Bat flight

Abstract

Bats are unique among extant flying animals, as they have compliant wings and an echolocation sensory system that distinguish them from birds and insects. Flying in the dark, guided by echolocation, has influenced the aerodynamics of bat flight perhaps more than previously realized and resulted in a characteristic flight that is now being revealed. Bats are unique among extant flying animals, as they have compliant wings and an echolocation sensory system that distinguish them from birds and insects. Flying in the dark, guided by echolocation, has influenced the aerodynamics of bat flight perhaps more than previously realized and resulted in a characteristic flight that is now being revealed. Bats evolved muscle-powered flight about 65 million years ago, alongside birds, pterosaurs (probably extinct when bats evolved) and insects. The oldest fossil bat dates 55 million years back and, hence, there is a 10 million year gap in the early evolution of bats where information about the initial adaptive radiation is still missing. The oldest well preserved bat fossils, Onychronycteris finneyi and Icanonycteris index, exhibit all of the features of modern bats, including elongated fingers to span out the wing surface and ear morphology suggesting that at least Icanonycteris was using echolocation. Since their earliest appearance, bats have diversified (Figure 1) and adapted to different ecological niches and many different flight strategies. The present count amounts to more than 1200 species, which means that one in five mammal species is a bat, only outnumbered by rodents. Their body size ranges from 2 g to 1.6 kg, a tenth of the size range of birds. Bat wings vary from short and broad in species that maneuver in cluttered habitats to long and narrow in species of the open airspace. Although the main wintering strategy in Northern hemisphere temperate climates is hibernation, some bats are migratory between summer reproductive areas and Southern wintering sites. Flight makes bats highly mobile and allows them to exploit many biomes and ecological niches. Here, we focus on the essential elements of bat flight. There are many features that distinguish the bat airframe, the wings and body, from that of birds and insects. These features have consequences for their flight performance. The most apparent one is perhaps how the wing surface is built. In birds and insects, the wings are mainly constructed from dead material (keratin feathers or chitin cuticle) giving them a limited ability to actively control the wing surface shape. Bats, on the other hand, have a wing constructed from live skin stretched by the elongated arm and fingers. The skin is 4–10 times thinner than expected, and the bones have a reduced mineralization, compared to other similar sized mammals, reducing the weight of the wing considerably. Skin is living tissue, packed with sensors, elastic fibers and in the case of bat wings also with specialized muscles (Figure 2A). The skin is anisotropic, with higher compliance (i.e. being permissive to load) parallel to the trailing edge, affecting how the skin deforms when subjected to aerodynamic forces, as reflected in strain measurements during flight. Intrinsic muscles in the wing membrane (Figure 2A), not connected to any bones, are thought to control the stiffness of the membrane and thereby the wing’s camber, the curvature of the wing profile. Recent studies have shown that these muscles are indeed active during specific phases of the wingbeat. Studies of artificial membranes with electrically controlled compliance have shown to be able to improve aerodynamic performance. Having the wing stretched by fingers gives bats a high morphing ability, i.e. the ability to change the shape of the wing (Figure 1H). The fingers can spread and bend to different degrees, changing the wing area by stretching the membrane or controlling the camber of the wing and as a consequence the lift coefficient of the wing. (The lift coefficient is a measure of the efficacy of a wing to generate lift and is defined as CL = 2L/ρU2S, where L is lift, ρ is air density, U is local speed about the wing and S is wing surface area.) Recent studies of 3D kinematics of bats show that area and camber are indeed controlled throughout the wingbeat and across flight speeds. We find a higher camber and larger area at slow flight speeds, where the lift-generating requirements of the wing are high. Camber is controlled by a range of mechanisms including bending of the fifth digit (pinky), deflection of the legs (increasing the camber at the inner wing) and through raising or lowering of the second digit (index finger) relative to the main wing surface. This latter mechanism allows the membrane in front of the third digit (middle finger) on the outer wing to function as a leading edge flap, which is stabilized through a mechanical locking mechanism, known as the ‘Norberg mechanism’, to give the wing a stiff leading edge. The leading edge flap affects the curvature of the leading edge, which may be a way to control the leading edge vortices (see below) that bats use to increase the force production of the wing at low speeds. In addition to the wings, many bat species have a tail membrane. The size of the tail varies between species (Figure 1), and has been suggested to play a role in maneuverability, but also seems to be correlated with ear size. However, bats also use the tail to capture prey (Figure 1F), and the size of the tail is related to foraging style. The skin of the tail is tougher than that of the wings, which has been suggested to be an adaptation to withstand puncture by the sharp structures of many of the prey insects. When punctures do occur in the tail or wing membranes they have a tendency to heal rapidly. The aerodynamic function of the tail in bats is currently not very well understood, but it is worth noting that also here we expect a difference compared to birds. The tail of birds is separated from the wings, providing an independent control surface. In bats, the tail as well as the wings are connected to the legs, resulting in the control of the tail and wings to be linked. For example, deflection of the legs to increase the camber of the wing will also deflect the tail and ‘scooping’ with the tail during prey capture will result in a very high camber on the inner wing (Figure 1E,F). A flying animal needs to provide weight support to stay aloft and thrust to counter the drag (drag arises mainly due to flow-separation and to a lesser degree from friction of the air moving across the body and wings). Weight support is generated by producing lift by the wings and thrust by tilting the lift generated by the wings forwards by flapping the wings. The strength of lift is proportional to the wing area, the angle of attack (angle at which air meets the wing), the camber (curvature) of the airfoil and the speed by which the wing moves through the air (actually it is speed squared). As noted above, bats are able to control all of these factors by adjusting the shape and movement of the wings. Due to continuous morphological changes of the highly compliant bat wing during the stroke cycle (planform, camber, angle of attack, twist) a reliable analytical aerodynamic analysis is not currently feasible. However, the generation of aerodynamic force by wings or bodies is matched by the shedding of vorticity (see below) into the wake, which according to fluid dynamic principles reflect the magnitude and time history of the forces on the wing (Kelvin’s theorem). The vorticity conveniently self-organizes as vortices that can be visualized and quantified. Vorticity measures the angular spin of a fluid element and if integrated over a material surface we obtain the circulation, which is directly proportional to the aerodynamic force. Scientists study the geometry and strength of wake vortices of bat flight in wind tunnels by using particle image velocimetry. An example of wake vortices from a bat is shown in Figure 3, where the three-dimensional wake is seen from different angles. The main features are the two wing-tip vortices trailing the wing-tip path, reflecting the overall lift generated. In-board from the tip-vortices are the root-vortices (Figure 3B) that are shed from the junction between wing and body, which reflect a steep gradient on lift generation as a consequence of low lift generation over the body. Also the upstroke is aerodynamically active at most speeds, even though the net force is lower than during the downstroke due to flexing of the wing. At cruising speeds, bats often generate two vortex loops at the transition from up- to downstroke (seen as symmetric kinks at the narrowest point of the wake in Figure 3B), which reflect forward thrust and negative weight support. These vortex loops seem unique to bats thus far and may reflect an optimal way of dealing with relatively high drag. Despite a rather large variation in the bats studied (mega and micro chiropterans, cruising and hovering specialists), these wake structures have been found in all studied species thus far. As can be imagined from the wake structures of Figure 3, bat flight is a rather complicated matter. Many bats are capable of slow flight and hovering (zero forward speed through the air), often used to obtain food from foliage or flowers. At slow speed and hovering, the wake is dominated by the downstroke that provides the majority of the aerodynamic force, but also the upstroke is active providing some weight support. During hovering the bat is not aided by a forward speed to generate aerodynamic force (in forward flight the speed ‘seen’ by the wing is the vector sum of forward speed and the flap speed), but has to achieve this by flapping the wings at high angle of attack. At high angle of attack, the flow separates at the leading edge of the wing, but instead of stalling (leading to the loss of lift) bats develop a leading edge vortex that remains stable throughout most of the downstroke. The formations of leading edge vortices is a universal phenomenon in slow animal flight, and the circulation of the leading edge vortex increases the lift significantly. Compared to birds, the leading edge of the bat wing forms a sharp edge, which facilitates the separation of the flow and the formation of leading edge vortices. During the kinematic upstroke the angle of attack of the outer wing is such that a leading edge vortex is developed on the morphological underside of the wing (Figure 4A), now facing upwards as the wing is rotated spanwise at the turn from down- to upstroke (Figure 1H). At the same time, a leading edge vortex of the same rotation direction as during the downstroke is present at the inner wing some time into the kinematic upstroke during slow flight (Figure 4B), hence resulting in two simultaneous leading edge vortices, of different rotation direction, at different wing positions. This is only possible to achieve with a very high wing twist, and is thus likely to be a unique feature of bats. The control of the complicated aerodynamics of bat flight is largely uncharted territory, but, for example, to maintain a stable leading edge vortex at the wing surface requires exquisite control of angle of attack, camber and wing twist. Tiny hairs covering the wing membrane, which transmit neural signals as a response to wind speed and direction, provide information about the flow near the wing. Interestingly, it seems as if these hairs are most sensitive to airflow coming from the trailing edge going towards the leading edge (Figure 2B). This is to be expected if these hairs’ main function is to monitor flow separation and possibly the presence of a leading edge vortex. Removal of wing hairs cause bats to increase flight speed and it decreases their ability to execute turning maneuvers, suggesting the hairs are involved in the control of slow flight. A dominant feature of bats is their ability to echolocate. This feature has allowed them to dominate the skies when birds can no longer find their way in the dark (although some birds have evolved a primitive form of echolocation). However, the ability to echolocate efficiently may conflict with the requirements of efficient flight. To lower the cost of echolocation, bats synchronize their calls with the wingbeats (except when prey is detected, the rate is increased as the so-called ‘feeding buzz’). This gives a call repetition rate of about 6–20 Hz, which is size related, with larger bats having a lower wingbeat frequency and thus lower call repetition rate than smaller bats. Compared to the flicker frequency of birds’ eyes (∼100 Hz), this gives the bats a low temporal resolution of the information about the environment around them. In addition, recent experiments suggest that the relatively noisy information from echolocation affects the bats’ ability to target prey, which is improved by visual input. Considering a lower fidelity of the sensory information would suggest it is safer to fly slower, and bats tend to fly slower than similar sized birds. The aerodynamic requirements of slow flight should select for a lower wing loading, which is also what we find in bats when comparing them to birds. It would also suggest that the larger the bat, the less likely it would be to fly in a cluttered environment. This is what is found, but it may also be related to maneuverability per se, which is lower for larger bats. To be able to echolocate efficiently you need ears. Bat ears are essentially parabolic shaped structures facing the air flowing over the bat as it is flying. Parabolic structures are known for their high drag, and it has been suggested that the ears may be tilted forward to reduce the drag and potentially aid in the lift production, but so far the results on model bats have not been able to conclusively show an aerodynamic benefit of protruding ears. Comparison of the wakes of freely flying birds and bats have on the other hand shown that the body of bats generate relatively little lift, indicating there may be an aerodynamic cost of ears. This is further confirmed when comparing the aerodynamic efficiency between birds and bats, where birds have both higher lift-to-drag ratio (a measure of the relative drag compared to the weight of the animal) and higher span efficiency than bats (a measure of the efficiency of generating lift by an airfoil, comparing the downwash distribution with the ideal uniform downwash). The independent evolution of powered flight in different animal groups makes a good case for comparative studies. Bats and birds have solved the same problem (flight) in very different ways, and although we find some differences in performance they show many similar characteristics and flight performance is perhaps more similar than different. This is partly dictated by the physical world in terms of numbers such as acceleration due to gravity, air density and viscocity, but within these bounds the adaptive diversity among birds and bats bears witness of the strength contained in the process of adaptive tinkering.