Bats descended from small, tree-dwelling mammals that lived during the late Paleocene or early Eocene, roughly 55 to 60 million years ago. The exact ancestor remains unknown because the fossil record has a frustrating gap right where the transition from ground-dwelling or gliding mammal to powered flyer should be. By the time recognizable bat skeletons show up in rock, around 52 to 53 million years ago, they already have fully developed wings and, in most cases, the inner-ear anatomy for echolocation. What happened in between is one of the more fascinating puzzles in mammalian evolution, and researchers have been filling it in with molecular genetics, aerodynamic modeling, and developmental biology.
The Frustrating Fossil Gap
The oldest known bat skeletons come from early Eocene deposits in Wyoming and date to roughly 52 million years ago. Species like Icaronycteris and the recently described Icaronycteris gunnelli are preserved as articulated skeletons complete with skulls, and they look unmistakably like modern bats: elongated finger bones, wing membranes, the works.1PubMed Central. The oldest known bat skeletons and their implications for Eocene chiropteran diversification The earliest complete bats from the Eocene, spanning about 49 to 53 million years ago, were already capable of flapping flight.2Mammal Review. The evolution of flight and echolocation in bats: another leap in the dark
That means the actual transition from non-flying mammal to bat happened earlier, probably sometime in the late Paleocene, but no fossil from that transitional window has been found. Compare this to whales, where paleontologists have a beautiful sequence from walking land animals to fully aquatic swimmers. Bats have no such sequence. The best guess, based on molecular clock estimates, is that the bat lineage diverged from other mammals around 60 to 65 million years ago, not long after the mass extinction that ended the age of dinosaurs. The ecological vacuum left by flying reptiles like pterosaurs, combined with an explosion of insect diversity, likely created exactly the opportunity a small nocturnal mammal needed to take to the air.
From Trees Down, Not Ground Up
How does a small mammal go from climbing branches to flying? Researchers have debated two broad scenarios for decades: “ground up,” where running and leaping gradually became flight, and “trees down,” where gliding from tree to tree transitioned into powered flapping. The evidence increasingly favors trees down.
A 2024 study compared limb proportions across bats, gliding mammals like flying squirrels and colugos, and non-gliding tree-dwellers. The results showed that gliders have forelimb proportions intermediate between non-gliding arboreal mammals and bats, with hindlimbs that are narrower but not longer, again falling between the two groups. This supports a pathway where a glider-like body plan preceded the bat-like body plan adapted for powered flight.3PubMed Central. Gliding toward an understanding of the origin of flight in bats
Another piece of the puzzle comes from atmospheric modeling. During the early Eocene, Earth’s atmosphere was denser than it is today due to higher carbon dioxide levels and warmer temperatures. A 2024 study modeled the flight performance of Onychonycteris finneyi, one of the most primitive known bats, which had relatively short wings that might not have supported powered flight in today’s atmosphere. In the thicker Eocene air, though, a gliding performance gradient emerged: as wing shape became more elongated, sink rate decreased until a point where available muscle power was enough to sustain level flapping flight. This strongly supports the idea that early bats glided first and that denser air made the transition to powered flight much easier than it would be now.4PubMed Central. Palaeoatmosphere facilitates a gliding transition to powered flight in the Eocene bat, Onychonycteris finneyi
How a Hand Becomes a Wing
A bat wing is, structurally, a hand. Four massively elongated fingers support a thin membrane of skin. The thumb remains free as a small claw used for gripping. Turning a normal mammalian hand into a wing required dramatic changes in how finger bones grow during embryonic development, and researchers have identified some of the key molecular players.
The most important finding involves a family of signaling molecules called bone morphogenetic proteins, specifically BMP2. In developing bat embryos, the forelimb digits show much higher rates of cartilage cell proliferation and growth compared to the hind limb digits or to a mouse’s digits. BMP2 can stimulate this cartilage growth and increase digit length in bat forelimb embryos, and its expression is ramped up specifically in forelimb digits. This suggests that an evolutionary up-regulation of the BMP pathway is one of the central mechanisms behind how bat fingers became so dramatically long.5PubMed Central. Development of bat flight: morphologic and molecular evolution of bat wing digits
BMP2 is not the whole story, though. Gene expression profiling of bat digits identified at least 14 genes strongly associated with the difference between short digits and long wing digits. These include members of the BMP signaling pathway, several Hox genes involved in limb patterning, and genes like Twist1 that are associated with bone and cartilage development in other mammals.6PubMed Central. Digital gene expression tag profiling of bat digits provides robust candidates contributing to wing formation The wing did not evolve through a single mutation. It required coordinated changes across a network of developmental genes, all pushing finger bones to grow longer while the membrane between them expanded.
Did Flight or Echolocation Come First?
This is one of the longest-running debates in bat biology. Most bats echolocate, emitting ultrasonic calls and listening to the returning echoes to navigate and hunt in total darkness. But did early bats develop echolocation before they could fly, after they could fly, or at the same time? The debate has been surprisingly hard to resolve because, as noted earlier, the oldest known bat fossils already show evidence of both abilities.2Mammal Review. The evolution of flight and echolocation in bats: another leap in the dark
The strongest piece of evidence for “flight first” comes from Onychonycteris finneyi, the primitive Eocene bat mentioned above. Its wing bones indicate it could fly, but its inner ear anatomy does not show the specializations associated with echolocation. If that interpretation is correct, the earliest bats could get around in the air before they could echolocate, and echolocation was added later as an adaptation for catching insects at night.
Inner ear morphology across the full bat family tree provides another angle. Ancestral reconstructions of cochlear structure suggest that the common ancestor of all living bats was probably capable of hearing ultrasonic frequencies, which is a prerequisite for echolocation but not proof that full echolocation was present.7PubMed Central. Evolutionary origins of ultrasonic hearing and laryngeal echolocation in bats inferred from morphological analyses of the inner ear It is possible that primitive ultrasonic hearing evolved early and was later refined into the sophisticated sonar system seen in modern echolocating species.
Recent work on the spiral ganglion, the bundle of nerve cells connecting the ear to the brain, has added nuance. One of the two major bat suborders, Yangochiroptera, has a highly unusual ganglion structure that allows more neurons, denser nerve connections, and more precise sound processing. This specialized anatomy differs from both the other bat suborder and from non-bat mammals, suggesting that the two bat lineages diverged significantly in how they refined their hearing hardware.8PubMed. Evolution of inner ear neuroanatomy of bats and implications for echolocation
Two Suborders, Not the Ones You Might Expect
For most of the twentieth century, bats were divided into “megabats” (the large, fruit-eating Old World flying foxes) and “microbats” (everything else, mostly small and echolocating). This split felt intuitive: big bats that see well versus small bats that echolocate. But molecular evidence upended it.
Genetic studies now strongly support a different classification into two suborders called Yinpterochiroptera and Yangochiroptera. In this arrangement, Old World fruit bats are grouped with certain echolocating families like horseshoe bats, while the remaining echolocating families form Yangochiroptera.9PubMed Central. Phylogenomic analyses of bat subordinal relationships based on transcriptome data This means echolocation either evolved independently in different bat lineages or, more provocatively, it evolved once in the common ancestor and was then lost in fruit bats. Either way, the old megabat-microbat divide was misleading.
The reclassification also reshuffles which traits are considered ancestral and which evolved convergently. Behaviors like roosting in foliage and building tent-like shelters from leaves evolved independently in both suborders, meaning these habits arose multiple times rather than being inherited from a shared ancestor.10Current Zoology. Convergences in the diversification of bats
What Bats Traded Away for Sonar
Echolocation is metabolically expensive and neurologically demanding. As some bat lineages invested more heavily in hearing, they appear to have dialed down other senses. The most striking trade-off involves vision. Old World fruit bats, which do not echolocate using their larynx, have excellent eyesight and navigate largely by sight. Echolocating bats, by contrast, rely primarily on sonar for navigation.11Oxford Academic. Blind as Bats: Bat Echolocation Study Reveals Key Evolutionary Trade-Offs with Other Senses
The genetic fingerprints of this trade-off are visible in the opsin genes responsible for color vision. An analysis of the SWS1 opsin gene, which enables ultraviolet light detection, found that it has been inactivated (turned into a non-functional “pseudogene”) in several bat lineages. Interestingly, this loss of UV vision does not map neatly onto echolocation ability. At least one echolocating species retains a functional SWS1 gene. Instead, the loss of UV vision correlates more strongly with cave-roosting habits, particularly among fruit bats. Species that spend their lives in dark caves have less evolutionary pressure to maintain color vision.12PubMed Central. As Blind as a Bat? Opsin Phylogenetics Illuminates the Evolution of Color Vision in Bats The familiar phrase “blind as a bat” turns out to be wrong on multiple levels: some bats see quite well, and the ones with reduced vision lost it because of where they live, not simply because they echolocate.
Convergent Evolution With Dolphins
One of the most surprising discoveries in bat research came from comparing hearing genes across mammals. Echolocating bats and toothed whales like dolphins independently evolved remarkably similar changes in the Prestin gene, which encodes a protein in the outer hair cells of the inner ear critical for processing high-frequency sound. When researchers built a family tree based on Prestin protein sequences, bottlenose dolphins clustered with echolocating bats rather than with their closer mammalian relatives.13PubMed Central. The hearing gene Prestin unites echolocating bats and whales A separate study confirmed that these identical amino acid changes in Prestin were driven by natural selection, not chance, and that dolphins show the same convergent pattern as echolocating bats.14Current Biology. Convergent sequence evolution between echolocating bats and dolphins
This is a striking example of convergent molecular evolution: two groups of mammals, separated by tens of millions of years and living in completely different environments (air versus water), arrived at the same molecular solution to the same problem. It suggests that when natural selection pushes hard enough on a particular function, the number of workable genetic solutions may be surprisingly limited.
Flight Reshaped the Entire Genome
Powered flight is the most metabolically intense form of locomotion in the animal kingdom. A bat in flight has a metabolic rate many times higher than a similar-sized mammal running at full speed. That kind of energy output generates enormous amounts of reactive oxygen species, the same damaging molecules that contribute to aging and DNA damage in every organism. Bats should, by that logic, burn out fast. Instead, they live far longer than similarly sized mammals, sometimes reaching 30 or 40 years.
Comparative genomic analysis of bats has revealed an unexpected concentration of positively selected genes in DNA damage repair pathways and in the NF-ÎşB signaling pathway, which governs inflammation and immune responses. These genetic changes appear to be linked to the origin of flight itself, as though the extreme metabolic demands of powered flight forced bats to evolve better molecular repair systems.15PubMed Central. Comparative analysis of bat genomes provides insight into the evolution of flight and immunity
One of the most remarkable genomic findings involves an entire gene family called PYHIN, which plays a role in the immune system’s ability to detect foreign DNA inside cells. Bats are the only major group of placental mammals to have lost the entire PYHIN family, with only a truncated remnant found in a single species. This loss removes an important set of inflammasome sensors, the molecular alarm systems that trigger inflammation when they detect viral or bacterial DNA. The loss may be an adaptation that reduces the self-damaging inflammatory responses triggered by the DNA fragments released during the metabolic stress of flight, and it may also partly explain why bats can harbor viruses that would cause severe illness in other mammals without getting visibly sick themselves.16PubMed Central. Unique Loss of the PYHIN Gene Family in Bats Amongst Mammals: Implications for Inflammasome Sensing
How Skulls Reveal Ecological Diversification
With over 1,400 living species, bats are the second most species-rich order of mammals after rodents. They eat insects, fruit, nectar, fish, frogs, blood, and other bats. That dietary diversity is reflected in skull shape, and researchers have used three-dimensional skull measurements across the bat family tree to study how ecological specialization drove physical change.
Skull evolution in bats is best described not as a gradual drift but as a series of jumps between distinct “adaptive zones,” each associated with a different ecological role. The cranium was more strongly shaped by echolocation type than by diet, while the mandible tracked dietary habits more closely. The family Phyllostomidae, the New World leaf-nosed bats, stands out dramatically: this single family exhibits more adaptive zone shifts than all other bat families combined, reflecting its extraordinary range of diets from blood-feeding vampire bats to nectar-sipping long-tongued bats to fig-eating fruit bats.17PubMed Central. Signatures of echolocation and dietary ecology in the adaptive evolution of skull shape in bats
Torpor as an Ancestral Survival Strategy
Many bat species use torpor, a state of dramatically reduced body temperature and metabolism, to survive periods when food is scarce. Some temperate-zone bats hibernate for months at a stretch, while tropical species use shorter bouts of daily torpor. The widespread use of torpor across diverse bat lineages, including tropical and subtropical species, suggests that the ability to drop body temperature and slow metabolism is not a late add-on but an ancestral trait that bats inherited from their earliest common ancestor.18Integrative and Comparative Biology. Hibernation and Torpor in Tropical and Subtropical Bats in Relation to Energetics, Extinctions, and the Evolution of Endothermy
Full hibernation, however, evolved separately multiple times. Ancestral state reconstructions indicate that the most recent common ancestor of bats used daily torpor rather than deep hibernation, and that full hibernation arose independently at least four times: once in the ancestor of vesper bats and free-tailed bats, once in the lineage containing horseshoe bats, and independently in at least two other species. The reconstructions also found that thermoregulation in bats never jumped directly from hibernation to fully warm-blooded homeothermy or the reverse; it always passed through the intermediate stage of daily torpor, suggesting that temperature regulation in mammals sits on a spectrum rather than being a set of discrete switches.19PubMed Central. Hibernation in bats (Mammalia: Chiroptera) did not evolve through positive selection of leptin
Bats and the Plants That Need Them
As bats radiated into new ecological niches, they did not just adapt to their environment; they reshaped it. Around 528 species of flowering plants spanning 67 families depend on nectar-feeding bats for pollination. Bat pollination evolved independently many times in plants from a variety of starting points, including lineages previously pollinated by insects, birds, or non-flying mammals. Two separate bat families, one in the Old World and one in the New World, independently evolved specialized nectar-feeding habits and the physical traits that go with them: elongated snouts, long tongues, and reduced teeth.20PubMed Central. The evolution of bat pollination: a phylogenetic perspective
Bat-pollinated plants include economically significant crops like agave (the source of tequila), durian, and certain species of banana and mango. The flowers that bats visit tend to open at night, produce strong musky odors, and offer copious dilute nectar, all traits that distinguish them from flowers adapted for daytime pollinators. This co-evolutionary relationship between bats and plants extends back millions of years and is another reminder that bat evolution has had outsized effects far beyond the bats themselves.