Which came first as bats evolved — flight or echolocation? Newly described fossils favour the flight-first hypothesis. But these creatures may have been otherwise equipped for flying at night.
A long-standing debate about the processes that led to the evolution of modern bats takes a new twist with the discovery of remarkable fossil bats recovered from the Green River formation in Wyoming. The fossils, which constitute a new genus and species, are described by Simmons et al. on page 818 of this issue1. Phylogenetic analysis and comparison with other fossil bats recovered from the same formation, and from the Messel formation in Germany, indicate that this is the most ancient species of bat yet discovered.
The problem of understanding bat evolution dates back at least to Charles Darwin, who in The Origin of Species enumerated a list of difficulties he saw with the theory of evolution by natural selection. The example often discussed is the origin of the eye. But Darwin also mentioned the vexed issue of how bats had arisen from terrestrial ancestors. The discovery of echolocation in bats about 50 years ago2 added an additional feature to the conundrum of the early evolution of bats. This currently boils down to one big question: which came first, echolocation or flight3,4?
For a long time, 'echolocation first' held sway. Ancestral 'pre-bats' were hypothesized to have been small terrestrial or arboreal echolocators that detected passing insects using their echolocation and snatched them from the air4. This favoured the extension of the arms and digits to facilitate prey capture, perhaps with webbing between the digits. Eventually, these animals started leaping out to capture insects, using their echolocation to guide them to a landing spot and their extended arms and digits as an aerofoil. From this point they started hunting from perches (known as perch hunting) and eventually developed fully powered flight (called aerial hawking; Fig. 1).
Supporters of the echolocation-first hypothesis pointed to the existence of terrestrial animals, such as certain shrews, that have rudimentary echolocation systems; and to the fact that the most primitive extant bats often use perch hunting, and lack a feature known as the calcar, which is also absent in the most ancient fossil bats. (The calcar is a cartilaginous spur projecting from the base of the lower limb and running along the edge of the membrane between the hind limbs and tail.) Moreover, the idea that bats might have evolved the ability to fly before they could orient themselves in darkness was seen as highly unlikely.
However, around the end of the 1980s, evidence accumulated, including work from my own group, that favoured the 'flight-first' hypothesis. One paper5 showed that, for a bat hanging at rest, echolocation is extremely energetically costly. This high cost probably explains why no terrestrial mammals have evolved full-blown echolocation systems such as those used by bats. However, a second paper6 showed that when a bat takes flight these costs disappear. This is because of a remarkable coupling of the beating of the wings with the ventilation of the lungs and production of the echolocation pulses7. When a bat hangs stationary and echolocates, it must contract its muscles specifically to generate a forceful expiratory burst, and this is where the large costs come from. When a bat is flying, it is already contracting these muscles, so in effect echolocation when flying is free (or at least substantially cheaper).
But what about the problem of bats flying in darkness before they could orient themselves? A hypothesis I favour8 is that the earliest ancestors of bats may have been diurnal, and had visual means of orientation — but were perhaps forced to become nocturnal by the appearance of avian predators, shortly after the dinosaurs became extinct around 65 million years ago. Some then evolved echolocation, whereas others became nocturnal vision specialists.
Until the discovery of the specimens reported by Simmons et al.1, the fossil record has been rather unhelpful in resolving these issues: the earliest-known bats, which have been recovered from Eocene deposits around 50 million years old, are fully formed bats very similar to extant ones9,10. It has been possible to show that these bats were all already capable of echolocation by examining the size of the cochleae in their ears; cochleae are massively enlarged in echolocators. Previously described fossil bats were all already capable of both flight and echolocation11,12.
The bat described by Simmons et al.1 is represented by two fossils dating to about 52.5 million years ago; one is shown on the cover of this issue. The bat's wing morphology is very similar to that of extant species, except that it has claws on its digits. But in all other respects this is clearly a bat capable of powered flight. Unlike other primitive fossil bats, this species also has a calcar — indicating that the absence of a calcar and perhaps therefore perch hunting are not ancestral traits. But the real insight provided by this fossil is the spectacular finding that it does not have enlarged cochleae. By inference, therefore, it was not capable of echolocation, providing the first direct evidence supporting the flight-first hypothesis. Examination of the bat's limb proportions suggests that it was probably arboreal, as has generally been assumed by proponents of both echolocation-first and flight-first hypotheses.
A remaining question is whether this bat was nocturnal or diurnal. Examination of the size of the eye sockets might help, as nocturnal non-echolocating animals generally have enlarged eyes and so enlarged eye sockets. Unfortunately, the two specimens described by Simmons et al. cannot answer this question, because their upper skulls are crushed and their eye sockets cannot be reconstructed. Perhaps that was too much to hope for. As it is, these outstanding fossils considerably advance our understanding of bat evolution.
Simmons, N. B., Seymour, K. L., Habersetzer, J. & Gunnell, G. F. Nature 451, 818–821 (2008).
Griffin, D. R. Listening in the Dark: The Acoustic Orientation of Bats and Men (Yale Univ. Press, New Haven, CT, 1958).
Teeling, E. C. et al. Nature 403, 188–192 (2000).
Jones, G. & Teeling, E. C. Trends Ecol. Evol. 21, 149–156 (2006).
Speakman, J. R., Anderson, M. E. & Racey, P. A. J. Comp. Physiol. A 165, 679–685 (1989).
Speakman, J. R. & Racey, P. A. Nature 350, 421–423 (1991).
Suthers, R. A., Thomas, S. P. & Suthers, B. J. J. Exp. Biol. 56, 37–48 (1972).
Speakman, J. R. Mamm. Rev. 31, 111–130 (2001).
Jepsen, G. L. Science 154, 1333–1339 (1966).
Gunnell, G. F. & Simmons, N. B. J. Mamm. Evol. 12, 209–246 (2005).
Habersetzer, J. & Storch, G. Naturwissenschaften 79, 462–466 (1992).
Novacek, M. J. Nature 315, 140–141 (1985).
About this article
Nature Communications (2018)