An integrative theory on the origin of bat flight

Abstract

Summary and Conclusions: A theory on the origin of bat flight has perhaps greater potential than any other theme in bat biology for integrating results from diverse fields of research, including phylogenetic paleontology, molecular control of organogenesis, gross embryology and development, functional anatomy, mechanics of locomotion, and roosting and trophic ecology. This limited review left aside several aspects that deserve a great deal of attention, including: physiological impact of evolving a large, hairless flight membrane; impact of genomic approaches in understanding the evolution of gene systems that allow bat metabolic performance (e.g., Shen et al., 2010); implications of specific muscle physiology of bats and their ontogeny (e.g., Hermanson, 1998, 2000) in the transition to flight; correspondence between ontogeny of flight (e.g., Elangovan et al., 2004, 2007), proposed adaptive landscapes (Adams, 2000; 2008), and the sequence of morphological changes implied in current phylogenies; the possible connection between rapid growth of long bones (e.g., Farnum et al., 2008a, b) and precociality (e.g., Jepsen, 1970) with morphological changes in evolutionary time; a thorough analysis of aerodynamic models in light of previously unknown morphologies now available from new fossils; and a full-scale allometric analysis of structural change (e.g., of wing development; Kunz and Robson, 1995; Wyant and Adams, 2007) applied to interspecific variation in the phylogeny of bats.However, as summarized below, a substantial body of knowledge from numerous fields of bat biology exists, and the current challenge consists of elaborating a comprehensive synthesis of all such knowledge, perhaps starting from the oldest fossils, the Early Eocene bats. These bats exhibited a mosaic of derived (e.g., greatly elongated, foldable hand wing) versus primitive wing features (e.g., clawed digits with rigid bony phalanges), as well as embryonic retentions (e.g., hand webbing, supinated hand) versus divergences from the mammalian developmental standard (e.g., elongated digits). These character states changed rapidly during the course of early bat evolution, but left morphological signatures in fossils that allow an increasingly finer level of detail in reconstructing the transition experienced by bats as the only volant mammals. In this way, fossils reveal not only the morphology of extinct organisms but also the ontogenetic stages attained in diverse developing systems. For instance, the claws of Onychonycteris show that this taxon underwent the standard mammalian process of ungual phalanx development, whereas this process is arrested to varying degrees in more derived bats. In general, bats seem particularly useful in showing how paleontological and developmental evidence can be combined in ways that illuminate each other and the evolution of a lineage. Evolution of the hand wing is key in the transition to flapping flight. In this chapter, a cumulative sequence of four major events is proposed: interdigital membrane retention, digital elongation, wing folding, and truncation of distal phalanx development. In terms of developmental control, the first two processes, which are linked in bats, are understood at the molecular level (e.g., Hockman et al., 2008) and indicate that specific changes in spatiotemporal expression of regulatory genes played a major role in shaping the bat hand wing (as in derived limbs of other mammals; Honeycutt, 2008). The extraordinary rate of ontogenetic digital elongation reported for bats (Farnum et al., 2008a, b) may give clues about the question of Cooper and Tabin (2008) on how fast, and how intergraded, was the evolution of a novelty such as the bat hand wing. Pattern of expression of specific receptor proteins and signaling pathways may help explain the structure of C-MC joints, hence the origin of the complex wing folding mechanism of bats. In turn, a simple heterochronic process (e.g., deceleration) may account for the truncation of phalanx development. In spite of the evolution of the hand wing and flapping flight, the center of lift of the bat wing is in the plagiopatagium (Swartz et al., 2005) as in all gliders (Jackson and Schouten, in press). The plagiopatagium and the other patagial tracts likely evolved before the handwing and independently of each other. Each tract develops as a separate entity representing a distinct dermal appendage. These appendages appear to be already fully evolved in known fossils, so character states of the patagia, as well as hind-limb and some wing traits (e.g., C-MC-folding, faceting of scapular fossae), indicate that primitive bats were functionally as advanced as some of their modern descendants. However, primitive bats also exhibited functionally poor character states in wing proportions (e.g., relatively low AR), muscle attachments (e.g., ribs without laminae), and traits probably affecting aerodynamic performance (e.g., clawed, rigid finger tips). Fossils also show that, as compared to the wing, the hind-limb was more advanced toward the character states exhibited by modern bats, so bat precursors likely used hind-limb suspension before being fully capable of flapping flight (e.g., reconstructions in Smith, 1977). In a gliding, or Darwinian, model of bat flight origin, the hypothetical transition sequence is from climbing and leaping to gliding to flapping. The first step is unproblematic as there are living representatives documenting the relative ease with which gliding has been achieved in independent clades of arboreal mammals. The transition to flapping flight has been questioned mainly because bats only seldom glide and gliders do not flap their aerofoils and differ in other aspects like their use of high angles of attack as compared to flying bats, all of which points to a wide performance gap in-between (see Bishop, 2008). However, bats are capable of gliding and, at low speed, they flap their wings at angles of attack typical of gliders. Therefore the transition envisioned by Scholey (1986a), Norberg (1986), Rayner (1986), and others, driven by improvements in cost of transport, is much less difficult than currently thought as long as 1. a performance gradient of increasing AR and decreasing WL is sustained in the evolving wing (see above); and 2. low-amplitude flapping is introduced later in the process, when the wing reaches an intermediate AR and a low WL. Interestingly, the onset of flapping in developing bats occurs when maximum AR and minimum WL combine favorably (Adams, 2008; Elangovan et al., 2004, 2007), so there could be a ontogenetic component correlated with evolving powered flight. AR and WL are oversimplified characters which may not describe bat flight with the necessary functional accuracy (Riskin et al., 2010; Swartz et al., 2005); still, these parameters allow depiction of a coarse-grained gradient of possible changes during the evolution of bat flight. The morphology of bats likely posed problems at the ecophysiological level, particularly the combination of being small-sized homeotherms with extensive, highly vascularized, hairless membranes. While these membranes may be greatly advantageous to dispose of waste heat during the nighttime (Pennycuick, 1986; Neuweiler, 2000), by the same principle of heat exchange bat precursors may have experienced problems of excessive heat loss during resting. Hind-limb suspension using a tendon-locking mechanism may have allowed bats to exploit tree cavities and take advantage of this energy-saving roost, as they do today in forests. Subsequent evolution of wing folding greatly improved control of heat loss at the roost, with additional advantages (e.g., wing protection). Other, somewhat contentious ecological aspects of bat evolution, particularly activity patterns and diet, are resolved by recourse to phylogenetic inertia given that nocturnality and insectivory are plesiomorphic in mammals so no change is required at the root of the chiropteran subtree.Hypothetical bat precursors are here envisioned as relatively small mammals with a forest-glider body plan but with naked membranes and some behavioral character states of manifest evolutionary inertia (insectivory, nocturnality), initially evolving hindlimb suspension and a tree-cavity dwelling habit, next rapidly evolving size (phyletic nanism), an elongated handwing, and a wing folding mechanism, initiating flapping at some low-speed situation such as breaking before landing, and continuing in a path of increasing AR and wing refinement (e.g., reduction of claws) that allowed an explosive adaptive radiation and invasion of all continents by the Early Eocene. Finally, if the gliding model is correct, the performance of just any gliding mammal would be enhanced by, for instance, wings evolving toward high AR. So why only bats among mammals developed flight when there are many independent gliding lineages, each of which represents a potential “source” of bat-like descendants? The answer probably is historical and lays on the nature of changes required to elaborate the bat hand wing. Cretekos et al. (2008) suggested that evolutionary changes like fore-limb elongation may require many changes in regulatory genes with minor or transient, but cumulative effect, which would produce sudden, large-scale modifications once combined in a bat ancestor. In addition, some of the regulatory effects that generate an elongated and webbed hand are intrinsically antagonistic if not for the rather elaborate patterning of expression in the handplate. For instance, an up-regulated Bmp2 is required to elongate the digits but its expression causes cell death in the interdigital tissue; so specific patterning and additional regulators are required to produce elongated digits and dactilopatagium (see Hockman et al., 2009); notice that a possible regulatory mechanism becomes evident for the colugo, with webbed autopods but short digits (e.g., similar to webbed aquatic birds; see Weatherbee et al., 2006). These effects, to be compatible overall, are orchestrated in ways that just do not favor chances of replication in other lineages, even those lineages that could be selected by these changes (e.g., other gliders). Moreover, those extraordinary transformations were not heavily penalized by selection probably because they did not happen in isolation; it is possible that pre-existing habits (e.g., hind-limb suspension and roosting in energy-saving cavities) and additional and timely adaptations (e.g., C-MC wing folding) enabled bat precursors to endure a seemingly difficult transition, ultimately crossing over from gliding to powered flight.