Article | Open

The complex evolutionary history of the tympanic middle ear in frogs and toads (Anura)

Published online:


Most anurans possess a tympanic middle ear (TME) that transmits sound waves to the inner ear; however, numerous species lack some or all TME components. To understand the evolution of these structures, we undertook a comprehensive assessment of their occurrence across anurans and performed ancestral character state reconstructions. Our analysis indicates that the TME was completely lost at least 38 independent times in Anura. The inferred evolutionary history of the TME is exceptionally complex in true toads (Bufonidae), where it was lost in the most recent common ancestor, preceding a radiation of >150 earless species. Following that initial loss, independent regains of some or all TME structures were inferred within two minor clades and in a radiation of >400 species. The reappearance of the TME in the latter clade was followed by at least 10 losses of the entire TME. The many losses and gains of the TME in anurans is unparalleled among tetrapods. Our results show that anurans, and especially bufonid toads, are an excellent model to study the behavioural correlates of earlessness, extratympanic sound pathways, and the genetic and developmental mechanisms that underlie the morphogenesis of TME structures.


The function of audition in frogs and toads (Anura) is primarily the perception of airborne sounds, including those involved in social communication1. Thus, hearing in anurans is thought to be a key trait for survival and reproduction. In most anurans, perception of airborne sounds is enabled by a tympanic middle ear (TME) composed minimally of a tympanic membrane, middle ear cavity, and middle ear bone (=columella, columella auris, stapes, plectrum) that conducts sound waves from the environment to the inner ear where they are transduced into electrical signals via hair cells1,2,3,4.

Among other tetrapods, a TME is absent in caecilians and salamanders1,5 but present in amniotes. Nevertheless, although the TME is primitively present in all extant amniotes, it is not homologous across amniote lineages, having evolved independently at least five times in turtles, lepidosaurs, archosaurs, an extinct lineage of parareptiles, and the synapsid ancestor of mammals6,7,8,9,10. TME losses are extremely rare in amniotes. All mammals, turtles, and archosaurs possess a complete TME, and even the amphisbaenians, snakes, and lizards that have lost the tympanic membrane and middle ear cavity retain a columella11, the sole exceptions being the pygopod lizard Aprasia repens12 and possibly the snakes Atractaspis and Xenocalamus13. In contrast, loss is widespread among anurans, with at least a few species of several families lacking the entire TME, a condition referred to as “earlessness”14.

Earlessness is especially common in the true toad family Bufonidae, in which the TME is completely lacking in more than 200 species. Bufonidae is one of the most diversified groups of amphibians, comprising more than 580 species in 51 genera naturally distributed over numerous ecoregions of the Americas, Africa, and Eurasia15. Several authors have noted the reduction and loss of TME structures (e.g.,16,17,18,19,20) or morphological variations in middle ear structure (e.g.,1,21,22,23) in bufonids, but the phylogenetic distribution of earlessness has never been studied either within Bufonidae or across Anura.

As such, the goals of this study are to explore the sequences of gains and losses of the TME evolution across all anurans and evaluate in a phylogenetic framework the patterns of diversification of TME in Bufonidae.

Materials and Methods

Tympanic middle ear morphology and data collection

Although many variations in the structure of the anuran auditory system exist, a generalized model can be described1,2,3: the lateral-most portion of this system is composed of a highly differentiated disc of thin, non-glandular skin, termed the tympanic membrane. The rim of the tympanic membrane is attached to a cartilaginous tympanic ring, the tympanic annulus. The middle ear cavity is a diverticulum of the pharynx that opens ventrally to the buccal cavity via the Eustachian tubes. The columella contacts the tympanic membrane laterally and the otic capsule medially and is divided into three portions1,24,25: (1) the pars externa plectri or extracolumella, a cartilaginous structure that contacts the tympanic membrane and often presents a slim, flattened strip of cartilage called the ascending process or pars ascendens plectri, that extends anterodorsally to contact the crista parotica of the prootic, (2) the pars media plectri or columellar shaft, an ossified, rod-shaped portion with a dilated medial end, and (3) the pars interna plectri, a mainly cartilaginous structure that is continuous with the pars media and extends posteriorly to lie medial to the operculum. The expanded medial end of the pars media and the entire pars interna constitute the stapedial footplate, which fills the rostral portion of the oval window of the otic capsule1,26. The footplate is connected to the suprascapula via the columellar muscle in some species1, although the individuality of this muscle has been questioned in at least some cases27,28. The otic operculum (operculum fenestrae ovalis), found only in caudates and anurans5, is an ovoid element that is usually cartilaginous or sometimes partially calcified4,25, that contacts the stapedial footplate and covers the caudal portion of the oval window. The operculum is present in all anurans. The opercularis muscle inserts on the operculum and originates on the suprascapular cartilage of the pectoral girdle1,26.

In all observed anuran species (also see29), the presence/absence of TME structures follows a consistent pattern. Absence of a medial structure is accompanied by the absence of the more lateral structures, such that absence of the columella entails absence of the tympanic annulus and tympanic membrane and absence of the tympanic annulus entails absence of the tympanic membrane but not the columella. Similarly, presence of a lateral structure is accompanied by the presence of the more medial structures, such that presence of the tympanic membrane entails presence of the tympanic annulus and columella and presence of the tympanic annulus entails presence of the columella but not the tympanic membrane. Consequently, we made the following assumptions when scoring the presence/absence of the tympanic structures (Fig. 1): (1) absence of the tympanic membrane and tympanic annulus when the columella is absent; (2) absence of the tympanic membrane when the tympanic annulus is absent; (3) presence of the tympanic annulus and columella when the tympanic membrane is present; and (4) presence of the columella when the tympanic annulus is present.

Figure 1: Schematic representation of tympanic middle ear structures in anurans, showing the assumptions followed here for coding absence and presence of the different elements.
Figure 1

TM, tympanic membrane; TA, tympanic annulus; CO, columella

In total, we scored the condition of the TME for 556 species and 51 genera of Bufonidae, representing >94% of all described species in the family. Among the sampled species, 239 were included in Pyron’s30 phylogenetic analysis. We also scored the conditions of these structures for 1860 of the 2538 non-bufonid anuran species (representing 53 families; see15) included by Pyron30, as well as 147 non-bufonid anuran species not included in this analysis. The only frog family not sampled by Pyron30 is the recently described Odontobatrachidae31. Although our outgroup sampling is not exhaustive, we included data for the vast majority of genera of all the families sampled by Pyron30. Details on material examined, considerations about character coding and character states scored for each transformation series, and literature sources are listed as Supplementary Material (section S1 of the Supplementary Information).

Ancestral state reconstructions

We employed the most recent and densely sampled phylogenetic hypothesis available for Anura, that of Pyron30, for ancestral state reconstruction, and we discuss bufonid species not included in Pyron’s30 study but present in other analyses (e.g., well-supported results of32,33,34). We focused original data collection primarily on Bufonidae and relied more extensively on literature accounts for non-bufonids, some of which were unclear or ambiguous about the occurrence of specific structures. In particular, taxonomic accounts often use imprecise terminology to describe the external morphology of the otic region35. Consequently, we analysed the phylogenetic distribution of each of the TME structures in Bufonidae but only the columella in analyses of Anura.

To test the homology of the middle ear structures individually and explain their variation among anurans, most parsimonious ancestral state reconstructions36 on Pyron’s30 phylogenetic hypothesis were performed using Mesquite v3.0337. We further explored alternative evolutionary scenarios for the complete loss of all TME structures within Bufonidae with maximum likelihood ancestral reconstructions using the package APE38 and stochastic character mapping39 using phytools40 in R41. We compared Akaike Information Criterion (AIC) values for a maximum likelihood model in which transition rates were allowed to vary (ARD) throughout the tree and a maximum likelihood model in which transition rates between states were equal (ER). We used the most supported transition rate from our maximum likelihood analyses (ER) to estimate the number of gains and losses across Bufonidae using stochastic character mapping. Stochastic character mapping allowed us to explore the probability of ear transitions under various evolutionary scenarios, giving us a better understanding of the likelihood of regains throughout this family. We considered three scenarios: (1) equal transition rates, no restrictions; (2) equal transition rates and restricting the ancestor to being eared; and (3) a Dollo’s model (no regains possible). We ran 10,000 simulations per scenario and counted a state change whenever nodes switched from greater than 50% support for one character state to greater than 50% support for the other character state.


Tympanic middle ear evolution in Anura

The occurrence of a columella is plesiomorphic in Anura, although the sister clade of all other anurans (Ascaphidae + Leiopelmatidae) lacks this structure (see Discussion). Given that the tympanic membrane and tympanic annulus are not fossilizable structures, their occurrence in fossil material cannot be assessed, making it impossible to determine if the presence of those structures is also plesiomorphic in Anura.

The TME is completely absent in at least some species of no fewer than 20 anuran families: Ascaphidae, Alsodidae, Batrachylidae, Bombinatoridae, Brachycephalidae, Brevicepitidae, Bufonidae, Calyptocephalellidae, Craugastoridae, Dicroglossidae, Hemisotidae, Leiopelmatidae, Leptodactylidae, Megophryidae, Microhylidae, Myobatrachidae, Nasikabatrachidae, Rhinophrynidae, Sooglossidae, and Telmatobiidae (section S1 of the Supplementary Information).

Ancestral character state reconstructions and detailed description of the occurrence of the columella in anuran families other than Bufonidae are provided in section S2 of the Supplementary Information. Ancestral state reconstruction using Pyron’s30 phylogenetic hypothesis shows that the complete loss of the TME, as evidenced by lack of the columella, occurred independently at least 25 times outside Bufonidae, plus two additional losses when taxa not included in Pyron’s30 study but present in other phylogenetic analyses are considered.

Tympanic middle ear evolution in Bufonidae

Based on the phylogenetic hypothesis of Pyron30 and the results of both the parsimony and probabilistic ancestral state reconstructions (Fig. 2 and section S3 of the Supplementary Information), the tympanic membrane, tympanic annulus, and columella were lost in the most recent common ancestor of bufonids, regained subsequently, and then repeatedly re-lost again. Below we summarize the results of the ancestral reconstructions of the tympanic membrane, tympanic annulus, and columella (the numbers of regains and re-losses of TME structures in Bufonidae differs somewhat when taxa not sampled by Pyron30 are considered; see sections S1 and S4 of the Supplementary Information and Discussion, below).

Figure 2: Partial phylogenetic tree of Pyron30 showing parsimony ancestral state reconstructions for the columella in Bufonidae.
Figure 2

The absence of the columella is a synapomorphy of Bufonidae (see section S2 of the Supplementary Information), with independent regains in a subclade of Atelopus, Frostius (not included in this analysis, but see text), and the sister clade of Nannophryne, followed by 10 independent losses.

Parsimony ancestral state reconstruction

The tympanic membrane was lost in the most recent common ancestor of Bufonidae and reappeared in the sister clade of Nannophryne. Subsequent independent losses occurred at least 25 times (see sections S3 and S4 of the Supplementary Information). The absence of the tympanic annulus is the inferred ancestral condition in bufonids, with two independent regains: (a) within Atelopus in a clade composed of A. flavescens, A. franciscus, A. pulcher, and A. spumarius, and (b) in the sister clade of Nannophryne. The gain of the tympanic annulus in the latter clade was followed by 10 independent losses (see sections S3 and S4 of the Supplementary Information). Finally, the phylogenetic distribution of the gains and losses of the columella is identical to that of the tympanic annulus for the taxa included in the hypothesis of Pyron30 (see Fig. 2 and section S4 of the Supplementary Information). However, if bufonid species not included by Pyron30 are also considered, the columella sometimes occurs without a tympanic annulus (see Discussion and section S1 of the Supplementary Information).

Stochastic mapping of complete tympanic middle ear loss under various constraints

Our stochastic character mapping estimated similar patterns of TME loss and regain within Bufonidae under various evolutionary scenarios (see section S3 of the Supplementary Information). When we ran an equal rates model of evolution we found results similar to the parsimony reconstructions of the tympanic annulus and columella with strong support for an ancestor lacking these structures, two regains, and 10 losses within the tree. When we assumed the ancestor had these structures and a model of equal transition rates, we found support for 12 losses and still two regains. When restricting regains from occurring (Dollo’s model), we found a total of 17 losses across bufonids.


The lack of the columella in Ascaphidae and Leiopelmatidae, which together form the sister clade of all other extant anurans (=Lalagobatrachia), has generated much discussion about the plesiomorphic condition in Anura29,42,43,44. However, since most proanurans and stem anuran fossils have a columella (i.e. Mesophryne beipiaoensis, Notobatrachus spp., Prosalirus bitis, Triadobatrachus massinoti, Yizhoubatrachus macilentus45,46,47,48,49), the lack of this structure in Ascaphidae + Leiopelmatidae appears to be synapomorphic. A columella could not be identified in the stem anuran Vieraella herbstii; however, the state of preservation of the specimen is poor50, leading some authors to consider the occurrence of a columella to be unknown (e.g.,47) and others to consider it to be absent (e.g.,43). Regardless, given the phylogenetic position of Vieraella herbstii46,47,50, this controversy has no bearing on our inferences of the evolutionary history of this structure in anurans.

Although available evidence clearly indicates the plesiomorphic presence of the columella in Anura, it is unknown if the ancestral anuran also possessed a tympanic membrane and annulus. The tympanic membrane and tympanic annulus are not fossilizable structures, so their precise phylogenetic origin is unknown. As such, two scenarios are compatible with current evidence: the tympanic annulus and membrane might have been present in the most recent common ancestor of Anura and lost with the columella in Ascaphidae + Leiopelmatidae, or they might have arisen in Lalagobatrachia.

Among non-bufonid anurans, the TME was completely lost at least 27 times (see above). All these losses involve small clades scattered across the major lineages of Anura, implying several putative synapomorphies (e.g., Atelognathus + Chaltenobatrachus, Brachycephalus, Nasikabatrachidae + Sooglossidae, Pseudophryne, Telmatobufo) or autapomorphies (e.g., Balebreviceps hilmani, Melanobatrachus indicus, Rhinophrynus dorsalis). For other anuran clades (e.g., Alsodes, Microhyla, Nanorana, Scutiger, Telmatobius), denser taxon sampling is necessary to obtain adequate evidence to understand the evolution of the TME (see section S2 of the Supplementary Information for a more exhaustive discussion about the evolution of this structure in these and other non-bufonid species).

Although TME structures were lost repeatedly in Anura, TME evolution in Bufonidae is especially complex. All of the families that are closely related to Bufonidae30 (also see51,52) have a complete TME, making the absence of these structures a synapomorphy of Bufonidae. As such, the lack of a TME is plesiomorphic in the earliest diverging lineages (i.e., Amazophrynella, most species of Atelopus, Dendrophryniscus, Oreophrynella, Osornophryne, Melanophryniscus, and Nannophryne).

Independent of the methodological approach, available evidence indicates that the complete TME was regained within Bufonidae in Frostius and the sister clade of Nannophryne, whereas the tympanic annulus and columella were regained in a subclade of Atelopus. Thus, these structures are not homologous with the equivalent structures found in the TME of other anurans, although it seems likely that the underlying genetic basis for their development is homologous (i.e., deep homology53, see below). Subsequent losses occurred several times in different clades, indicating a complex evolutionary history of the TME in Bufonidae (see Fig. 2 and section S3 of the Supplementary Information).

The complex evolutionary history of the TME in Bufonidae is even more unusual when compared to other tetrapods. As noted above, although extant caecilians and salamanders do not possess a tympanic membrane or middle ear cavity, it has been hypothesized that a TME might have been present plesiomorphically and that the lateral elements might have been lost independently5,54. Regardless, although the remaining middle ear structures underwent extensive modification, the columella was lost only once in each group, having been greatly reduced or lost in salamandrid salamanders1,55 and lost in adult scolecomorphid caecilians (present as a cartilaginous element in fetal and juvenile Scolecomorphus kirkii56).

Among amniotes, TME loss is extremely rare. There are no documented losses among turtles (e.g.,57), archosaurs (e.g.,58,59,60), or mammals (e.g.,61), despite the remarkable middle ear transformations in fossorial and marine mammals62,63. Among lepidosaurs, numerous lineages have lost the lateral-most components of the TME, including Serpentes, Amphisbaenia, Agamidae, Diploglossidae, Gymnophthalmidae, Lanthonotidae, Phrynosomatidae, and Scincidae11,60,64,65,66, but the columella appears to be present in all but the pygopod lizard Aprasia repens12 and, possibly, the lamprophid snakes Atractaspis and Xenocalamus13.

With few exceptions, the development of TME structures in anurans follows a consistent sequence that might explain the consistent pattern of co-occurrence of middle ear structures and provides clues about the mechanisms involved in their loss and gain. First, the medial end of the pars media plectri develops as a chondrification within the connective tissue membrane spanning the fenestra ovalis adjacent to the already formed operculum27,67,68,69,70. Next, the pars interna plectri begins to chondrify and, with the incipient pars media, form the future stapedial footplate. Subsequently, a socket-like structure begins to be defined, articulating with the anterior edge of the operculum. The lateral-most portion of the stapedial footplate elongates to complete the formation of the shaft of the pars media plectri, which extends laterally towards the outside of the head. Meanwhile, the tympanic annulus and pars externa plectri develop as cartilaginous condensations associated with the posterior margin of the palatoquadrate. As the palatoquadrate swings posteriorly during metamorphosis, so too do the tympanic annulus and pars externa plectri. As the ontogenetic sequence of development of these structures progresses, they are positioned in the same medial-lateral plane. At this point, the partes media and externa plectri connect synchondrotically to each other and the tympanic annulus induces the differentiation of the tympanic membrane27,70.

The sequences of losses and gains appear to be related to the relative timing of the development of structures (heterochronies) and tissue differentiation phenomena. For example, Helff69 demonstrated the inductive effects of the tympanic annulus on the tegument to produce the differentiation of the tympanic membrane, which explains why the tympanic membrane never occurs in the absence of a tympanic annulus. Similarly, Hetherington27, Smirnov71, and Fabrezi and Goldberg68 emphasized the relatively late development of TME structures. Hetherington27 and Smirnov71 also observed that several species undergo post-metamorphic development of previously absent or undeveloped structures (e.g., Sclerophrys regularis, Pseudacris crucifer, Bombina orientalis). Meanwhile, Smirnov72 pointed out that developmental heterochronies (progenesis, neoteny, and post-displacement) seem to play a major role in the post-metamorphic development of these structures. All these events occur in specific sequences and their disruption in particular stages could produce the observed patterns of losses in the subsequent stages of development of the TME. Therefore, research into the genetic basis for the absence of induction of lateral elements promises to be a fruitful line of investigation.

Additionally, genetic mechanisms that directly regulate the expression of these ear structures might be involved. Knowledge of the origin of the components of the vertebrate auditory system is incipient generally and for anurans particularly. However, recent studies of Xenopus laevis support a model in which the cartilaginous elements of the TME are derived from three neural crest cell streams (see73): (1) the mandibular stream forms the tympanic annulus, (2) the hyoid stream gives rise to the partes media and externa plectri, and (3) the branchial stream forms the pars interna plectri. The consistent patterns of co-occurrences observed in anurans suggest that a direct role of regulatory genes and/or transcription factors might be involved in the tissular differentiation of the tympanic membrane due to inductive phenomena from the tympanic annulus. Also, it is likely that the development of the tympanic annulus and pars externa plectri (in the margins of the palatoquadrate) and the partes interna and media plectri (in the otic capsule) results from the initiation of a common developmental module, as in the morphogenesis of many other structures74,75. Unfortunately, information on the developmental control genes that lead to the formation of elements in the amphibian middle ear is unavailable. However, some genetic pathways involved in this differentiation process have been identified in other vertebrates and could be examined in frogs76.

The lateral–medial dependency between the presence and absence of tympanic middle ear structures appears also to be related to functional constraints: a tympanic membrane without a tympanic annulus or columella would have no acoustic function, as would a tympanic annulus without a columella. In contrast, the tympanic annulus retains its acoustic function in the absence of a tympanic membrane, and the columella remains acoustically functional even in the absence of both structures, as evidenced by the middle ears of salamanders1. This asymmetric functional dependency appears to have allowed these three structures to evolve sequentially rather than as a single transformation series (i.e., presence or absence of all the three structures), with losses and gains of each element occurring sequentially in a lateral–medial dependency, across the bufonid tree.

The losses, regains, and re-losses of TME structures in Bufonidae make true toads an excellent model to study the behavioural correlates of TME morphology. Previous studies have hypothesized a relationship between earlessness and aquatic or fossorial habitats and lack of acoustic communication or production of low-frequency calls77. Additionally, based on the limited evidence presently available (see section S5 of the Supplementary Information), the loss of TME structures in Bufonidae appears to be coincident with the origin of a scramble competition mating system in which males in dense aggregations attempt amplexus indiscriminately and struggle for possession of females78. In this mating system, acoustic territorial defence is absent and reliance on hearing for mate choice is greatly reduced or eliminated, as is the effectiveness of prezygotic isolating barriers like advertisement calls78, which presumably results in the natural interspecific hybridization observed in many bufonid species (see79 and references therein).

Nevertheless, although most of the species of early diverging clades of Bufonidae for which the mating system is known exhibit scramble competition (see section S5 of the Supplementary Information), the reproductive behaviour of most species is unknown, making the character state reconstruction of this behaviour at the root node of Bufonidae ambiguous. Similarly, both within Bufonidae and across Anura, many of the groups that lack a TME employ high frequency (>1 kHz) advertisement calls during reproductive communication (e.g., Atelopus80, Bombina81, Brachycephalus82, Melanophryniscus83, Osornophryne84, Rhinophrynus85, and Sechellophryne + Sooglossus86). Indeed, despite the absence of a TME and the occurrence of a scramble competition mating strategy, interspecific acoustic diversity is maintained in most genera (e.g.,80,87,88,89) suggesting that acoustic signals play a still unclear role in communication and/or mate choice. The maintenance of call diversity and widespread production of advertisement calls may be explained by extratympanic hearing pathways in earless frogs. Multiple extratympanic pathways, including a lung pathway (e.g., Atelopus22,23, Bombina81, Nectophrynoides asperginis21), an opercularis pathway (reviewed by3,25,90,91), and bone conduction enhanced by resonation of the oral cavity (Sechellophryne92) have been shown effective or hypothesized so far in a few earless species. Given that in at least some anurans airborne sounds are transferred via both tympanic and extratympanic pathways (reviewed by3,25), anurans may experience relaxed selective pressures on the TME if TME plasticity does not greatly affect acoustic acuity. If the generality of alternative sound transfer pathways for aerial sounds is corroborated across anuran diversity, then the pre-existence of alternative pathways for airborne sound transmission might explain the high rate of TME loss in anurans. Nevertheless, currently proposed sound localization pathways in anurans all require middle ear coupling93, leaving an alternative mechanism by which earless species localize audible sounds unknown.

Concluding remarks

Since the tympanic annulus and membrane first arose in combination with the plesiomorphically present columella, either prior to the origin of Anura or in Lalagobatrachia (the clade formed by all anurans except Ascaphidae and Leiopelmatidae), our analysis indicates that the TME was completely lost at least 38 times in anurans, usually in small clades within diverse families. Bufonidae is exceptional within both Anura and among all tetrapods in that the loss of all TME structures preceded a radiation of more than 150 earless species followed by independent regains and many additional losses in most derived clades. In contrast, among the approximately 26500 species of amniotes the TME was completely lost only three times and was never regained. Available evidence suggests that losses/gains of each TME structure constitute independent transformation series that occur in a lateral-medial dependency, where heterochronic events and regulation via specific genetic mechanisms are implied during development. The complex pattern of TME evolution, extensive morphological and reproductive diversity, and maintenance of bioacoustic diversity despite the loss of TME structures make Bufonidae a promising model to study extratympanic pathways of sound transmission, the physiological and behavioural consequences of middle ear loss, and the underlying genetic and developmental mechanisms that shaped its remarkable TME diversity.

Additional Information

How to cite this article: Pereyra, M. O. et al. The complex evolutionary history of the tympanic middle ear in frogs and toads (Anura). Sci. Rep. 6, 34130; doi: 10.1038/srep34130 (2016).


  1. 1.

    The amphibian ear. (Princeton University Press, 1985).

  2. 2.

    , , , & The frog inner ear: picture perfect? Journal of the Association for Research in Otolaryngology 16, 171–188 (2015).

  3. 3.

    & Evolution of the amphibian ear In Evolution of the vertebrate auditory system (eds , & ) Ch. 6, 164–199 (Springer Science & Business Media, 2004).

  4. 4.

    The ear and hearing in the frog, Rana pipiens. Journal of Morphology 141, 461–477 (1973).

  5. 5.

    & Evolution of the amphibian ear with implications for lissamphibian phylogeny: insight gained from the caecilian inner ear. Fieldiana Life and Earth Sciences 2, 59–76 (2012).

  6. 6.

    & Evolution of a sensory novelty: tympanic ears and the associated neural processing. Brain Research Bulletin 75, 365–370 (2008).

  7. 7.

    The evolution of tetrapod ears and the fossil record. Brain, Behavior and Evolution 50, 198–212 (1997).

  8. 8.

    An evolutionary perspective on middle ears. Hearing Research 263, 3–8 (2010).

  9. 9.

    & The evolution and development of middle ears in land vertebrates In The middle ear: science, otosurgery and technology (eds , & ) Ch. 3, 7–30 (Springer, 2013).

  10. 10.

    & Impedance-matching hearing in Paleozoic reptiles: evidence of advanced sensory perception at an early stage of amniote evolution. PloS One 2, e889 (2007).

  11. 11.

    Peripheral hearing mechanisms in reptiles and birds. (Springer, 1990).

  12. 12.

    & Cranial anatomy of the pygopodid lizard Aprasia repens, a gekkotan masquerading as a scolecophidian In All animals are interesting: a festschrift in honour of Anthony P., Russell BIS Verlag, Oldenburg, Germany (eds et al.) 303–350 (BIS Verlag, 2015).

  13. 13.

    & The snake skull In Biology of the reptilia Vol. 20 (eds , & ) Ch. 2, 349–692 (Society for the Study of Amphibians and Reptiles, 2008).

  14. 14.

    , & Structure and function of the amphibian middle ear In The evolution of the amphibian auditory system (eds et al.) Ch. 2, 69–91 (Wiley, 1988).

  15. 15.

    Amphibian species of the world: an online reference. Version 6.0 (11 April 2016). (2016).

  16. 16.

    Morphology and phylogenetic position of the west african Didynamipus sjoestedti Andersson, 1903 (Anura Bufonidae). Monitore Zoologico Italiano 15, 187–215 (1981).

  17. 17.

    & A new species of semiarboreal toad with a salamander-like ear (Anura: Bufonidae: Rhinella). Herpetologica 70, 198–210 (2014).

  18. 18.

    et al. Assessing the molecular phylogeny of a near extinct group of vertebrates: the neotropical harlequin frogs (Bufonidae; Atelopus). Systematic and Biodiversity 1, 45–57 (2011).

  19. 19.

    Comparative morphology and evolution of frogs of the genera Atelopus, Dendrophryniscus, Melanophryniscus, and Oreophrynella. Bulletin of Los Angeles County Museum of Natural History 12, 1–66 (1971).

  20. 20.

    Two new genera of african bufonids, with remarks on the phylogeny of related genera. Copeia 1960, 225–233 (1960).

  21. 21.

    , & Acoustic communication in the Kihansi spray toad (Nectophrynoides asperginis): insights from a captive population. Journal of Herpetology 45, 45–49 (2011).

  22. 22.

    et al. Whispering to the deaf: communication by a frog without external vocal sac or tympanum in noisy environments. PLoS One 6, e22080 (2011).

  23. 23.

    , & Biomechanical and neurophysiological studies on audition in eared and earless harlequin frogs (Atelopus). Journal of Comparative Physiology A 183, 265–271 (1998).

  24. 24.

    A. Ecker’s und R. Wiedersheim’s Anatomie des Frosches, part 1. (Braunschweig: Friedrich Vieweg und Sohn, 1896).

  25. 25.

    Pathways for sound transmission to the inner ear in amphibians In Hearing and sound communication in amphibians (eds et al.) Ch. 6, 147–183 (Springer Science+Business Media, LLC, 2007).

  26. 26.

    & Characteristic features in the formation of anuran sound-conducting systems. Journal of Morphology 192, 1–11 (1987).

  27. 27.

    Timing of development of the middle ear of Anura (Amphibia). Zoomorphology 106, 289–300 (1987).

  28. 28.

    & Vibrometric studies of the middle ear of the bullfrog Rana catesbeiana II. The operculum. Journal of Experimental Biology 205, 3167–3176 (2002).

  29. 29.

    Patterns of cranial diversity among the Lissamphibia In Patterns of structural and systematic diversity Vol. 2 The Skull (eds & ) Ch. 6, 255–343 (University of Chicago Press, 1993).

  30. 30.

    Biogeographic analysis reveals ancient continental vicariance and recent oceanic dispersal in amphibians. Systematic Biology 63, 779–797 (2014).

  31. 31.

    et al. The first endemic west african vertebrate family–a new anuran family highlighting the uniqueness of the Upper Guinean biodiversity hotspot. Frontiers in Zoology 11, [1–10] (2014).

  32. 32.

    et al. A new genus and species of arboreal toad with phytotelmonous larvae, from the Andaman Islands, India (Lissamphibia, Anura, Bufonidae). ZooKeys 555, 57–90 (2016).

  33. 33.

    , , , & An extraordinary new species of Melanophryniscus (Anura, Bufonidae) from southeastern Brazil. American Museum Novitates 3762, 1–31 (2012).

  34. 34.

    et al. Gradual adaptation toward a range-expansion phenotype initiated the global radiation of toads. Science 327, 679–682 (2010).

  35. 35.

    & Frogs of the genus Eleutherodactylus in western Ecuador. University of Kansas Natural History Museum, Special Publication 23, 1–236 (1997).

  36. 36.

    Toward defining the course of evolution: minimum change for a specific tree topology. Systematic Zoology 20, 406–416 (1971).

  37. 37.

    & Mesquite: a modular system for evolutionary analysis. Version 3.03. URL (2015).

  38. 38.

    , & APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20, 289–290 (2004).

  39. 39.

    , & Stochastic mapping of morphological characters. Systematic Biology 52, 131–158 (2003).

  40. 40.

    Phytools: an R package for phylogenetic comparative biology (and other things). Methods in Ecology and Evolution 3, 217–223 (2012).

  41. 41.

    R Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL (2013).

  42. 42.

    & Evolution of the amphibian tympanic ear and the origin of frogs. Biological Journal of the Linnean Society 24, 83–99 (1985).

  43. 43.

    & Phylogenetic significance of the amphicoelous frogs, Ascaphidae and Leiopelmatidae. Ethology Ecology & Evolution 5, 233–245 (1993).

  44. 44.

    The anatomy of the head of the New Zealand frog, Leiopelma. Transactions of the Zoological Society of London 27, 255–305 (1951).

  45. 45.

    & A new species of Notobatrachus (Amphibia, Salientia) from the Middle Jurassic of northwestern Patagonia. Journal of Paleontology 8, 403–407 (2008).

  46. 46.

    & A new frog (Amphibia: Anura) from the Lower Cretaceous of western Liaoning, China. Cretaceous Research 25, 761–769 (2004).

  47. 47.

    & Mesozoic anurans from Liaoning Province, China, and phylogenetic relationships of archaeobatrachian anuran clades. Journal of Vertebrate Paleontology 21, 460–476 (2001).

  48. 48.

    & Redescription of Triadobatrachus massinoti (Piveteau, 1936) an anuran amphibian from the early Triassic. Palaeontographica A 206, 1–16 (1989).

  49. 49.

    & An early Jurassic jumping frog. Nature 377, 49–52 (1995).

  50. 50.

    & The earliest known frogs of the Jurassic of South America: review and cladistic appraisal of their relationships. Müncher Geowissenschaftliche Abhandlungen (A) 30, 131–158 (1996).

  51. 51.

    et al. The amphibian tree of life. Bulletin of the American Museum of Natural History 297, 1–370 (2006).

  52. 52.

    et al. Phylogenetic systematics of dart-poison frogs and their relatives (Amphibia: Athesphatanura: Dendrobatidae). Bulletin of the American Museum of Natural History 299, 1–262 (2006).

  53. 53.

    , & Fossils, genes and the evolution of animal limbs. Nature 388, 639–648 (1997).

  54. 54.

    & Evolution of the tetrapod ear: an analysis and reinterpretation. Biological Journal of the Linnean Society 11, 19–76 (1979).

  55. 55.

    The developmental morphology of salamander skulls In Amphibian biology Vol. 5, Osteology (eds & ) Ch. 4, 1684–1781 (Surrey Beatty & Sons, 2003).

  56. 56.

    , , , & Morphology and function of the head in foetal and juvenile Scolecomorphus kirkii (Amphibia: Gymnophiona: Scolecomorphidae). Biological Journal of the Linnean Society 96, 491–504 (2009).

  57. 57.

    Comparative cranial morphology of recent and fossil turtles. Bulletin of the American Museum of Natural History 164, 65–376 (1979).

  58. 58.

    Morphology of the bony stapes (columella) in the Passeriformes and related groups: evolutionary implications. University of Kansas Museum of Natural History, Miscellaneous Publications 63, 1–34 (1975).

  59. 59.

    , & The evolution of the meatal chamber in crocodyliforms. Journal of Anatomy 228, 838–863 (2016).

  60. 60.

    , , & The middle ear of reptiles and birds In Comparative hearing: birds and reptiles (ed , & ) Ch. 2, 13–69 (Springer-Verlag, 2000).

  61. 61.

    et al. The placental mammal ancestor and the post–K-Pg radiation of placentals. Science 339, 662–667 (2013).

  62. 62.

    Middle ear structures in fossorial mammals: a comparison with non-fossorial species. Journal of Zoology 255, 467–486 (2001).

  63. 63.

    Adaptive features of the middle ear of mammal in ontogeny. Acta Zoologica Bulgarica 65, 101–116 (2013).

  64. 64.

    The loss of the external ear opening in scincid lizards. Journal of Herpetology 36, 544–555 (2002).

  65. 65.

    , , & A molecular perspective on the evolution of microteiid lizards (Squamata, Gymnophthalmidae), and a new classification for the family. Biological Journal of the Linnean Society 74, 315–338 (2001).

  66. 66.

    Evolutionary changes in the middle ear of certain agamid and iguanid lizards. Proceedings of the Zoological Society of London 108, 543–549 (1938).

  67. 67.

    The ontogenesis of the sound-conducting apparatus of Bufo angusticeps Smith. Morphologisches Jahrbuch 97, 477–544 (1956).

  68. 68.

    & Heterochrony during skeletal development of Pseudis platensis (Anura, Hylidae) and the early offset of skeleton development and growth. Journal of Morphology 270, 205–220 (2009).

  69. 69.

    Studies on amphibian metamorphosis. III. The influence of the annular tympanic cartilage on the formation of the tympanic membrane. Physiological Zoology 1, 463–495 (1928).

  70. 70.

    & The ontogenesis of the sound conducting apparatus of the egyptian toad, Bufo regularis Reuss, with a review of this apparatus in Salientia. Journal of Morphology 103, 359–375 (1959).

  71. 71.

    Postmetamorphic skull development in Bombina orientalis (Amphibia Discoglossidae), with comments on neoteny. Zoologischer Anzeiger 223, 91–99 (1989).

  72. 72.

    The anuran middle ear: developmental heterochronies and adult morphology diversification. Belgian Journal of Zoology 121, 99–110 (1991).

  73. 73.

    & Segmentation of the vertebrate skull: neural-crest derivation of adult cartilages in the clawed frog, Xenopus laevis. Integrative and Comparative Biology 48, 681–696 (2008).

  74. 74.

    & Reversing opinions on Dollo’s law. Trends in Ecology & Evolution 23, 602–609 (2008).

  75. 75.

    , & Resynthesizing evolutionary and developmental biology. Developmental Biology 173, 357–372 (1996).

  76. 76.

    Can you hear me now? Understanding vertebrate middle ear development. Frontiers in Bioscience 16, 1675–1692 (2011).

  77. 77.

    The effects of body size on the evolution of the amphibian middle ear In The evolutionary biology of hearing (eds , & ) Ch. 21, 421–437 (Springer, 1992).

  78. 78.

    The ecology and behaviour of amphibians. (The University of Chicago Press, 2007).

  79. 79.

    et al. Phylogenetic relationships of toads of the Rhinella granulosa group (Anura: Bufonidae): a molecular perspective with comments on hybridization and introgression. Cladistics 32, 36–53 (2016).

  80. 80.

    , , & Vocalization of eight species of Atelopus (Anura: Bufonidae) with comments on communication in the genus. Copeia 1990, 631–643 (1990).

  81. 81.

    & Lung-based hearing in an “earless” anuran amphibian. Journal of Comparative Physiology A 184, 395–401 (1999).

  82. 82.

    , , & A new species of Brachycephalus (Anura: Brachycephalidae) from the highlands of the Atlantic Forest, southeastern Brazil. Herpetologica 70, 89–99 (2014).

  83. 83.

    , & Acoustic repertoire of Melanophryniscus cupreuscapularis (Céspedez and Álvarez 2000)(Anura: Bufonidae): advertisement, encounter, and release calls. Journal of Herpetology 49, 53–59 (2015).

  84. 84.

    & Sexual dimorphism in Osornophryne guacamayo with notes on natural history and reproduction in the species. Journal of Herpetology 35, 148–151 (2001).

  85. 85.

    , , & Sexual size dimorphism and acoustical features of the pre-advertisement and advertisement calls of Rhinophrynus dorsalis Duméril & Bibron, 1841 (Anura: Rhinophrynidae). Mesoamerican Herpetology 2, 154–166 (2015).

  86. 86.

    & A new species of frog, genus Sooglossus (Anura, Sooglossidae) from Silhouette Island, Seychelles. Amphibia-Reptilia 23, 445–458 (2002).

  87. 87.

    Evolution in the genus Bufo (University of Texas Press, Austin and London, 1972).

  88. 88.

    , & The advertisement and release calls of Melanophryniscus pachyrhynus (Miranda-Ribeiro, 1920) from the central region of Rio Grande do Sul, southern Brazil. Acta Herpetologica 8, 115–122 (2013).

  89. 89.

    & Aspects of the reproductive ecology and behavior of the Tepui toads, genus Oreophrynella (Anura, Bufonidae). Copeia 1989, 445–451 (1989).

  90. 90.

    & Functional aspects of anuran middle ear structures. Journal of Experimental Biology 61, 71–93 (1974).

  91. 91.

    & Electromyography of the opercularis muscle of Rana catesbeiana: an amphibian tonic muscle. Journal of Morphology 175, 17–26 (1983).

  92. 92.

    et al. How minute sooglossid frogs hear without a middle ear. Proceedings of the National Academy of Sciences 110, 15360–15364 (2013).

  93. 93.

    , & Auditory brain stem processing in reptiles and amphibians: roles of coupled ears In Insights from comparative hearing research (eds et al.) Ch. 8, 193–225 (Springer, 2013).

Download references


For access to collections and specimen loans we thank Darrel R. Frost (AMNH), James Aparicio (CBF), Célio F.B. Haddad (CFBH), John D. Lynch (ICN), Linda Trueb, William E. Duellman, and Rafe Brown (KU), Glaucia Maria Funk Pontes (MCP), Ignacio de la Riva and José González (MNCN), Miguel T. Rodriguez and Paulo E. Vanzolini (MZUSP), Ingrid Fernandez and Lucindo González (MNK), David C. Cannatella (TNHC), Ronald Nussbaum (UMMZ), W. Ron Heyer and Ronald Crombie (USNM), and Ariovaldo Giaretta (ZUEC). We also thank MicroCT Core Facility (San Antonio, Texas), and Katyuscia Araujo-Vieira, Aaron Bauer, Sara Bertelli, Laura Nicoli, Diego Pol, Marco Rada, Guillermo Rougier, Juliana Sterli, Miguel T. Rodrigues, and Marcelo Weksler for sharing information about TME structures in different tetrapod groups. Agustin Elias-Costa contributed in the production of Figure 1. We are grateful to the anonymous reviewers for their insightful comments and suggestions that improved the manuscript. We thank ANPCyT, CONICET, FAPESP, CNPq, Fulbright Commission Argentina, and NSF funding for financial support: NSF IOS-1350346; PIP 1112008010-2422, 112201101-00875, and 112201101-00889; PICT 2011-1524, 2010-1740, 2012-2687, and 2013-404; CNPq proc. 305234/2014-5, and FAPESP procs. 2012/09401-5, 2012/10000-5, 2013/20423-3, 2013/50741-7, and 2014/03585-2.

Author information

Author notes

    • Martín O. Pereyra
    •  & Molly C. Womack

    These authors contributed equally to this work.


  1. Museo Argentino de Ciencias Naturales “Bernardino Rivadavia”-CONICET, Buenos Aires, C1405DJR, Argentina

    • Martín O. Pereyra
    • , J. Sebastián Barrionuevo
    • , Boris L. Blotto
    •  & Julián Faivovich
  2. Department of Biology, Colorado State University, Fort Collins, CO 80523, USA

    • Molly C. Womack
    •  & Kim L. Hoke
  3. Departamento de Zoologia, Instituto de Biociências, Universidade de São Paulo, São Paulo, SP 05508-090, Brazil

    • Boris L. Blotto
    • , Mariane Targino
    • , Jhon Jairo Ospina-Sarria
    •  & Taran Grant
  4. Laboratorio de Genética Evolutiva, Instituto de Biología Subtropical (CONICET-UNaM), Facultad de Ciencias Exactas Químicas y Naturales, Universidad Nacional de Misiones, Posadas, N3300LQF, Argentina

    • Diego Baldo
  5. Centro de Investigación de la Biodiversidad y Cambio Climático (BioCamb), Ingeniería en Biodiversidad y Cambio Climático, Facultad de Medio Ambiente, Universidad Tecnológica Indoamérica, Diego de Robles y Vía Interoceánica, 17-1200-841, Quito, EC170103, Ecuador

    • Juan M. Guayasamin
  6. Colegio de Ciencias Biológicas y Ambientales COCIBA, Laboratorio de Biología Evolutiva, Universidad San Francisco de Quito, Campus Cumbayá, Quito, 170901, Ecuador

    • Juan M. Guayasamin
  7. Centro Jambatu de Investigación y Conservación de Anfibios, Fundación Otonga, Geovanni Farina 566 y Baltra, San Rafael, Quito, Ecuador

    • Luis A. Coloma
  8. Universidad Regional Amazónica Ikiam, Muyuna, Tena, Ecuador

    • Luis A. Coloma
  9. Departamento de Biodiversidad y Biología Experimental, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Buenos Aires, C1428EGA, Argentina

    • Julián Faivovich


  1. Search for Martín O. Pereyra in:

  2. Search for Molly C. Womack in:

  3. Search for J. Sebastián Barrionuevo in:

  4. Search for Boris L. Blotto in:

  5. Search for Diego Baldo in:

  6. Search for Mariane Targino in:

  7. Search for Jhon Jairo Ospina-Sarria in:

  8. Search for Juan M. Guayasamin in:

  9. Search for Luis A. Coloma in:

  10. Search for Kim L. Hoke in:

  11. Search for Taran Grant in:

  12. Search for Julián Faivovich in:


All authors conceived of the study. M.O.P., M.C.W., K.L.H., T.G. and J.F. coordinated data collection and analysis. M.O.P., M.C.W., J.S.B., B.L.B., D.B., M.T., J.J.O.-S., L.A.C., T.G. and J.F. collected the data. M.O.P. and M.C.W. performed the analyses. M.O.P., M.C.W., J.S.B., D.B., T.G. and J.F. wrote the manuscript. B.L.B., M.T., J.J.O.-S., J.M.G., L.A.C. and K.L.H. provided critical comments. All authors approved the final version of the manuscript for submission.

Competing interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to Taran Grant or Julián Faivovich.

Supplementary information


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Creative CommonsThis work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit