Sound reasons for tympanic hearing in mammalian precursors

Mammals feature a very sensitive and unique hearing system (1, 2) that is capable of detecting airborne sounds over a wide range of frequencies, exceeding that of many other vertebrates. Elephants can detect frequencies as low as 17 Hz (3), while bats have pushed their high frequency hearing to an extreme, with the ability to detect frequencies beyond 100,000 Hz (4). The evolutionary foundation of this high-performing mammalian auditory system was formed more than 200 Mya, in early-branching Synapsida, a group that contains modern mammals (Fig. 1). In a recent study, Wilken et al. (5) used high-resolution CT scans and biomechanical modeling to overcome the difficulty of inferring auditory capabilities from fossilized ear structures, testing by which mechanism these mammal precursors detected sound. Using a comprehensive finite element model of vibration harmonics, Wilken et al. assessed the hearing efficacy of Thrinaxodon liorhinus, a well-known cynodont from the Early Triassic from about 250 Mya. They discovered that sound reception through the tympanum, or ear drum, a sound transmitting membrane, was already as effective as in modern mammals, even though their middle ear machinery was still quite different from that of modern mammals.

Fig. 1.

Simplified phylogeny of Synapsida with hypothetical sequence of events related to the evolution of ears and brains after refs. 5 and 6. Thrinaxodon is illustrated by a surface model of the skull and mandible of T. liorhinus, University of California Museum of Paleontology specimen V40466. Abbreviations: Carb., Carboniferous; E, Early; Jur., Jurassic; L, Late; M, Middle; Ma, Million years. 3D mesh data were downloaded from Morphosource (7), which is licensed under CC BY 4.0.

A physical challenge for the auditory system of all terrestrial vertebrates is the transmission of soundwaves from air into the fluid-filled inner ear, the cochlea, where sensory hair cells detect vibrations and translate sound waves into electrical signals. The much higher density and lower compressibility of water compared to air cause a reflection of the majority of sound waves at the air–fluid interface. This acoustic impedance mismatch leads to a substantial loss of sound pressure, meaning that the detection of soundwaves is challenging because they are simply too weak to be registered after traveling from air into fluid. However, mammals mitigate this physical challenge by two different mechanisms. First, mammals have a lever system of middle ear bones that increase the vibrational force by having a proportionally longer in-lever. Second, the tympanic membrane, which vibrates in response to incoming soundwaves, is substantially larger than the oval window, the membrane-covered entrance to the inner ear. Hence, the vibrational force is transferred to a much smaller membrane area, which in turn increases sound pressure in the inner ear.

Tympanic hearing has evolved multiple times independently in amniotes (8, 9) and is present in many of the major lineages, a strong indication of the evolutionary fitness benefits tied to this trait as it improved the ability to detect airborne sounds. The tympanic ear of mammals with its specialized middle ear machinery is the result of a prolonged evolutionary transformation, documented by a rich fossil record of early synapsids, the mammal precursors (10). The mammalian middle ear machinery consists of three tiny ossicles: the malleus, incus, and stapes (for reference, the human stapes is only about 3 mm in length). While the stapes, derived from the hyomandibular, a gill arch bone in fishes, is already found in the middle ears of the earliest tetrapods (11), malleus and incus are unique to the synapsid lineage. Intriguingly, the mammalian middle ear is formed from derivatives of ancestral jaw bones (2). Malleus and incus are evolutionarily homologous to articular and quadrate, which formed the jaw joint before becoming integrated into the middle ear when the dentary became a single lower jaw bone. The tympanic membrane, the sound receiver that transfers vibrations to the middle ear machinery, was anchored on the ectotympanic, a bone derived from the angular. While there is a surprisingly detailed picture of how the anatomy of the mandibular bones of synapsids changed, the function of these structures remained elusive.

The conundrum of the mammalian ear evolution lies in the dual function of the involved bones for the intermediate, transitional stages of the ears of early synapsids (12, 13). Synapsids initially employed the articular and quadrate for both chewing and hearing, using bone conduction to transfer vibrations directly to the inner ear. However, at some point during their evolution, synapsids started sensing airborne sounds facilitated by the tympanic membrane. Synapsids with functional tympanic ears would no longer have been constrained to bone conduction of vibrations, transitioning from primarily seismic hearing to mostly airborne hearing. But how exactly synapsids detected auditory cues and when tympanic, airborne hearing first appeared was difficult to assess until now.

Using a comprehensive finite element model of vibration harmonics, Wilken et al. assessed the hearing efficacy of Thrinaxodon liorhinus, a well-known cynodont from the Early Triassic from about 250 Mya.

The biomechanical modeling approach (5) helps to pinpoint the timing of evolution of tympanic hearing. Paleontologists often face the challenge of identifying the function of an anatomical structure, in particular when the involved structures are in part soft tissues and do not have a matching analog among living organisms as is the case for the transitional ears of mammalian precursors. Biomechanics can help guide such functional interpretations by outlining what actions are physically possible and estimating how well the system may work in a set of different assumptions. For example, Thrinaxodon still had large articular and quadrate bones which were load-bearing elements in the jaw joint unlike in true mammals, but Thrinaxodon did have the osteological signatures of a tympanic membrane (5). How did Thrinaxodon hear? Did it conduct sound via the robust mandibular bones or did it already use its tympanic membrane? The comparison of how well sound vibrations are passed on to the inner ear by either using bones or the tympanic membrane might provide the needed clues to answer these questions. Indeed, the estimated sound pressure level at the inner ear when sound was directly received by a tympanic membrane was much higher than the sound pressure level achieved when vibrations were conducted through mandibular bones. These results strongly suggest that tympanic hearing was more effective than bone conduction, providing robust evidence that early cynodonts used a soft tympanum to detect airborne sound, even though Thrinaxodon did not have the middle ear machinery of mammals yet. Placed in a larger comparative evolutionary framework, the modeling results confirm earlier hypotheses that tympanic hearing evolved deep in synapsid history, preceding the complex mammalian middle ear structure (5).

The contribution of Wilken et al. does not only represent a biomechanical analysis of one of the most important transformations in synapsid (and hence our own) history but also is particularly interesting because of its integrative character. As a complex functional system, the auditory apparatus consists of many different components, all of which play a role in determining how well an organism can hear. The new analysis represents an integrated analysis of many structures involved in hearing. Such an integrated analysis might yield more reliable results than work focused on isolated components of the ear-jaw system and may outperform estimates from cochlear size and shape (14). Cochlea morphometrics are used widely across amniotes but the application of this auditory performance proxy to early synapsids is complicated by the absence of well-defined cochlear cavities (15). With the harmonic response analysis, paleontologists have another feasible quantitative approach in their toolkit to estimate and compare auditory performance. The results from harmonic analyses provide a credible range of hearing frequency (low and high frequency limit) and sensitivity (best sensitivity threshold), i.e., yield a quantitative estimate of the hallmarks of auditory performance.

Paleontologists can now advance the understanding of the evolution of a key mammalian trait in the context of the organisms’ environment (16) and explore the ecological, behavioral, and evolutionary consequences of tympanic hearing. The deep origin of this traits in synapsids, well before the appearance of the middle ear machinery (Fig. 1), demonstrates the functional and thus ecological importance of tympanic hearing. Expanding the harmonic response analysis to additional species would eventually allow for testing if tympanic hearing evolved once or perhaps even multiple times within synapsids (15). Future comparative studies may address whether nonfossorial habits tend to favor the evolution of tympanic hearing. Fossorial species are in frequent ground contact, favoring vibration detection through bone conduction when the head touches the substrate, while a shift to above ground activities may co-occur with transition to tympanic, airborne hearing (15). Another axis for future investigations is the preferred diel activity pattern. The mammalian middle ear is often seen as a key innovation that helped mammals to explore nocturnal niches (17), but analyses of visual structures such as scleral ring and orbit morphology suggest that at least some early synapsids also explored nocturnal niches (18). While Thrinaxodon’s hearing at high frequencies remained limited, any small improvements of auditory performance may have facilitated more extensive nocturnal activity, mitigating the diminished importance of vision in low light environments.

What is perhaps most exciting about the ability to estimate hearing range and sensitivity in synapsids is the potential long-term impact for better understanding the factors that control brain evolution. Sensory systems are at the direct interface between organism and environment and thus should be considered highly relevant for evolutionary fitness because sensory input provides detailed information about potential opportunities and threats and facilitates communication. While the sensory input from vision decreased with increasing nocturnality, the sensory information collected by smell, touch, and hearing increased while synapsids compensated for the diminished role of vision (19). This net increase of sensory input equals a larger amount of data to be processed by the brain, which is often connected to the enlarged brains of probainognathians and early mammaliaforms (Fig. 1). Having more precise information about auditory sensory input advances the ability to investigate the role of the senses in mammalian brain evolution. Modeling of visual range and volume can provide estimates of visual input (20), and the size of olfactory bulbs can serve as proxy of the diversity of the olfactory receptor diversity (21). Emerging evidence points to a neurosensory revolution during the Triassic evolution of synapsids (6), and studies such as Wilken’s provide one of the pieces to solve this grand puzzle.