Hearing diversity in 34 000 fish species: A personal perspective^a)

One of the things most exciting about doing science over a long career is to look back and think about how one's interests and work have “evolved.” Indeed, in retrospect, it would seem rather boring to spend one's whole career asking the same questions or using the same research approaches.

As I look at my career, I started by using behavioral approaches to “ask” fish what they could hear (e.g., Popper, 1970). I then moved to questions that included topics such as: sound source localization (Popper , 1973), which was actually supposed to be the basis of my doctoral work; morphology and ultrastructure (Popper, 1971, 1976, 1977, 1980; Sokolowski and Popper, 1987); physiological responses of the ear (Fay , 1974; Fay and Popper, 1974, 1975; Lu and Popper, 2001; Plachta , 2004; Meyer , 2012); hair cell development and regeneration (Popper and Hoxter, 1984; Presson and Popper, 1990; Lombarte , 1993; Lanford and Popper, 1996; Lanford , 1996; Smith , 2006); evolution of hair cells and the ear (Chang , 1992; Popper , 1992; Fay and Popper, 2000; Coffin , 2004), and any number other topics that were directed at understanding the biology of fish hearing.¹

However, over the past 25 years or so I started to move in a rather new direction—one that I like to call “translational fish bioacoustics.” Here, just like biomedical researchers who strive to transition their bench science into clinical applications (Strand, 2020), I took my basic science and applied it to work protecting fishes from anthropogenic (human-generated) sound (e.g., Popper and Carlson, 1998; Popper, 2003; Popper , 2005).

Indeed, “translational fish bioacoustics” is a growing field these days as regulators and others start to appreciate that anthropogenic sounds produced by a broad range of sources in both marine and freshwater environments, including (but not limited to) offshore wind farms, seismic air guns, commercial shipping, and military activities have the potential to alter the lives of fishes and have a significant impact on fitness and populations. As a result, we have begun to realize that to understand the potential effects of anthropogenic sounds we need to understand a great deal more about fish hearing and general fish bioacoustics. However, we also now see that there are great gaps in our knowledge of fish bioacoustics that range from what fishes hear and how well they hear the sounds to how they respond to sounds of different amplitudes and types (Hawkins , 2015). Because of these gaps, it is not yet possible to successfully suggest ways, in the forms of criteria and mitigation, to “protect” fishes from potential harm resulting from the sounds.

So, coming back to my career, I started to think about the issues I have explored over the past 57 years and which of those still most interest me. Put another way, had I another career ahead of me (knowing what I know now), what questions and issues would I most like to focus upon over the next 10–15 years. (Assuming that grant funding was not partly driving the questions upon which I would work!) Then put a different way, if I could suggest the research questions to be asked by the current and next generation of investigators working on fish bioacoustics, what would those suggestions be?

The answer is one I have discussed many times with close friends and colleagues, most notably Dick Fay, Tony Hawkins, and Chris Platt (Popper, 2020)—the biology of fish hearing and sound communication! In effect, what I know now is that despite over 100 years of research on fish hearing (Sand , 2024), we really know very little about what and how fishes hear and process sounds. To really explore and answer questions about anthropogenic sound we need to know far more about these basic topics.

So, the purpose of this paper is to propose the kind(s) of research I personally would like to pursue in the next decades. While I realize that, at my time of life, I am not going to be doing this work, I am hoping that perhaps I will help some future investigators see avenues of research that will help reveal much more about the biology of fish hearing and intrigue them as much as they intrigue me.

I am fascinated that much of life is a series of serendipitous events. That is, opportunities continuously open to us in life – a young woman coming into my graduate student office in 1967 to talk with my officemate one day and that leads to our being married now for almost 55 years, or one happens to pick up a random journal and sees an article that leads to a totally new insight into one's research. The critical thing about serendipity, however, is that it requires having an open mind. Indeed, as pointed out by Busch (2020) “cultivating serendipity is first and foremost about looking at the world with open eyes and seeing opportunities others do not. It is not just about being in the right place at the right time and having something happen to us (blind luck), but rather a process in which we can be actively involved.”²

The fact is that serendipity has been fundamental in directing my career. I could share example after example (and perhaps readers are now thinking back on the role of serendipity in their lives), but just give a few examples.

A major serendipitous event in my life occurred when Dick Fay moved to the University of Hawai'i as a postdoc while I was an assistant professor of zoology (Popper and Fay, 2016). Dick and I instantly “hit it off,” both professionally and personally (as did our wives and later our kids). (Dick and Cathy Fay and my wife Helen and I toasted our 50th “anniversary” over Zoom on December 26, 2021.)

My view, and I know that Dick felt the same way (see Fay , 2023), is that both of us would have had fine careers, but our synergy led to things that neither of us would have ever thought about, or pursued including organizing eight scholarly meetings as well as founding and editing the now 77 volume Springer Handbook of Auditory Research (SHAR) (Fay and Popper, 2014).

In fact, SHAR is a product of serendipity. I happened to go to the Society for Neuroscience meeting in the late 1970s. In browsing the book exhibit, I wandered to the booth for the publisher Springer-Verlag (now Springer Nature). By chance, at the booth was Dr. Mark Licker, a senior Springer editor. I mentioned that Dick and I were organizing a special session for the Acoustical Society of America (ASA) meeting in Honolulu and casually asked if Springer might be interested in publishing the proceedings. Mark was immediately interested in the idea, leading to Dick and I publishing our first book (Popper and Fay, 1980). As Dick and I started to organize other meetings (Tavolga , 1981; Atema , 1988; Webster , 1992), we kept returning to Mark, and he and his successors were interested in what we were doing.

So, when Dick and I one day came up with the idea of SHAR and proposed it to our then editor, Dr. William Curtis (Fay and Popper, 2014), Springer quickly agreed to the crazy idea of eight volumes on hearing and signed us to a contract. Had I not gone to the meeting, and had Mark not been at the Springer booth at the right moment, there might be no SHAR!

A final example of serendipity in my career (and there are many others) occurred in the late 1980s when I had a call from a staff member of the U. S. Congress who wanted to know about the use of sound to control the movement of salmon in rivers. This was not a topic I knew anything about, but I agreed to do a literature search and write a report on the feasibility of using sound to keep returning salmon from entering irrigation ditches in California (apparently the issue was raised by a member of Congress from that state) (Popper, 1995).

Based on this report, I was invited to several meetings on fish passage and met Tom Carlson. By that time, Tom and I were both quite interested in fish passage and sound and we both felt that the literature in the field was weak and in need of critical review. Thus, we collaborated on a paper that has been very widely cited (Popper and Carlson, 1998). The paper also led to invitations to a range of different opportunities to consider potential uses of sound to control fish.

Of course, that serendipitous phone call led me on a new career path into “translational fish bioacoustics.” I wound up with a substantial number of opportunities for my lab to get involved in studies on the potential effects of anthropogenic sound on fishes (a few examples include McCauley , 2003; Popper , 2005; Popper , 2007; Popper and Hastings, 2009; Halvorsen , 2012; Popper , 2014; Hawkins , 2015; Dahl , 2020; Popper , 2022). However, that is not the topic of this review!

As I consider this paper, it strikes me that just writing the same things I have discussed before is not something I want to do again (just read my older papers!). Instead, as I think back on fish hearing and what I have learned, and the question that I would like to ask, my goal in this paper is to raise just a few of the issues and questions that still puzzle me and would likely be the ones I would like to work on. Indeed, there are no answers to the questions I raise, but perhaps others will read this and get some ideas and start to explore questions about fish hearing that are very different and in my view far more exciting, than the “mundane” questions of what sounds fishes hear!

I also want to point out that this paper is not meant to be anything like a review of all the past work on any topic. Rather, I have focused on citing very few papers on any topic, and there is some emphasis on some of the older literature (and some of my papers to show the extent of my work)—something I think that many younger investigators often do not know about, or ignore, despite many such papers, even today, being at the forefront of creative thinking and research on fish hearing.

I am, in truth, a comparative biologist, and have been from my first class in “comparative anatomy” at New York University instructed by Professor Douglas B. Webster where I “fell in love” with the comparative approach to doing biology! While, as discussed below, I see immense value in studying one or a few species, I am going to take the approach in this paper to focus on some of the comparative questions that interest me the most and which I think need further exploration if we are to really start to understand fish hearing. Indeed, my hope is that what excites me might “rub off” on some student(s) or colleagues in the future and they will decide to follow my lead, and that of Gustav Retzius (1881), and explore comparative fish hearing.

Of course, there is one very important caveat to what I say in this paper—while my focus is on hearing, I know, and totally appreciate, that many of the structures involved in fish hearing are also involved with other functions. Just two examples. First, the swim bladder clearly has much more than an auditory role. In fact, it likely first evolved to deal with buoyancy issues and its roles in hearing and in sound production came later. Thus, while I will deal with hearing and the swim bladder, buoyancy, and other issues no doubt tempers the hearing role of the organ.

Second, the ear is both a hearing organ and involved in vestibular senses. In fact, many have argued that the three otolith end organs are “simultaneously” involved in both senses (though we do not know how!). Thus, while I focus on hearing, much of ear structure and function probably have evolved to take on multiple tasks.

It is critical that the reader understand that this paper reflects my views and interests! Indeed, each reader will likely have their own views as to what is interesting and important about fish hearing, and that is great. My goal here is to share my ideas. Thus, in reading this paper, please keep in mind that this is a very personal perspective!

One of the most fascinating things about fishes is the extraordinary diversity among the more than 34 000 extant species (more than all other vertebrate species combined!) (see Helfman , 2009). This diversity is found in every facet of the lives of fishes, but certainly, it is not limited to how they feed, where they live, their anatomy and physiology, their ecology, and everything else (e.g., Greenwood , 1966; Helfman , 2009). Indeed, different species live in shallow water, at immense depths and pressures, and even on land (some even glide through the air!), and they have adapted their basic body plan to meet where and how they live. Of course, as reflected first in the amazing treatise of Gustav Retzius (1881), this diversity includes the ear.

As a result of this diversity, one can imagine that what we see in fish ears and auditory periphery (we know far less about the central nervous system) may reflect great diversity in how and what fishes hear. Or all fishes may hear sounds in basically the same way, but there are diverse options to get at the same end because of various fish taxa taking somewhat different evolutionary paths. The questions that arise focus on how to explore fish hearing so that we ultimately understand the diversity we see in ears, and to understand what this diversity means for sound detection and processing.

There are perhaps two extremes in how one might examine fish hearing, neither of which is either satisfying or likely to fully help us understand the subject. One approach is to investigate, in great depth, hearing in a single “representative” species. However, what is a representative fish species, and are data from a single species likely to help us understand the hearing of 34 000 species?

The other extreme is to sample hearing in a substantial (unknown, but probably very large) number of species to get a sense of variation in hearing and hearing mechanisms. The problem here, however, is just as intractable as focus on a single species. Which species, and how many species, does one select to give an appropriate sample of fish hearing? We now know something about hearing in a bit over 100 species (Ladich and Fay, 2013), but is that enough, and are those 100+ species “representative” of the diversity in structure and function of the auditory system among fishes?

Of course, the best way forward is a combination of these approaches though that is quite hard with fishes. This is because of the number of species, making it impossible to get a clear appreciation of variation from a comparative approach. Yet, it is equally impossible to get a sense of how fishes will respond to signals by just studying one or two species, often selected because of availability and not for particular characteristics. Indeed, imagine trying to study hearing in a 11 m long giant oarfish (Regalecus glesne) (https://bit.ly/3XocUtT) or the hadal snailfish (Pseudoliparis amblystomopsis) which lives at a depth of 7700 m (https://bit.ly/3r7gUCU).

One of the strengths of my 50+ year collaboration with Richard (Dick) Fay is that we came to our interest in fish hearing from two rather different approaches. Dick was trained as a psychologist and his focus was on understanding hearing in depth in a single species – in his case, the goldfish (wonderfully summarized in Fay, 2011; Fay , 2023),³ though Dick later branched out to, collaborating with my former doctoral student Peggy Walton, to explore central auditory processing in the oyster toadfish, Opsanus tau (e.g., Edds-Walton and Fay, 2005, 2009).

Indeed, much of what we know about fish hearing comes from in-depth work on one or a few species such as Fay's goldfish work, studies of hearing and acoustic behavior of various toadfish species (e.g., Winn, 1972; Fine, 1978; Sokolowski and Popper, 1988; Bass and Marchaterre, 1989), and investigations of hearing in Atlantic cod (Gadus morhua) by Anthony Hawkins, Olav Sand, and their collaborators (reviewed in Hawkins, 2022).

These studies, and particularly those by Fay, delved deeply into fish hearing capabilities. They taught us that the auditory system in fishes (or at least in goldfish!) is not unlike that of other vertebrates in that hearing range and sensitivity are only one (and perhaps least important) part of hearing (reviewed in Popper and Fay, 1997; Fay and Popper, 2000).

Importantly, Fay's studies demonstrated that the important roles for hearing in fishes include things like detection of signals in noise, discrimination between signals, determination of the direction of a sound source, and the recognition of complex communication signals to identify “friend from foe.” Thus, the most important questions are whether these capabilities are found in all fishes, and how fishes “achieve” these capabilities. Moreover, are the mechanisms used in the detection and processing of sound the same in all species, or have different, highly diverse, species “invented” different ways to achieve these ends?

Despite all we know, and will continue, to learn from studies of a few species, the very fact that there are 34 000 species with tremendous diversity in hearing systems forces us to wonder—how does this diversity reflect in hearing capabilities? Sadly, but understandably, comparative hearing gets scant attention from investigators today (e.g., Brown , 2023). Yes, different investigators study different species, but few are asking true comparative questions, or even putting together the data in the literature in a true comparative sense to try and understand the “big picture” of fish hearing.

The diversity to be found in fish ears was most clearly demonstrated by the great anatomist Gustav Retzius (1881) in his monumental descriptions of the ears of vertebrates. Yet, more than 140 years after this work, we have only the most limited understanding of the functional significance of the diversity of fish ears and how hearing may have evolved in the living fishes (much less their ancestors).

Some of the diversity lies in the size, shape, and general characteristics of the semicircular canals (Platt, 1983), but it is particularly apparent in the otolith organs of the ears—the saccule, lagena, and utricle. The variation may include the overall shape of the otolith organs, the size and shape of the otolith within each end organ, the size and shape of the sensory epithelia (maculae), and in many characteristics (as well as the number) of the sensory hair cells.

A striking issue is that we really have little understanding of the functional significance of the differences we see in the otolith end organs. It is possible that the differences are just evolutionary “experiments” in detecting sounds and that all the differences seen reflect different ways to do the same thing. Alternatively, and more likely, the differences may reflect different hearing capabilities within closely related groups of fishes.

For example, Sheryl Coombs and I examined the inner ears of two closely related sympatric species, Myripristis kuntee (subfamily Myripristinae—soldierfish) and Adioryx xantherythrus (now Sargocentron, subfamily Holocentrinae—squirrelfish) and found that they have very different ear structures and very different hearing capabilities (Coombs and Popper, 1979). Myripristis has a very large saccule, and the anterior end of the swim bladder tightly adheres to the bone around it. In contrast, Adioryx has a very different saccular structure, a swim bladder that terminates some distance from the ear, and hearing range and sensitivity is not nearly as great as that in Myripristis. We have no detailed knowledge of the functional significance of these differences, but the differences in hearing and ear morphology in such closely related species must be related to a functional difference for the fish.

While it would be possible to cite numerous other examples of inner ear and hearing differences, the point is that there is immense diversity in fish ear structure, and it is likely this has a functional significance. While we do suspect that this means different hearing range and sensitivity, the far more important questions are likely to be whether these capabilities are solely related to sensitivity and the bandwidth over which various species hear or are there other far more important differences which are, to date, very poorly studied on a very few species. For example, do differences in ears reflect differences in discrimination between sounds or determining sound source direction or detection of signals in noise, etc.? Or, put another way, are we asking the wrong questions when we ask about fish hearing capabilities? Instead, should we be asking questions that delve more deeply into hearing capabilities if we are to understand fish hearing?

Fundamentally related to comparative hearing and what fishes hear are questions concerning how fish ears work, and whether all ears of all fishes respond to sound in the same way? It is clear that the ears of fishes detect particle acceleration and that they only detect pressure if there is a separate pressure receptor, such as the swim bladder, which “converts” pressure to particle motion near enough to the ear for the ear to be able to detect it (e.g., de Vries, 1950; Dijkgraaf, 1960; Popper and Hawkins, 2018; Schulz-Mirbach , 2019a).

There is also general consensus that the mode of stimulation of the ear is that, in a sound field, the fish's body and ear tissues move with the water motion while the far denser otoliths move at a different amplitude and phase (de Vries, 1950; Dijkgraaf, 1960). Since the tips of the cilia on the sensory hair cells are directly coupled to otoliths, this relative movement causes bending of the cilia, thereby causing ionic transfer across the hair cell membranes, resulting in the release of neurotransmitters that stimulate the fibers of the innervating cranial nerve VIII, thereby sending signals to the brain (e.g., Ashmore , 2010).

However, while this general mechanism makes considerable sense, the most interesting aspects of ear function in fishes is likely to be in the details (e.g., Schulz-Mirbach , 2019a; Schulz-Mirbach , 2019b). Indeed, there are numerous questions about how the ear works in fishes. Furthermore, it is likely that the motions in the ear in response to an impinging sound field and stimulation of the sensory cells are every bit as complex (though in different ways) as might be found in the ears of frogs, birds, and even mammals! However, the truth is that we know almost nothing about how the ear functions in fishes other than that the stimulation comes from an interaction of the otolith and sensory cells. The question becomes even more complex once we consider the immense diversity in the structure of the ears in different species!

This leads to questions that have yet to be explored. For example, what is the pattern of motion of the otolith. One way of thinking is that otoliths move back and forth relative to the epithelium. However, it is far more interesting to wonder why the variation in the shapes and sizes of otoliths in different species occurs.

Indeed, it is hard to imagine that this variation does not have functional significance! This leads to the suggestion that the motions of the otolith are likely rather complex and change as the direction and other aspects of the particle motion component of the sound field changes. Thus, the hair cells that are stimulated, and/or the levels of their responses, when the sound comes from one direction are very likely different when the sound comes from a different direction. Moreover, it is reasonable to suggest that there are inter-specific differences in the stimulation of the sensory cells by sounds of different frequencies, amplitudes, and other factors.

Moreover, the complexity of the stimulation of the ear is further seen by the substantial differences in the hair cell orientation patterns of the sensory cells, particularly in the saccule, although there is also interesting variation in patterns in the lagena of many species (e.g., Popper, 1977). We are quite certain that hair cells oriented in different directions extract directional information from the motion of the otolith, and thus provide fish with a peripheral direction detector (unlike in terrestrial vertebrates where sound direction information from the two ears is processed in the brain) (Popper, 1977; Nedelec , 2016; Hawkins and Popper, 2018). However, it is also possible that other information is being gathered about the sounds by having hair cells in “orientation groups” so that they respond best to sounds from different directions.

Put another way, the combination of complex motion of the otoliths and the complex patterns of the sensory cells found in many fish species, must be considered when thinking about fish hearing mechanisms. It is hard to imagine that the complex interactions between otolith and sensory cells are “random.” Instead, considering how evolution works, it is parsimonious to argue that the complex nature of the ear has evolved to do far more than just detect the presence of sound and its level – and probably even more than just localizing sounds.

Indeed, as Dick Fay demonstrated in goldfish (e.g., Fay, 1974; Fay , 1978; Fay and Passow, 1982), it is likely that much of the structure of the fish ear performs many of the same auditory functions as found in terrestrial vertebrates. This is not, in fact, an unreasonable hypothesis considering the very long evolutionary history of fishes!

Finally, and coming back to comparative hearing, why do species differ in otolith structure, overall ear shape, ciliary bundle length, extent of epithelia covered by the otolith, “rigidly” of connection between the otolith, and the cilia of the hair cells, etc.? These, and related questions, are at the heart of what make fishes such exciting subjects for research!

All fishes detect particle motion (e.g., de Vries, 1950; Hawkins and Popper, 2018; Popper and Hawkins, 2018)! Far fewer species also detect sound pressure using a transduction mechanism involving an auxiliary structure, generally a bubble of air such as the abdominally located swim bladder. The gas bubble is set into motion by an impinging sound and re-radiated in the form of particle motion which stimulates the inner ear. It is also generally accepted that for this mechanism to enable pressure detection, the air bubble must be close to, or in contact with, the ear. Otherwise, the particle motion attenuates as it travels from the swim bladder to the ear sufficiently so it can no longer produce otolith motion (Alexander, 1966). Thus, it has been argued that many species with swim bladders far from the ear, such as tunas, salmonids, etc. (e.g., Hawkins and Johnstone, 1978) likely only detect particle motion, whereas fishes where the swim bladder has an anterior projection coming near the ear, like in gadids and the aforementioned holocentrids (Myripristis and Adioryx), detect both particle motion and sound pressure (e.g., Chapman and Hawkins, 1973).

While investigators of fish hearing have basically accepted these ideas, the fact is that work on swim bladder contributions to hearing has only been done on a few species (e.g., Sand and Hawkins, 1973), and that there are far more questions than answers. For example, while Alexander (1966) did considerable work on swim bladder function, he only explored a limited number of species, and his results, and the limited work from other labs, have not been tested more widely, nor has it been broadened very much to additional diverse species. Indeed, there has not been a real functional comparative approach to the role of the swim bladder in hearing that would help us understand its contributions to hearing in different species, and how it may vary based on its anatomy, whether the swim bladder is closely tied to surrounding tissue or free to move, its distance from the ear, etc.

At the same time, the swim bladder is a multi-functional organ. While it clearly plays a role in hearing and in sound production in many species (e.g., Fine , 2001), its major function is that of a hydrostatic organ, helping fishes adjust to depth (e.g., Steen, 1970; Blaxter and Tytler, 1978). Thus, in considering the role of the swim bladder in fish hearing (and sound production), one must also consider its role as a major hydrostatic organ.

Bones connecting the anterior end of the swim bladder to the inner ear were first described by, and later named for, Weber (1820)—the Weberian ossicles. Over 10 000 species, all members of the superorder Ostariophysi (especially the series Otophysi), have Weberian ossicles, and we know that most otophysans studied (albeit, still very few of the 10 000 species) hear better both in terms of wider frequency range and better sensitivity than most (though far from all) species without the ossicles (Poggendorf, 1952; Ladich, 2023).

While the enhancement of hearing has been assumed, there have only been a few, rather indirect, studies focused on the function of the Weberian ossicles. The first specific work directly on ossicular function was by Poggendorf (1952) who successfully ablated the bones in a catfish (Amiurus nebulosus) and showed a decrease in hearing sensitivity and bandwidth. The only other study, by Ladich and Wysocki (2003), showed hearing loss in goldfish with removal of the most posterior ossicle, the tripus. Not unexpectedly, the specific losses differed in the two studies—something that could be due to methodology, species, or many other factors. So, while the data suggests important roles for the bones, we have no clear information about specific function.

One of the problems in understanding the ossicles is that it is very hard to remove them or cut the fibrous connection between them without massively harming the fish. Thus, experimentation is difficult, and extensive controls are critical. Moreover, while other authors suggested function for the ossicles (e.g., Kleerekoper and Roggenkamp, 1959; Fay and Popper, 1974), these have really only looked at the effects of removal or alternation of the swim bladder, and so it is impossible to differentiate effects of loss of ossicle function when the structure driving them, the swim bladder, is not intact.

Still, it is generally agreed that the Weberian ossicles directly couple the movement of the swim bladder walls to the fluids of the inner ear, thereby resulting in motion of the saccular (and possibly lagenar) otolith and stimulation of the sensory cells (Alexander, 1962). Indeed, the morphology of the saccular otolith in Otophysi is very different than in non-otophysans (Popper, 1971), and has extensions that likely “catch” fluid movement induced by the Weberian ossicles.

This leads to a range of questions about the actual function of the Weberian ossicles and their role(s) in hearing. The basic questions relate to the movements of the ossicles and how movements might change with different frequencies and other aspects of signals impinging upon the swim bladder. There are also questions related to comparative function. Indeed, as pointed out by Ladich (2023), even within one otophysan group, the Siluriformes (catfishes) there may be from one to four ossicles between the swim bladder and the inner ear. Questions then arise as to the functional differences related to the morphological differences, and, of course, questions about the evolution of the different patterns.

In effect, all we know about the Weberian ossicles is that they appear to enhance hearing bandwidth and sensitivity by directly coupling motions of the pressure receptor (the swim bladder) to the ear, overcoming the attenuation of the signal through body tissues in species where the swim bladder is not directly coupled to the ear. Yet beyond that, do the differences in number of bones, bone morphology, and perhaps in connections to the ear and other aspects of structure, impact the signals that the bones transmit or anything else about how the signals get to the ear. It is known that the bones of the mammalian middle ear can affect signals (Puria , 2013) and so one wonders if the Weberian ossicles might behave in similar ways.

When I first started doctoral work, I proposed to my advisor, William N. Tavolga (a truly great scholar and one of the “founders” and true pioneers of fish bioacoustics—see Tavolga, 1964, 1967), that I investigate sound source localization in fishes. My interest in localization arose since it was apparent that this ability is of critical importance to animals (e.g., Masterton , 1969) but, at that time (1966), the literature on localization by fishes was scanty, and contradictory (e.g., von Frisch, 1923; Reinhardt, 1935; von Frisch and Dijkgraaf, 1935). One conclusion reached by van Bergeijk (1964, 1967) was that fishes cannot localize sound since the ears were very close together, the speed of sound in water very high, and so there was not the interaural differences in detection that is required for localization by many terrestrial vertebrates (van Bergeijk, 1964, 1967; Moulton and Dixon, 1967).

Tavolga, fully knowing that studying sound localization in fishes was an immensely difficult problem, predicted it would take me 17 years to complete my doctoral work. Ultimately, I agreed that Bill was right, and I switched topics, but my interest in fish sound localization has never waned (e.g., Popper , 1973; Hawkins and Popper, 2018). Indeed, despite advances on the topic, and the logic that fishes should be able to localize sound so they can tell the position of “friend, food, and foe,” we still know very little about how well fishes localize sound, the mechanisms they use, and if all fishes localize equally well and with the same (or similar) mechanisms (Fay, 2005; Hawkins and Popper, 2018)!

It was not until the exquisite work of Arie Schuijf and Rob Buwalda (and colleagues) in the Netherlands that it became clear that at least the few species of bony fishes studied could determine sound direction (e.g., Schuijf , 1971; Schuijf and Buwalda, 1975; Schuijf and Buwalda, 1980). (There was, before that, some evidence that sharks could localize sound—Nelson, 1967; Nelson and Johnson, 1972). The critical piece of thinking was that directional determination in fishes did not rely as much on central comparison of signals from two ears as occurs in terrestrial vertebrates (including all mammals) (Popper and Fay, 2005), but that much of the determination of direction took place in the ears themselves. The mechanism likely involved in directional hearing became particularly apparent with the pretty much simultaneous but independent results from three labs that fish ears have tens of thousands of sensory hair cells that are oriented into groups that respond best to sounds from different directions (Dale, 1976; Enger, 1976; Popper, 1976).

The fact is, while we are quite certain fishes can localize sound, we still know very little about their capabilities, nor do we have any sense of differences in capabilities and mechanisms (if any) in different species. For example, are localization capabilities at all correlated with differences in hair cell orientation patterns in different fishes, and is the complex structure of the otoliths in many species involved in localization? We also suspect that all three otolithic end organs are involved with localization since, when combining the input from hair cells oriented in different directions in all six end organs in each species, there is probably a full three-dimensional coverage of the direction of sound from all directions (Zeddies , 2010; Zeddies , 2012; Coffin , 2014)

There are also questions about localization capabilities—something we know almost nothing about other than from the aforementioned work of Schuijf and colleagues and some more recent work from the laboratory of Joseph Sisneros on the plainfin midshipman (Poriththys notatus) (Zeddies , 2010; Zeddies , 2012; Coffin , 2014). What, for example, is the ability of fishes to discriminate between sounds from different directions (often referred to as the minimal audible angle, MAA), but the few data from fishes suggest that their MAA is 15 degrees or more (e.g., Buwalda , 1983), compared, for example, to marine mammals and many terrestrial species that can discriminate sounds only 1–2 degrees apart (e.g., Renaud and Popper, 1975). While Schuijf and colleagues suggested that fishes use both sound pressure and particle motion to detect direction (e.g., Van den Berg and Schuijf, 1983), this has only been shown in a few species, and considering that most species do not detect sound pressure (e.g., salmon, tunas), or they are primarily detectors of particle motion, does this mean that most fishes cannot localize sounds well?

I still think that sound localization is one of the most interesting and important questions regarding fish hearing, and I would argue that much of the structure of the ear has evolved not for hearing sensitivity or range of sounds to hear, but to localize sounds. As pointed out by Masterton (1969), sound source localization is a very fundamental part of hearing, and so it makes some sense that fishes, which have been hearing for millions of years more than terrestrial vertebrates, would have evolved very sophisticated systems to determine the direction and distance of sounds that could impact their lives.

Clearly, there are many other aspects of fish hearing that remain open, particularly from the perspective of inter-specific differences amongst the numerous species and how they live. To my thinking, the issues raised up to this point are the most interesting (if not some of the most important), but there are many other issues that deserve some consideration, and a few will be mentioned here, albeit briefly. The real “problem” is that despite fishes being basal vertebrates, and integral to the lives of humans, one being as a major source of protein in the diet (e.g., Tacon and Metian, 2013), we know little about many aspects of the lives of fishes, with hearing and the use of sound for communication areas of particularly little knowledge.

Considerable work has been done on the auditory regions of the central nervous system (CNS), but just on a few species (e.g., Northcutt, 1980; McCormick, 2001; Edds-Walton, 2016). However, most studies have focused on neuroanatomy and have not dealt with function (see, e.g., Page, 1970; Bodnar and Bass, 1997; Crawford, 1997; Edds-Walton and Fay, 2009). Because of this, relatively little is still actually known about the roles of different brain regions in a single fish in hearing or whether there is substantial interspecific difference in the role of the brain in hearing and in sound processing.

Moreover, while there has been considerable discussion over decades about the evolution and interaction of the auditory and lateral line systems, there is a lack of data, and particularly comparative and functional data, on how the two systems interact in the CNS and, perhaps, provide fishes with combined information related to the two senses.

For many decades, investigators of fish hearing have had to deal with there being substantial levels of both sound pressure and particle motion in water. While fish ears have been known to be detectors of particle motion for many decades (e.g., de Vries, 1950; Dijkgraaf, 1960), most hearing work on fishes has focused on measurements of sound pressure and not particle motion. In great part, this is because measuring particle motion, even today, is not nearly as easy (or as inexpensive) as measuring sound pressure (e.g., Popper and Hawkins, 2018; Nedelec , 2021), and particle motion can only be predicted from sound pressure in very deep water far away from any surfaces. Thus, unless experimental tanks are highly specialized (and very expensive!), the only way to truly measure what fishes hear is in the field—thereby adding a whole level of complexity to research (much of this field work is nicely reviewed in Hawkins and Chapman, 2020).

However, despite this issue, future studies of fish hearing must focus on particle motion since every species (including sharks) detect that and do not directly detect sound pressure (Popper and Hawkins, 2018, 2021; Popper , 2022). Indeed, far fewer species, whose swim bladder comes close to, or connects to the ear (in some way), as per the otophysans, can detect sound pressure in any useful way.

There is also growing interest in the idea that sounds, and particularly those from anthropogenic sources such as pile driving, wind farms, and seismic exploration inject sound into the substrate, and this energy is likely detectable by fishes and aquatic invertebrates that live in, on, or just above the bottom (e.g., Hawkins , 2021; Meekan , 2021; Roberts and Howard, 2022). It is clear, based on limited data, that in examining sound detection and fishes, substrate vibration will need to be taken into consideration for numerous species.

Very clearly, there are many other questions that are exciting and need to be explored. Just to mention a few, without any discussion, these include:

• What are the most critical uses of sound by fishes – is it frequency and sensitivity (unlikely!), or more complex functions such as discrimination, localization (mentioned earlier), and avoiding masking?

• How influential has acoustic communication, which is found in numerous fish species, been in the evolution of hearing? Are there differences in hearing capabilities and sound processing in species that communicate acoustically compared to those, like the goldfish, that do not produce sounds?

• Why did species, such as the Otophysi and Clupeiformes, evolve high-frequency hearing? Is this related to the species evolving in very shallow water (e.g., rivers, streams) where low frequencies do not propagate long distances (e.g., Rogers and Cox, 1988)?

• While many species produce and use sound for communication (e.g., Rice , 2022), numerous other species hear, but do not produce or use sound for communication with other fishes. How do the “silent” fishes use sound and is it used in the same way as in soniferous species?

As mentioned earlier, the goal of this paper is to present my very personal ideas about fish hearing. I do suspect that my colleagues interested in fish hearing would have other, though probably overlapping, ideas as to the most important and interesting questions. My set of issues is based on my more than 57 years working with fishes. I also bring to this paper my background and training and my fascination with comparative issues (thanks Doug Webster and Bill Tavolga!). So, while my work for the past 25 years or so has been heavily (but not exclusively) focused on “translational bioacoustics,” my real scholarly love is comparative hearing, and trying to ask questions that will unravel the breadth and depth of hearing by fishes. Hopefully, some students, at some point, will read this paper and start to think comparatively, and be in a good position to really tackle what must be some of the most interesting and exciting questions in (fish) hearing—the meaning of the extraordinary diversity in the auditory systems of fishes!

I have been fortunate to have a career filled with talented and outstanding students and colleagues who have contributed immensely to the work from my lab cited in this paper, and to the ideas that I am sharing. I also want to thank several funding agencies who supported much of the work that I discuss in this paper. This includes the National Institutes of Health – the source of my very first grant in 1971, the National Science Foundation, and the Office of Naval Research. I thank several long-time friends, Sheryl Coombs, Michael Fine, Dennis Higgs, and Christopher Platt for reading the MS critically and making invaluable suggestions for improving it. I have also, as readers will glean, had the opportunity to know and to work and collaborate with many of the giants of fish bioacoustics over the past 60 years or so. I dare not mention names (though I will mention four great colleagues and friends, Dick Fay, Tony Hawkins, Chris Platt, and Bill Tavolga!) for fear of missing people—but they know who they are, and I thank them for allowing me the honor of working with them. Finally, I deeply thank my wife Helen, and our daughters Michelle Popper Levit and Melissa Popper Levinsohn, for supporting, encouraging, and, quite often, tolerating, my work. I am fortunate to have these three amazing women (and our daughters' spouses and children) in my life. I particularly want to thank Helen for being the consummate scientific editor. Helen started editing my writing with my doctoral dissertation (and typed it on a Corona portable electric typewriter—which we still have—using carbon paper), and she continues to this day (after more than 54 years of our being together) to be the toughest, and very best, critic and editor anyone could ever want.

Arthur Popper has no conflict of interest.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.