Abstract

Gliding has evolved independently at least six times in mammals. Multiple hypotheses have been proposed to explain the evolution of gliding. These include the evasion of predators, economical locomotion or foraging, control of landing forces, and habitat structure. Here we use a combination of comparative methods and ecological and biomechanical data collected from free-ranging animals to evaluate these hypotheses. Our comparative data suggest that the origins of gliding are often associated with shifts to low-quality diets including leaves and plant exudates. Further, data from free-ranging colugos suggest that although gliding is not more energetically economical than moving through the canopy, it is much faster, allowing shorter times of transit between foraging patches and therefore more time available to forage in a given patch. In addition to moving quickly, gliding mammals spend only a small fraction of their overall time engaged in locomotion, likely offsetting its high cost. Kinetic data for both take-off and landing suggest that selection on these behaviors could also have shaped the evolution of gliding. Glides are initiated by high-velocity leaps that are potentially effective in evading arboreal predators. Further, upon landing, the ability to control aerodynamic forces and reduce velocity prior to impact is likely key to extending distances of leaps or glides while reducing the likelihood of injury. It is unlikely that any one of these hypotheses exclusively explains the evolution of gliding, but by examining gliding in multiple groups of extant animals in ecological and biomechanical contexts, new insights into the evolution of gliding can be gained.

Introduction

Selection pressures acting on an organism’s locomotor repertoire can result in novel modes of locomotion. While some locomotor transitions have occurred in only a few cases, such as the independent origins of flapping flight in birds, bats, pterosaurs, and insects, others have occurred numerous times. Gliding for example, has evolved independently at least thirty times in vertebrates alone and is likely to be included in the locomotor repertoire of most arboreal organisms (Dudley et al. 2007). In mammals, gliding has evolved at least nine times, including six extant and three extinct lineages (Mein and Romaggi 1991; Storch et al. 1996; Meng et al. 2006), with the earliest known fossils dating to 130 mya (Meng et al. 2006). Further, gliding mammals are a diverse group, including at least 60 species from these distantly related mammalian clades (Wilson and Reeder 1993). Despite their diversity in lineage or form, they all exhibit a gliding lifestyle and might therefore experience common selective pressures resulting in common themes in behavior and ecology. This study will address those possible commonalities and discuss them in light of several hypotheses that have been proposed to explain the evolution of gliding.

Several hypotheses, yet to be tested, have been proposed to explain the evolution of gliding flight. Animals might leap or glide between trees in response to a specific habitat structure (Emmons and Gentry 1983; Dudley and DeVries 1990; Dial et al. 2004), to avoid predators (Emmons and Gentry 1983), or to increase locomotor or foraging efficiency (Norberg 1983; Scholey 1986; Scheibe and Robins 1998; Dial 2003; Scheibe et al. 2006). Alternatively, gliding might have evolved as a means of minimizing the forces associated with landing after long leaps (Paskins et al. 2007; Byrnes et al. 2008). However, in order to gain insight into the selective forces that might have shaped the ability to glide, detailed study of the locomotor behavior of free-ranging gliding animals is necessary.

Directly addressing these hypotheses has been hampered in two ways. First of all, an historical distinction between gliding and parachuting, in which gliding is defined by descent angles <45° to the horizontal (Oliver 1951), has set up an operational definition that gliding occurs in the steady-state and that maximizing the ability to cover long distances “improves” performance. However, numerous examples show that gliding rarely, if ever, occurs in a steady-state in a variety of organisms, including insects (Yanoviak et al. 2005), reptiles (McGuire and Dudley 2005; Socha et al. 2010), and mammals (Bishop 2006; Byrnes et al. 2008). Furthermore, in natural settings gliding animals rarely maximize the distance covered (Table 1). As a result, our working definition follows Dudley et al. (2007) in that gliding is any aerial behavior that involves active regulation of aerodynamic forces and therefore does not include any parachuting behaviors that are strictly passive, with no active modulation of aerodynamic forces. The second challenge to addressing these hypotheses has been the ability to collect unbiased data on the movements of gliding animals in natural habitats. However, incorporating new technologies, including data-logging, into traditional methods of observation have the potential to overcome this challenge. Therefore, the goals of this article are to use available biomechanical and ecological data to address the numerous hypotheses proposed for the evolution of gliding and to suggest avenues of future research into gliding in animals.

Table 1

Mean and maximum glide distances for gliding mammals

Taxon Mean (m) Maximum (m) Mass (g) References 
Petaurus breviceps 20.4 42 69–150 Jackson 1999 
Petaurus gracilis 29.7 60 350–450 Jackson 1999 
Glaucomys sabrinus 16.4 93 93 Vernes 2001 
Petaurista leucogenys 19.5 50 1010 Stafford et al. 2002 
Galeopterus variegatus 31.4 145 >1000 Byrnes et al. 2011 
Taxon Mean (m) Maximum (m) Mass (g) References 
Petaurus breviceps 20.4 42 69–150 Jackson 1999 
Petaurus gracilis 29.7 60 350–450 Jackson 1999 
Glaucomys sabrinus 16.4 93 93 Vernes 2001 
Petaurista leucogenys 19.5 50 1010 Stafford et al. 2002 
Galeopterus variegatus 31.4 145 >1000 Byrnes et al. 2011 

Habitat structure

Gliding animals rely on the potential energy stored by climbing to travel horizontal distances by air. As a result of this dependence on the vertical dimension of arboreal habitats to initiate glides, it is no surprise that several hypotheses relating the structure of forested habitats to gliding have been proposed. It has been suggested that a low density of lianas (Emmons and Gentry 1983) or greater height of the canopy (Dudley and Devries 1990) could explain the worldwide pattern in the diversity of gliders. More recently, by studying patterns in the amount of free space in the canopy in both tropical and temperate forests, Dial et al. (2004) demonstrated that sites with a high diversity of gliders in Asia and Australia were characterized by greater free space in the canopy than in the other sites studied. The canopy-height (Dudley and DeVries 1990) and vertical-stratification (Ando and Shiraishi 1993) hypotheses are especially attractive, given the biomechanical–ecological prediction that larger gliders may be confined to higher strata of the canopy. If taxa are vertically stratified by body size, as has been shown for the flying lizard Draco (Inger 1983), it might be expected that the tallest forests will harbor the greatest diversity of taxa. This prediction is supported by the greatest diversity of gliding mammals in the Dipterocarp forests of Southeast Asia. Although this example is intriguing, many counter examples illustrate the complexity of this relationship. For example, the tallest forests in the world, the North American redwood forests, harbor only one species of gliding mammal. Similarly, in Australia, the greatest diversity of gliders does not occur in the tallest forests. Therefore, in order to test this hypothesis fully, more detailed information about heights and structure of canopies in forests where gliders and non-gliders are actually found is needed.

In addition, historical information about the structure of forests in which fossil gliders first appeared in various regions of the world would greatly enhance our understanding of the interplay between forest structure and the evolution of gliding. Fossils that have been attributed to gliders have been found through geologic time in both modern-day temperate and tropical regions (e.g., Mein 1970; Storch et al. 1996; Meng et al. 2006) and on most continents, with the first appearance of a gliding mammal nearly 130 mya (Meng et al. 2006). Unfortunately, many of these discoveries have been identified solely by dentition (e.g., Black 1963; Mein 1970). However, the difficulty of distinguishing gliding forms in the fossil record solely from dental characters has been well established (Thorington et al. 2005), making their historical distribution difficult to assess. More recently, well-preserved fossil animals with gliding affinities have been discovered (e.g., Storch et al. 1996; Meng et al. 2006), but these belong to extinct groups, thus giving little insight into the evolution of extant gliders. This type of data only exists in any form for the marsupial gliders and shows an intriguing potential correlation between the drying of the Australian continent and concomitant opening of forest canopies with the evolution of the three independent lineages of marsupial gliders (Archer 1984; Jackson 1999).

Avoidance of predators

The three-dimensional structure of arboreal habitats could also have influences on the ability of gliders to escape predation. Flying animals have been shown to have lower mortality than do non-volant animals (Pomeroy 1990). Similarly, after allowing for evolutionary history, arboreal mammals have greater longevity than do their terrestrial counterparts (Shattuck and Williams 2010). Being arboreal, like flying, allows for escape in three-dimensions, increasing the variability of escape paths for these animals. Some birds take advantage of this opportunity and avoid predation by varying their take-off trajectory (Bonser and Rayner 1996). Arboreal mammals could take advantage of similar strategies and it has been debated whether gliding could increase longevity in flying squirrels (Holmes and Austad 1994; Stapp 1994). Furthermore, gliding mammals can leap from arboreal perches and escape the canopy without injury (Emmons and Gentry 1983). Gliding mammals, including flying squirrels (Essner 2002; Scheibe et al. 2007) and colugos (Byrnes et al. 2008), and other arboreal mammals, including tree squirrels (Essner 2002) and primates (Garber 2005; Channon et al. 2010; Legreneur et al. 2010), are capable of leaping from the canopy at high velocity. This ability to change velocity quickly at take-off could be paramount for escaping predators (Howland 1974). However, unlike non-gliding animals that often crash into the canopy of adjacent trees after leaps to escape predators (Stern and Goldstone 2005), gliders can make controlled landings on targeted trees (Paskins et al. 2007; Byrnes et al. 2008). Finally, by rarely venturing to the ground some gliding mammals might minimize risks from terrestrial predators. Although many temperate gliding squirrels frequently come to the ground, colugos, for example, rarely terminate glides on the ground (only 4 of 267 glides) (Fig. 1) and thus could minimize exposure to ground-dwelling predators.

Fig. 1

Angle of orientation (θ) at landing for 267 glides by colugos. Asterisk denotes glides ending in animals landing on the ground (4 of 267 glides, 1.5%). Inset shows how landing angle is calculated.

Fig. 1

Angle of orientation (θ) at landing for 267 glides by colugos. Asterisk denotes glides ending in animals landing on the ground (4 of 267 glides, 1.5%). Inset shows how landing angle is calculated.

Foraging

Diet

Diet has been shown to affect the ecology both of gliders and non-gliders alike. Dietary variables predict many ecological traits including duration of foraging (Goldingay 1989; Comport et al. 1996) and population density (e.g., Wasserman and Chapman 2003). It is therefore likely that selection might act on characteristics that improve an individual’s ability to forage efficiently. The specialized diets of many gliders has led to the suggestion that gliding could have evolved because of its advantage in obtaining scattered or protein-deficient foods, such as leaves or plant exudates, that might not have been able to be harvested otherwise (Goldingay 2000).

Here we use comparative methods to address the hypothesis that gliding might have evolved because of its advantage in obtaining low-quality or scattered resources.

To test this hypothesis, dietary variables were obtained from the literature (Appendix 1, Supplemental Material). Diet was characterized as the most common food resource utilized. A composite phylogeny of 169 taxa (Fig. 2) was constructed including all gliding and closely related groups. The following major groups were included: squirrels (Mercer and Roth 2003), other rodents (Montgelard et al. 2002), primates (Poux and Douzery 2004), bats (Teeling et al. 2002), and diprotodont marsupials (Osborne et al. 2002). These groups were arranged in the context of the higher order mammalian phylogeny (Springer et al. 2004). Polytomies were resolved in the trees by using other phylogenetic information (Thorington et al. 2002; Mercer and Roth 2003) or randomly when no additional information was available. For randomly resolved polytomies, alternate resolutions had no influence on the outcome of the analyses.

Fig. 2

Phylogenetic reconstruction of diet in mammals. Gliding mammals are denoted by boldface type. The tree is a composite phylogeny of 169 mammalian taxa including the six independent lineages of gliding mammals and their close relatives.

Fig. 2

Phylogenetic reconstruction of diet in mammals. Gliding mammals are denoted by boldface type. The tree is a composite phylogeny of 169 mammalian taxa including the six independent lineages of gliding mammals and their close relatives.

Ecological variables were recoded as binary characters, and ancestral character states were reconstructed using parsimony with ambiguities reconstructed using both ACCTRAN and DELTRAN resolutions. Associations between gliding and dietary resources were tested using a concentrated changes test (Maddison 1990). Maddison’s (1990) concentrated changes test is a test of correlated evolution that determines whether a change in a specific character (in this case gliding) is associated with a specific state of a second character (diet). The probability of observing a concentration of changes in one state of a character is compared to the null hypothesis that the changes are distributed randomly on the phylogeny (Maddison 1990). Due to the large tree, data were simulated with n = 5000 in place of using the actual changes. All analyzes were performed using MacClade 4.0 PPC (Maddison and Maddison 2000). Significance was assessed using a rejection level of P < 0.05.

Gliding originated in six independent lineages in the phylogeny. These six origins occurred in only three of the eight categories of diet. Gliding evolved twice each in ancestors with folivorous, frugivorous, and exudivorous diets. Of these, there were significant associations based on the concentrated changes test for both exudivory (ACCTRAN P = 0.011; DELTRAN P = 0.023) and folivory (ACCTRAN P = 0.026; DELTRAN P = 0.049). There was not a significant association between gliding and frugivory (ACCTRAN P = 0.268; DELTRAN P = 0.321) despite gliding having evolved twice in ancestors exhibiting frugivory.

The results of this analysis strongly support an hypothesis relating to foraging behavior. The significant associations between gliding and these two dietary resources support Goldingay’s (2000) hypothesis that gliders may be able to use poor-quality, especially protein-deficient, food resources, which are unavailable to non-gliders. Exudates (Goldingay 1990), seeds (Koenig and Mumme 1987), and some leaves (Kavanagh and Lambert 1990) have been shown to be of low quality or are widely dispersed sources of nitrogen. Some gliders have been observed to track patches of nutrient-rich foliage over great distances (Kavanagh and Lambert 1990), an impossible or energetically costly strategy for non-gliders. Although there are some non-gliding mammals, including Ledbetter’s possum and some species of marmosets, that have exudivorous diets, Goldingay (2000) argued that such animals rely on dense patches of forage to utilize these resources. Although fruit is another often-scattered dietary resource, it was not significantly associated with gliding. Fruit contains a high content of energy, and is eaten by a wide variety of organisms. This finding suggests gliding allows use of low-quality, but not necessarily scattered, sources of food.

Comparison to ballistic leapers

Without detailed information on the movement and foraging patterns of gliders and non-gliders, it is difficult to directly test the hypothesis that gliding improves foraging efficiency. However, by using simple models based on ballistic motion, we can estimate the influence of the aerodynamic force produced by gliding mammals on the distance they travel. The velocity of a body governed by ballistic motion is given by the equations  

(1)
formula
and  
(2)
formula
where v0 is the take-off velocity, ϕ0 is the take-off angle (assumed to be 45° to maximize ballistic range), g is the acceleration due to gravity, and t is time. Using animal-borne data-loggers, we have recorded the take-off and landing impulse (Byrnes et al. 2008), glide distance (Byrnes et al. 2011), and the height climbed to initiate glides in free-ranging Malayan colugos (Byrnes et al., in press). Using these data, we can determine the benefit of producing aerodynamic force on foraging range by comparing our data to that of a model ballistic leaper, constrained with the kinetic parameters we measured. For every recorded glide made by a colugo, we can model a ballistic leap with the same take-off or landing kinetics and determine the distances covered. If we constrain take-off velocity and change in vertical distance to equal the observed data for each glide (Fig. 3A), glide distance exceeds leap distance in most cases (on average by a factor of 2.5), yet in numerous instances ballistic leaping improves range compared to gliding. This estimate is problematic however, because allowing vertical displacement to be equivalent between glides and ballistic leaps results in landing velocities that are approximately ten times greater for leaping, with consequently much higher landing forces. To account for this, we also calculated foraging distance for a second set of kinetic parameters for our ballistic model in which take-off velocity and landing velocity were constrained to equal the observed data from glides. With these more realistic values for landing velocity in the leaping model, gliding distance was more than 20 times greater than the distance of the leap predicted from the model and the distance leapt rarely exceeded the distance of the glide (Fig. 3B). Gliding appears to be of benefit not only because it can increase the distance traveled by an equivalent leap but also allows these large distances to be traveled without excessive momentum at landing (Fig. 4). Despite traveling long distances each night, colugos spend only minutes each day moving between the patches in which they forage. Time itself can be a valuable currency that shapes behavior (Dunbar and Dunbar 1988; Dunbar 1992), and by gliding, colugos, like other gliders (Goldingay 1989; ,Comport et al. 1996; Scheibe et al. 2006), minimize the time spent in locomotion. In turn, gliding might maximize time spent acquiring resources as well as the rate at which resources can be acquired (Charnov 1976).

Fig. 3

Estimated distance of ballistic leaps versus distance of glides. (A) Distances of ballistic leaps are calculated using equal take-off velocity (from Byrnes et al. 2008) and change in altitude (Byrnes et al., in press) from the corresponding glides. (B) Distances of ballistic leaps are calculated using take-off and landing velocities (Byrnes et al. 2008) equal to those from corresponding glides. Calculations of distances of ballistic leaps assume maximum range (45° take-off angle) in both (A) and (B).

Fig. 3

Estimated distance of ballistic leaps versus distance of glides. (A) Distances of ballistic leaps are calculated using equal take-off velocity (from Byrnes et al. 2008) and change in altitude (Byrnes et al., in press) from the corresponding glides. (B) Distances of ballistic leaps are calculated using take-off and landing velocities (Byrnes et al. 2008) equal to those from corresponding glides. Calculations of distances of ballistic leaps assume maximum range (45° take-off angle) in both (A) and (B).

Fig. 4

Landing impulse versus distance traveled for ballistic leaping (gray symbols) and gliding (black symbols). Landing impulses for ballistic leaps were calculated assuming a 45° take-off angle as well as equal take-off velocities and changes in altitudes for glides. Landing impulse for ballistic leaping increases significantly with distance traveled, but landing impulse for gliding is unrelated to glide distance.

Fig. 4

Landing impulse versus distance traveled for ballistic leaping (gray symbols) and gliding (black symbols). Landing impulses for ballistic leaps were calculated assuming a 45° take-off angle as well as equal take-off velocities and changes in altitudes for glides. Landing impulse for ballistic leaping increases significantly with distance traveled, but landing impulse for gliding is unrelated to glide distance.

Locomotor economy

The locomotor-economy hypothesis has received the greatest attention, but measuring the potential costs of gliding locomotion and thus its benefits over other forms of locomotion is logistically demanding. As a result, mathematical models have primarily been used in previous investigations of the potential costs of gliding (Scholey 1986; Scheibe and Robins 1998; Dial 2003). These models compare the cost of climbing to a height necessary to glide a given horizontal distance to the cost of using quadrupedal locomotion to travel the same distance. The metabolic cost per unit time associated directly with gliding has been shown to be low in other vertebrates (Baudinette and Schmidt-Nielson 1974; Sapir et al. 2010). Given the low metabolic cost per unit time and the very short intervals of time over which gliding occurs, this direct cost of gliding should be negligible and has generally been ignored.

These models give insight into the comparative costs of gliding in large and small animals, but rely on generalized estimates of locomotor behavior and glide performance. However, two important predictions have emerged from the models. First, large gliders must glide a much longer distance compared to small gliders before gliding is energetically cheaper than running. For example, Scheibe and Robins (1998) calculated that a small glider, the North American flying squirrel, Glaucomys volans, must glide only 3 m compared to between 50 and 100 m for the red giant flying squirrel, Petaurista petaurista, (Scholey 1986; Scheibe and Robins 1998) before gliding is a cheaper mode of transport. Secondly, due to differential scaling relationships for running and climbing, the energetic benefit of gliding may be greatest at intermediate body sizes (Dial 2003). Unfortunately, few empirical studies have been conducted on the energetic cost of vertical or incline climbing (e.g., Taylor et al. 1972; Wunder and Morrison 1974) and there is not a predictable relationship between energetic cost, incline angle and body size (Full and Tullis 1990), making direct comparison of the costs of climbing to launch a glide as opposed to the costs of running a given distance difficult. Additional study of the metabolic cost of vertical climbing in a diverse group of taxa would be helpful for fully understanding this relationship.

Empirical data have been added to these energetic models but the results have been inconclusive. In a field study of another small North American flying squirrel, it was estimated that Glaucomys sabrinus, must glide ∼10 m before gliding is cheaper than running the same horizontal distance (Scheibe et al. 2006). The comparably sized marsupial glider, Petaurus norfolcensis, however, must glide ∼30 m before gliding would be less costly (Flaherty et al. 2008). These large differences in cost-effective glide distance could be the result of substantial differences in the locomotor ecology of the two species. However, these studies used empirical data from only the initial glides after release at a trap site, making it possible the results could also reflect differences in the escape response of the two species. Further, it is difficult to apply the costs to an animal’s overall energy budget because these studies focus on single glides.

To understand the influence of gliding on the overall energy budget of an animal, detailed study of the locomotor behavior over an ecologically relevant interval of time is required. By using animal-borne data-loggers, we were able to collect detailed data on locomotor behavior for Malayan colugos over several days (Byrnes et al. 2008, 2011). From these data, we were able to quantify every glide and every bout of climbing leading to a glide to determine all vertical and horizontal movements. Combining these detailed data with estimates of the metabolic costs of horizontal (Taylor et al. 1982) and vertical (Hanna et al. 2008) movement from scaling equations, we determined that climbing a given distance to launch a glide was no more economical than running an equivalent distance (Byrnes et al., in press).

The locomotor ecology of gliding mammals also does not support the locomotor-economy hypothesis. Colugos spend only a small fraction of their daily time budget (<2%) engaged in locomotor behaviors (Byrnes et al. 2011). Similarly, two marsupial gliders from independent lineages, the greater glider (Petauroides volans) and yellow-bellied glider (Petaurus australis) use locomotion minimally, spending ∼6% (Comport et al. 1996) and 4% (Goldingay 1989) of their daily time budget engaged in locomotor activities, respectively. Therefore, despite the high cost of climbing, little of the overall daily energy balance is expended for locomotion. For example, based on the scaling of field metabolic rate (Nagy 2005), a 1 kg colugo would be predicted to expend ∼800 kJ per day. On average, colugos expend only 12 kJ climbing each day (Byrnes et al., in press), only 1.5% of their estimated daily energy expenditure. Furthermore, P.volans expends ∼520 kJ per day (Foley et al. 1990). If P. volans were to move similar distances to those of colugos, locomotor costs would total just 2.5% of the total daily energy budget. Therefore, locomotor economy may not play a large role in the ecology of extant colugos and its role in the evolution of gliding is not well supported by data from extant species.

Despite not providing the direct energetic savings that have been hypothesized for locomotion, gliding is a rapid form of locomotion and thus can save travel time, possibly indirectly influencing the energy budget. Time itself is a pressure that can influence behavior (Dunbar and Dunbar 1988; Dunbar 1992) and by gliding, the time spent traveling between the trees used for foraging can be minimized by colugos and other gliders (Goldingay 1989; Comport et al. 1996; Scheibe et al. 2006). In contrast, moving through the canopy requires negotiating the terminal branches of trees, and this could be slow or indirect. The velocities of other arboreal mammals on narrow substrates are relatively slow, approaching 1 ms1 (Delciellos and Vieira 2007; Stevens 2008). Conversely, gliding mammals can travel at velocities up to 10 ms1 or more (Stafford et al. 2002; Scheibe et al. 2006; Byrnes et al. 2008), thereby reducing travel time by as much as 10-fold. Therefore, although climbing and gliding may not maximize locomotor economy, moving quickly between trees allows for more of the active period of an animal to be spent foraging, possibly increasing the net energy balance.

Avoidance of injury

Gliding mammals navigate a complex and discontinuous arboreal habitat under the darkness of night. Discontinuous arboreal substrates require precise placement of limbs and modulation of locomotor forces while moving tens of meters above the ground. Falls from elevated substrates can be common (Schlesinger et al. 1993) and can result in significant injury (Schultz 1939; Jurmain 1997). As a result, arboreality requires high levels of dynamic and postural control. As gaps between substrate elements increase in size, many arboreal animals rely on leaping or gliding between distant supports. When traveling long distances, they leap from trees, glide tens of meters, and then must land safely on the boles of trees they are unable to visualize at take-off. High-speed collisions with stationary objects can prove fatal in other volant animals, including birds (Klem 1990) and bats (Crawford and Baker 1981). Furthermore, impacts with trees during landing are sometimes fatal to juvenile colugos that cling to their mothers’ abdomens during glides (G. Byrnes, personal observation). Because they travel at high-speed while gliding, their ability to modulate aerodynamic forces while airborne and thereby reduce velocity prior to landing is critical for survival.

Unlike many arboreal animals, including monkeys, birds, and tree squirrels, that often land in the compliant terminal branches of trees (Stern and Goldstone 2005), gliding mammals land on the rigid, large-diameter boles of trees. As a result, gliding mammals experience conflicting pressures at take-off and landing. To maximize ballistic range, or to escape predators, gliding mammals leap at high velocity during take-off from their perches (Essner 2002; Scheibe et al. 2007). Furthermore, they accelerate under the force of gravity for much of their glide. To land safely, however, gliders must elicit a landing maneuver that rapidly sheds excess velocity prior to impact.

The importance of the landing maneuver has been recognized since naturalists’ early descriptions in field reports (e.g., Hingston 1914). More detailed descriptions of the landing maneuver have also been presented (Scholey 1986; Ando and Shiraishi 1993; Stafford et al. 2002). In this maneuver, angle of attack is increased to >60° and the animal lands with either just the forefeet or with all four feet simultaneously (Nachtigall 1979; Scholey 1986; Scheibe et al. 2007), reorienting the patagial membrane to the oncoming flow to act as a parachute reducing velocity prior to impact.

To measure landing forces, an instrumented force pole has been used for leaping strepsirhine primates (Demes et al. 1995, 1999) and flying squirrels (Paskins et al. 2007). Over leaps or glides of short distances of up to 2.5 m, peak impact forces increase with distance both in the southern flying squirrel (Glaucomys volans) (Paskins et al. 2007) and in primates (Demes et al. 1995, 1999). However, once an animal enters a steady-state glide with no net acceleration, landing forces no longer increase. Using an animal-borne accelerometry system to measure the landing kinetics of free-ranging colugos (Galeopterus variegatus), Byrnes et al. (2008) found that peak landing forces decreased with increasing length of glide for glides ranging from 2 to 145 m. Furthermore, there is a critical length of glide up to which landing forces drop significantly and then level off. This glide length may correspond to the time it takes the animal to complete the landing maneuver and to land with all four limbs simultaneously. As a result of this maneuver, velocity just prior to landing is reduced by up to 60% (Byrnes et al. 2008). This ability to slow down prior to impact is a significant advantage to gliding animals over their leaping precursors. The ability to modulate aerodynamic forces and reduce velocity prior to impact allows gliding animals to travel long distances and reduce landing impulses compared to ballistic leaping (Fig. 4). The ability to avoid injury by implementing complex aerodynamic control at landing likely influenced the animals’ ability to travel longer distances by gliding.

Conclusions and future directions

This article has discussed how examining both the biomechanics and ecology of extant gliders can give insights into the evolution of gliding in mammals. Recent biomechanical studies of gliding (i.e., Yanoviak et al. 2005; Bishop 2006; Socha et al. 2010) have shown compelling evidence that gliding rarely occurs in the steady-state. Further, evidence from the locomotor ecology of gliding mammals suggests that they glide infrequently (i.e., Comport et al. 1996; Byrnes et al. 2011) and do not regularly maximize the distance covered. These results should shift our focus from the assumption that maximizing the distance of the glide is the most important metric of glide performance to understanding the aerodynamic control required to perform these aerial behaviors and for assessing the ecological conditions that drive their use. In trying to apply this new framework, we have discussed the many hypotheses that have been suggested for the origins of gliding. It is likely that none of these factors alone contributed to the evolution of gliding. However, given our current understanding, hypotheses relating to foraging have the most support thus far. Using comparative methods, we have shown that gliding is associated with low-quality diets. Further, using simple ballistic models we have shown that by producing aerodynamic forces gliding mammals can forage over greater distances with equivalent effort. Although gliding appears to be related to foraging behavior, it might not be because gliding is an economical form of locomotion. Instead it likely saves time in traveling and therefore allows for dependence on low-quality or possibly scattered resources or provides more time for acquiring resources. It is also likely that gliders, like many arboreal animals, are able to use high-velocity leaps to escape predators. However, leaping at high velocity increases risks of injury upon landing, especially when leaping long distances. The ability to modulate aerodynamic forces during gliding has allowed gliding animals to decouple take-off velocities and landing velocities and simultaneously improve range. Therefore, it is likely that the ability to modulate forces while airborne was a key transition in the evolution of controlled gliding.

To further understand the evolution of gliding, future research should address the following questions: (1) How do extant gliders use this form of locomotion? and (2) Why don’t more arboreal animals glide? In other words, what are the costs associated with a gliding versus an arboreal lifestyle? To answer the first question, it is now possible to get a detailed glimpse of the locomotor ecology of gliders by using a variety of newly available technologies. We have shown the utility of using animal-borne data loggers in our studies of both the ecology and biomechanics of free-ranging colugos. These, or similar, methods could be used on a variety of other gliding mammals to understand relationships between both biomechanical and ecological variables and other factors including body size or lineage. Further, understanding the differential costs of gliding and non-gliding lifestyles is critical to an understanding of why animals glide. For example, it has recently been shown that flying squirrels have greater metabolic costs while running on the level compared to non-gliding tree squirrels (Flaherty et al. 2010). More generally, we do not have a good understanding of the costs of arboreal locomotion compared to terrestrial locomotion. By understanding these costs, and combining them with detailed information about the movements of animals in the field, we can gain a better picture of the differential costs of a gliding lifestyle.

Funding

Summer Graduate Research Fellowship from the Department of Integrative Biology, University of California, Berkeley and from Wildlife Reserves Singapore (to G.B.); Royal Society International Travel Grant (to A.J.S. and G.B.).

Supplementary Data

Supplementary data are available at ICB online.

Acknowledgments

We would like to thank Robert Dudley and Steve Yanoviak for organizing this symposium and inviting us to participate. We would also like to thank the other participants of the symposium for their discussions of the topic. We also thank the editor and two anonymous reviewers for their comments that have improved the article.

References

Ando
M
Shiraishi
S
Gliding flight in the Japanese Giant Flying Squirrel Petaurista leucogenys
J Mamm Soc Japan
 , 
1993
, vol. 
18
 (pg. 
19
-
32
)
Archer
M
Archer
M
Clayton
G
Evolution of arid Australia and its consequences for vertebrates
Vertebrate zoogeography and evolution in Australia
 , 
1984
Perth
Hesperian Press
Baudinette
RV
Schmidt-Nielson
K
Energy cost of gliding flight in herring gulls
Nature
 , 
1974
, vol. 
248
 (pg. 
83
-
4
)
Bishop
KL
The relationship between 3-D kinematics and gliding performance in the southern flying squirrel, Glaucomys volans
J Exp Biol
 , 
2006
, vol. 
209
 (pg. 
689
-
701
)
Black
CC
Holarctic evolution and dispersal of squirrels (Rodentia: Sciuridae)
Evolutionary Biology
 , 
1963
, vol. 
6
 (pg. 
305
-
22
)
Bonser
RHC
Rayner
JMV
Measuring leg thrust forces in the common starling
J Exp Biol
 , 
1996
, vol. 
199
 (pg. 
435
-
9
)
Byrnes
G
Libby
T
Lim
NT-L
Spence
AJ
Gliding saves time but not energy in Malayan colugos
J Exp Biol
 , 
2011
Byrnes
G
Lim
NT-L
Spence
AJ
Take-off and landing kinetics of a free-ranging gliding mammal, the Malayan colugo (Galeopterus variegatus)
Proc R Soc B
 , 
2008
, vol. 
275
 (pg. 
1007
-
13
)
Byrnes
G
Lim
NT-L
Yeong
C
Spence
AJ
Sex differences in the locomotor ecology of a gliding mammal, the Malayan colugo (Galeopterus variegatus)
J Mamm
 , 
2011
, vol. 
92
 (pg. 
444
-
51
)
Channon
AJ
Crompton
RH
Günther
MM
D’Août
K
Vereecke
EE
The biomechanics of leaping in gibbons
Am J Phys Anthropol
 , 
2010
, vol. 
143
 (pg. 
403
-
16
)
Charnov
EL
Optimal foraging, the marginal value theorem
Theor Popul Biol
 , 
1976
, vol. 
9
 (pg. 
129
-
36
)
Comport
SS
Ward
SJ
Foley
WJ
Home ranges, time budgets and food-tree use in a high-density tropical population of greater gliders, Petauroides volans minor (Pseudocheiridae: Marsupialia)
Wildl Res
 , 
1996
, vol. 
21
 (pg. 
401
-
19
)
Crawford
RL
Baker
WW
Bats killed at a North Florida television tower: a 25–year record
J Mamm
 , 
1981
, vol. 
62
 (pg. 
651
-
2
)
Delciellos
AC
Vieira
MV
Stride lengths and frequencies of arboreal walking in seven species of didelphid marsupials
Acta Theriol
 , 
2007
, vol. 
52
 (pg. 
101
-
11
)
Demes
B
Fleagle
JG
Jungers
WL
Take-off and landing forces of leaping strepsirhine primates
J Hum Evol
 , 
1999
, vol. 
37
 (pg. 
279
-
92
)
Demes
B
Jungers
WL
Gross
TS
Fleagle
JG
Kinetics of leaping primates: influence of substrate orientation and compliance
Am J Phys Anthropol
 , 
1995
, vol. 
96
 (pg. 
419
-
29
)
Dial
R
Energetic savings and the body size distributions of gliding mammals
Evol Ecol Res
 , 
2003
, vol. 
5
 (pg. 
1151
-
62
)
Dial
R
Bloodworth
B
Lee
A
Boyne
P
Heys
J
The distribution of free space and its relation to canopy composition at six forest sites
For Sci
 , 
2004
, vol. 
50
 (pg. 
312
-
25
)
Dudley
R
Byrnes
G
Yanoviak
SP
Borrell
B
Brown
RM
McGuire
JA
Gliding and the functional origins of flight: biomechanical novelty or necessity?
Ann Rev Ecol and Syst
 , 
2007
, vol. 
38
 (pg. 
179
-
201
)
Dudley
R
DeVries
P
Tropical rain forest structure and the geographical distribution of gliding vertebrates
Biotropica
 , 
1990
, vol. 
22
 (pg. 
432
-
4
)
Dunbar
RIM
Time: a hidden constraint on the behavioural ecology of baboons
Behav Eco Sociobiol
 , 
1992
, vol. 
31
 (pg. 
35
-
49
)
Dunbar
RIM
Dunbar
P
Maternal time budgets of gelada baboons
Anim Behav
 , 
1988
, vol. 
36
 (pg. 
970
-
80
)
Emmons
LH
Gentry
AH
Tropical forest structure and the distribution of gliding and prehensile-tailed vertebrates
Am Nat
 , 
1983
, vol. 
121
 (pg. 
513
-
24
)
Essner
RL
Jr
Three-dimensional launch kinematics in leaping, parachuting and gliding squirrels
J Exp Biol
 , 
2002
, vol. 
205
 (pg. 
2469
-
77
)
Flaherty
EA
Ben-David
M
Smith
WP
Quadrupedal locomotor performance in two species of arboreal squirrels: predicting energy savings of gliding
J Comp Physiol B
 , 
2010
, vol. 
180
 (pg. 
1067
-
78
)
Flaherty
EA
Scheibe
JS
Goldingay
R
Locomotor performance in the squirrel glider, Petaurus norfolcensis, and the sugar glider Petaurus breviceps
Aust Mamm
 , 
2008
, vol. 
30
 (pg. 
25
-
35
)
Foley
WJ
Kehl
JC
Nagy
KA
Kaplan
IR
Borsboom
AC
Energy and water metabolism in free-living greater gliders, Petauroides volans
Aust J Zool
 , 
1990
, vol. 
38
 (pg. 
1
-
9
)
Full
RJ
Tullis
A
Energetics of ascent: insects on inclines
J Exp Biol
 , 
1990
, vol. 
149
 (pg. 
307
-
17
)
Garber
PA
Blomquist
GE
Anzenberger
G
Kinetmatic analysis of trunk-to-trunk leaping in Callimic goeldii
Int J Primatol
 , 
2005
, vol. 
26
 (pg. 
223
-
40
)
Goldingay
R
Time budget and related aspects of the foraging behavior of the yellow-bellied glider, Petaurus australis
Aus Wildl Res
 , 
1989
, vol. 
16
 (pg. 
105
-
12
)
Goldingay
RL
The foraging behaviour of a nectar-feeding marsupial, Petaurus australis
Oecologia
 , 
1990
, vol. 
85
 (pg. 
191
-
9
)
Goldingay
RL
Goldingay
RL
Scheibe
JS
Gliding mammals of the world: diversity and ecological requirements
Biology of gliding mammals
 , 
2000
Germany, Furth
Filander
(pg. 
9
-
44
)
Hanna
JB
Schmitt
D
Griffin
TM
The energetic cost of climbing in primates
Science
 , 
2008
, vol. 
320
 pg. 
898
 
Hingston
RWG
The attitudes and movements of the large red flying squirrel—Petaurista inornatus
J Bombay Nat Hist Soc
 , 
1914
, vol. 
23
 (pg. 
344
-
50
)
Holmes
DJ
Austad
SN
Fly now, die later: life-history correlates of gliding and flying mammals
J Mamm
 , 
1994
, vol. 
75
 (pg. 
224
-
6
)
Howland
HC
Optimal strategies for predator avoidance: the relative importance of speed and manoeverability
J Theor Biol
 , 
1974
, vol. 
47
 (pg. 
333
-
50
)
Inger
RF
Morphological and ecological variation in the flying lizards (genus Draco)
Fieldiana Zoology, new series
 , 
1983
, vol. 
18
 (pg. 
1
-
35
)
Jackson
SM
Glide angle in the genus Petaurus and a review of gliding in mammals
Mammal Rev
 , 
1999
, vol. 
30
 (pg. 
9
-
30
)
Jurmain
R
Skeletal evidence of trauma in the African apes, with special reference to the Gombe chimpanzees
Primates
 , 
1997
, vol. 
38
 (pg. 
1
-
14
)
Kavanagh
RP
Lambert
MJ
Food selection by the greater glider, Petauroides volans: is foliar nitrogen a determinant of habitat quality?
Aust Wildl Res
 , 
1990
, vol. 
17
 (pg. 
285
-
9
)
Klem
DJ
Bird injuries, cause of death, and recuperation from collisions with windows
J Field Ornithol
 , 
1990
, vol. 
61
 (pg. 
115
-
9
)
Koenig
WD
Mumme
RL
Population ecology of the cooperatively breeding acorn Woodpecker
 , 
1987
Princeton
Princeton University Press
Legreneur
P
Thévenet
F-R
Libourel
P-A
Monteil
KM
Montuelle
S
Pouydebat
E
Bels
V
Hindlimb interarticular coordinations in Microcebus murinus in maximal leaping
J Exp Biol
 , 
2010
, vol. 
213
 (pg. 
1320
-
7
)
Maddison
WP
A method for testing the correlated evolution of two binary characters: are gains and losses concentrated on certain branches of a phylogenetic tree?
Evolution
 , 
1990
, vol. 
44
 (pg. 
539
-
57
)
Maddison
WP
Maddison
DR
MacClade 4.0 PPC
 , 
2000
Sinauer Associates, Sunderland Massachussetts
McGuire
JA
Dudley
R
The cost of living large: comparative gliding performance in flying lizards (Agamidae: Draco)
Am Nat
 , 
2005
, vol. 
166
 (pg. 
93
-
106
)
Mein
P
Les sciuropteres (Mammalia, Rodentia) Neogenes D’Europe occidentale
Geobios
 , 
1970
, vol. 
3
 (pg. 
7
-
77
)
Mein
P
Romaggi
J-P
Un gliridé (Mammalia, Rodentia) planeur dans le Miocène Supérieur de l’Ardèche: une adaptation non retrouvée dans la nature actuelle
Geobios
 , 
1991
, vol. 
13
 (pg. 
45
-
50
)
Meng
J
Hu
Y
Wang
Y
Wang
X
Li
C
A Mesozoic gliding mammal from northeastern China
Nature
 , 
2006
, vol. 
444
 (pg. 
889
-
93
)
Mercer
JM
Roth
VL
The effects of Cenozoic global change on squirrel phylogeny
Science
 , 
2003
, vol. 
299
 (pg. 
1568
-
72
)
Montgelard
C
Bentz
S
Tirard
C
Verneau
O
Catzeflis
FM
Molecular systematics of Sciurognathi (Rodentia): the mitochondrial cytochrome b and 12S rRNA genes support the Anomaluroidea (Pedetidae and Anomaluridae)
Mol Phylogenet Evol
 , 
2002
, vol. 
22
 (pg. 
220
-
33
)
Nachtigall
W
Gleitflug des flugbeutlers Petaurus breviceps papuanus II. Filmanalysen zur einstellung von Gleitbahn und rumpf sowie zur steuerung des gleitflugs
J Comp Phys A
 , 
1979
, vol. 
133
 (pg. 
89
-
95
)
Nagy
KA
Field metabolic rate and body size
J Exp Biol
 , 
2005
, vol. 
208
 (pg. 
1621
-
5
)
Norberg
RA
Optimal locomotion modes of foraging birds in trees
Ibis
 , 
1983
, vol. 
125
 (pg. 
172
-
80
)
Nowak
RM
Walker's mammals of the world
1999
6th ed
Baltimore
Johns Hopkins University Press
Oliver
JA
“Gliding” in amphibians and reptiles, with a remark on an arboreal adaptation in the lizard, Anolis carolinensis carolinensis Voigt
Am Nat
 , 
1951
, vol. 
85
 (pg. 
171
-
6
)
Osborne
MJ
Chrisidis
L
Norman
JA
Molecular phylogenetics of the Diprodontia (kangaroos, wombats, koala, possums, and allies)
Mol Phylogenet Evol
 , 
2002
, vol. 
25
 (pg. 
219
-
28
)
Paskins
KE
Bowyer
A
Megill
WM
Scheibe
JS
Take-off and landing forces and the evolution of controlled gliding in northern flying squirrels Glaucomys sabrinus
J Exp Biol
 , 
2007
, vol. 
210
 (pg. 
1413
-
23
)
Pomeroy
D
Why fly? The possible benefits for lower mortality
Biol J Linn Soc
 , 
1990
, vol. 
40
 (pg. 
53
-
65
)
Poux
C
Douzery
EJP
Primate phylogeny, evolutionary rate variations, and divergence times: a contribution from the nuclear gene IRBP
Am J Phys Anthro
 , 
2004
, vol. 
124
 (pg. 
1
-
16
)
Sapir
N
Wikelski
M
McCue
MD
Pinshow
B
Nathan
R
Flight modes in migrating European bee-eaters: heart rate may indicate low metabolic rate during soaring and gliding
PLoS ONE
 , 
2010
, vol. 
5
 pg. 
e13956
 
Scheibe
JS
Paskins
KE
Ferdous
S
Birdsill
D
Kinematics and functional morphology of leaping, landing, and branch use in Glaucomys sabrinus
J Mamm
 , 
2007
, vol. 
88
 (pg. 
850
-
61
)
Scheibe
JS
Robins
JH
Steele
MA
Merritt
JF
Zegers
DA
Morphological and performance attributes of gliding mammals
Ecology and evolutionary biology of tree squirrels
 , 
1998
 
Special Publication Vol. 6. Martinsville, VA: Virginia Museum of Natural History, p. 131–44
Scheibe
JS
Smith
WP
Bassham
J
Magness
D
Locomotor performance and cost of transport in the northern flying squirrel Glaucomys sabrinus
Acta Therio
 , 
2006
, vol. 
51
 (pg. 
169
-
78
)
Schlesinger
WH
Knops
JMH
Nash
TH
Arboreal sprint failure: lizardfall in a California oak woodland
Ecology
 , 
1993
, vol. 
74
 (pg. 
2465
-
7
)
Scholey
K
Nachtigall
W
The climbing and gliding locomotion of the giant red flying squirrel Petauista petaurista (Sciuridae)
Bat Flight—Fledermausflug
 , 
1986
Stuttgart, Germany
Gustav-Fischer
(pg. 
187
-
204
)
Schultz
AH
Notes on the diseases and healed fractures of wild apes and their bearing on the antiquity of pathological conditions in man
Bull Hist Med
 , 
1939
, vol. 
7
 (pg. 
571
-
82
)
Shattuck
MR
Williams
SA
Arboreality has allowed for the evolution of increased longevity in mammals
Proc Nat Acad Sci
 , 
2010
, vol. 
107
 (pg. 
4635
-
9
)
Socha
JJ
Miklasz
K
Jafari
F
Vlachos
PP
Non-equilibrium trajectory dynamics and the kinematics of gliding in a flying snake
Bioinsp Biomim
 , 
2010
, vol. 
5
 (pg. 
1
-
12
)
Springer
MS
Stanhope
MJ
Madsen
O
de Jong
WW
Molecules consolidate the placental mammal tree
Trends Ecol Evol
 , 
2004
, vol. 
19
 (pg. 
430
-
8
)
Stafford
BJ
Thorington
RW
Jr
Kawamichi
T
Gliding behavior of Japanese giant flying squirrels (Petaurista leucogenys)
J Mamm
 , 
2002
, vol. 
83
 (pg. 
553
-
62
)
Stapp
P
Can predation explain life-history strategies in mammalian gliders?
J Mamm
 , 
1994
, vol. 
75
 (pg. 
227
-
8
)
Stern
M
Goldstone
R
Red colobus as prey: the leaping habits of five sympatric old world monkeys
Folia Primatol
 , 
2005
, vol. 
76
 (pg. 
100
-
12
)
Stevens
NJ
The effect of branch diameter on primate gait sequence pattern
Am J Primatol
 , 
2008
, vol. 
70
 (pg. 
356
-
62
)
Storch
G
Engesser
B
Wuttke
M
Oldest fossil record of gliding in rodents
Nature
 , 
1996
, vol. 
379
 (pg. 
439
-
41
)
Taylor
CR
Caldwell
SL
Rowntree
VJ
Running up and down hills: some consequences of size
Science
 , 
1972
, vol. 
178
 (pg. 
1096
-
7
)
Taylor
CR
Heglund
NC
Maloiy
GMO
Energetics and mechanics of terrestrial locomotion. I. Metabolic energy consumption as a function of speed and body size in birds and mammals
J Exp Biol
 , 
1982
, vol. 
97
 (pg. 
1
-
21
)
Teeling
EC
Madsen
O
Van Den Bussche
RA
de Jong
WW
Stanhope
MJ
Springer
MS
Microbat paraphyly and the convergent evolution of a key innovation in Old World rhinolophoid microbats
PNAS
 , 
2002
, vol. 
99
 (pg. 
1431
-
6
)
Thorington
RW
Jr
Pitassy
D
Jansa
SA
Phylogenies of flying squirrels (Pteromyinae)
J Mamm Evol
 , 
2002
, vol. 
9
 (pg. 
99
-
135
)
Thorington
RW
Jr
Schennum
CE
Pappas
LA
Pitassy
D
The difficulties of indentifying flying squirrels (Sciuridae: Pteromyini) in the fossil record
J Vert Paleo
 , 
2005
, vol. 
25
 (pg. 
950
-
61
)
Vernes
K
Gliding performance of the northern flying squirrel (Glaucomys sabrinus) in a mature mixed forest of eastern Canada
J Mamm
 , 
2001
, vol. 
82
 (pg. 
1026
-
33
)
Wasserman
MD
Chapman
CA
Determinants of colobine monkey abundance: the importance of food energy, protein and fibre content
J Animal Ecol
 , 
2003
, vol. 
72
 (pg. 
650
-
9
)
Wilson
DE
Reeder
DM
Mammal species of the world, a taxonomic and geographic reference
 , 
1993
2nd ed
Washington, DC
Smithsonian Institution Press
Wunder
BA
Morrison
PR
Red squirrel metabolism during incline running
Comp Biochem Physiol A
 , 
1974
, vol. 
48
 (pg. 
153
-
61
)
Yanoviak
SP
Dudley
R
Kaspari
M
Directed aerial descent in arboreal ants
Nature
 , 
2005
, vol. 
433
 (pg. 
624
-
6
)

Author notes

From the symposium “The Biomechanics and Behavior of Gliding Flight” presented at the annual meeting of the Society for Integrative and Comparative Biology, January 3–7, 2011, at Salt Lake City, Utah.

Supplementary data