the wake dynamics and aerodynamic performance for the three studied species was compared. This showed that birds outperform bats in aerodynamic efficiency, which could be ascribed to differences in aerodynamic function of the body and of the wing upstroke, and which were proposed to be a result of differences in phylogenetic constraints between birds and bats.


Blå Hallen, Ekologihuset, Sölvegatan 37, Lund, Sweden
  • Tom Daniel (Prof.)


  • Biological Sciences
  • animal flight
  • aerodynamics
  • vortex wake
  • actuator disk
  • flight performance
  • span efficiency
  • bird
  • bat
  • wind tunnel
  • PIV
  • Pied Flycatcher
  • Glossophaga soricina
  • Leptonycteris yerbabuenae


  • Anders Hedenström
  • Christoffer Johansson
  • ISBN: 978-91-7473-086-9

Box 117, 221 00 LUND
Telefon 046-222 00 00 (växel)
Telefax 046-222 47 20
lu [at] lu [dot] se

Fakturaadress: Box 188, 221 00 LUND
Organisationsnummer: 202100-3211
Om webbplatsen