Keywords

1 Introduction

It has been almost 60 years since the first reports on the sonic behavior of cichlids. Fryer and Iles (1972) reviewed the discoveries by Bauer (1963), Myrberg et al. (1965), and Rodman (1966) and suggested that the use of sounds by cichlids may be important as a means of species recognition and mate location. However, research on cichlid bioacoustics progressed slowly until about the year 2000 (Fig. 1, earlier reviews of cichlid sounds include Lobel 2001; Amorim et al. 2004 and Longrie et al. 2013).

Fig. 1
figure 1

Summary of scientific papers describing sound production in cichlids per year (black columns) and cumulatively (gray line). Citations for papers that describe the production of sounds or behavioral use of sounds (N = 51) were discovered through literature searches using the Web of Science database (with coverage from 1864-present), and the literature cited list of relevant papers (e.g., Baerends and Baerends-van Roon 1950 in the case of Raj 1916)

Full size image

Historically, research was hindered by the expense and difficulty of using hydrophones and the technology for acoustic analyses, but these former obstacles have been eliminated by modern video camcorders, personal computers, and especially new software for acoustic analysis. The first underwater study of the sonic behavior by cichlids in the wild showed that cichlids were indeed quite acoustically active in Lake Malawi and sounds differed among neighboring species (Lobel 1998). Since then, cichlid bioacoustics has become an active topic for research. Another field study, also in Lake Malawi , documented possible population variation in cichlids calls over a geographic range (Danley et al. 2012). The nature/nurture question was examined in Oreochromis niloticus, and suggested that the ontogeny of sound production in juveniles was innate and not learned (Longrie et al. 2008). Acoustic features that could provide signature differences were examined by Bertucci et al. (2012a, b). The Tanganyikan cichlid, Neolamprologus pulcher , has been proposed as a silent species based on the lack of sounds recorded under laboratory conditions (Pisanski et al. 2015), prompting a need for closer investigation of acoustic activity in other species. Laboratory experiments have shown using acoustic playback experiments that increases in noise can impact behavior by reducing behavioral activity levels that normally coincide with sound production (Slabbekoorn et al. 2012).

It has been well known for a long time that cichlids display complex behaviors during the behavior of mate selection (Baerends and Baerends-van Roon 1950; Keenleyside 1991; Barlow 2000). Cichlids are clearly multi-modal in their communication (Rosenthal and Lobel 2006; Amorim et al. 2008; Verzijden et al. 2010). Vision and coloration are key in the mate selection process for the Pseudotropheus zebra species complex of Lake Malawi (Couldridge and Alexander 2002). But visual cues are not necessarily the sole communication modality, nonvisual cues are implicated in experimental studies (Blais et al. 2009). Recent reviews have focused on the importance of visual cues and color pattern s in the evolution of cichlid and other fishes (Rowland 1999; Seehausen et al. 1999). While visual cues are undoubtedly important, olfactory and acoustic s cues also play significant roles. For example, Chien (1973) determined that the combination of visual and olfactory cues presented together increased spawning rates in the cichlid, Pterophyllum spp., above the rates observed when only one cue was presented; i.e., the effects of the various stimuli are additive (Rowland 1999). Visual cues are apparently not the primary cue in all species examined. Furthermore, chemical cues have received attention as being important in cichlid communication (Maruska and Fernald 2012). The exact role of acoustic s in this mate selection process by cichlids is an ongoing hot topic today. During courtship, females prefer acoustic over silent males but the visual presence of the male is required (Estramil et al. 2014). Sounds and visual cues are synchronous during agonistic encounters (Bertucci et al. 2010). Blais et al. (2009) determined experimentally that visual signals alone cannot explain mate choice and that other signals must also be involved as a factor in female choice. However, they did not assess other cues such as olfaction or acoustic s.

Cichlid fishes are widely studied with regard to understanding the processes of speciation (e.g., Sturmbauer 1998; Kornfield and Smith 2000; Salzburger and Meyer 2004; Kocher 2004; Salzburger 2009). The rapid and extensive species radiation (“species flocks”) seen in cichlids is unrivaled in other vertebrate groups (Liem 1973; Echelle and Kornfield 1984; Barlow 2000; Kornfield and Smith 2000; Salzburger and Meyer 2004; Turner 1999; Turner et al. 2001). One underlying mechanism for this evolutionary success is the cichlid’s complex pharyngeal jaw, a highly adaptable morphological feature (Liem 1973, 1991) coupled with the cichlid’s complex mating behavior, short generation time with numerous offspring. Cichlids provide an opportunity to observe speciation at different stages of progression, allowing investigation of the process and mechanisms of evolution (Kornfield and Smith 2000; Kocher 2004). Species divergence in closely related sympatric cichlids, especially in the Great Lakes of Africa has been hypothesized to have occurred, in part, due to assortative mating. Cichlid mate choice studies confirm that assortative mating does occur in several species in the field and in captivity (Blais et al. 2009; Egger et al. 2008; Knight and Turner 2004; Salzburger et al. 2006). The key to understanding this process is to decipher the factors involved in cichlid mate selection. The influences of visual and chemical cues as prezygotic isolating mechanisms have been examined (Couldridge and Alexander 2002; Blais et al. 2009; Kidd et al. 2006; Maan et al. 2004; Plenderleith et al. 2005). Acoustic signals have not yet been similarly investigated.

This review will highlight the possible morphological mechanisms and biological significance for the occurrence and variability of sounds produced by cichlid species during aggression and courtship. Sounds are also associated with other behaviors, particularly feeding and swimming, which are not necessarily intentionally produced. The essential scientific question is whether certain sound patterns produced by cichlids are ethologically meaningful. For reviews of fish bioacoustics and communication including terminology (see Moulton 1960; Demski et al. 1973; Ladich 1997; Kasumyan 2009; Ladich 2015; Blaxter 1981; Fine et al. 1977; Hawkins 1986; Hawkins and Myrberg 1983; Myrberg 1980, 1981; Myrberg et al. 1978; Popper and Fay 1973, 1993; Schwarz 1985) for cichlid communication (see Nelissen 1991; Lobel 2001; Amorim et al. 2004; and Longrie et al. 2013) and for animal communication in general (see Bradbury and Vehrencamp 1998). Sound production is widespread throughout the phylogeny of fishes with evidence of sound production from 72 families (Rice et al. 2020).

2 Checklist of Sound Producing Cichlid Species

There are now a total of 39 cichlid species identified as sound producers (Table 1), and this represents an increase of an additional 19 species since Lobel (2001). One of the most conspicuous results is that all the recent studies have been on African species. Surprisingly, the last acoustic study of a Central/South American cichlid species was by Schwarz in 1980! The primary literature for cichlid acoustic biology numbers 55 publications (as of 2019). Table 1 lists the documented sound producing cichlids using the scientific name as published; and it cross-references to the most recent scientific name in current usage in FishBase (http://www.Fishbase.org), which is cross-linked to the definitive reference for current fish taxonomy, the Eschmeyer Catalog of Fishes (http://www.calacademy.org/scientists/projects/catalog-of-fishes). Cichlid species nomenclature is frequently changing and is the subject of a great deal of scientific discussion. In this review, we will try to use the most referenced names or the name cited in the original literature. Our purpose is not to update the taxonomy but only to clearly specify which species is being referenced.

Table 1 Sound producing cichlids list of synonymies
Full size table

The six specific questions regarding cichlid bioacoustics include:

  1. 1.

    Are specific sounds associated with specific behaviors?

  2. 2.

    What are the temporal patterns of calling activity?

  3. 3.

    What morphological structures produce the sounds?

  4. 4.

    What are the key characteristics of sound patterns that distinguish species?

  5. 5.

    Is there any relationship between body size and call characteristics?

  6. 6.

    How might an increase in ambient noise confound acoustic communication ?

2.1 Calling Activity and Caller Sex

The behavioral processes of courtship and spawning are similar across cichlid species, often involving approach, leading, circling, and quivering (Baerends and Baerends-van Roon 1950; Barlow 2000; Ripley and Lobel 2004, 2005). Although the specifics of circling behavior may differ between genera (Stauffer et al. 1993), “quivering” is associated with circling and courtship in many cichlids as they swim around each other and ripple their abdominal muscles. Sound production correlates with quivering in Tramitichromis intermedius, Copidichromis conophoros (Lobel 1998; Ripley and Lobel 2004), Simochromis diagramma (Nelissen 1975), and in Pseudotropheus species (Amorim et al. 2004, 2008). It is likely that many cichlids that share this quiver behavior will also produce sounds. One known exception, Oreochromis mossambicus , does not produce sound during the specific quiver behavior but it does produce sounds during other courtship behavior (Amorim et al. 2003). Whether a fish produces sound during quivering or other behaviors, sound production appears to be regularly associated with cichlid courtship (Lobel 1998; Amorim et al. 2003, 2004, 2008). The hypothesis is that female fish could use these courtship-associated sounds as one cue to assess a potential mate’s identity and quality. Actual spawning associated sounds are hypothesized to coordinate reproductive timing synchronizing gamete release.

Diel patterns of sound production in cichlids have been suggested for two cichlid species, both exhibiting a diurnal peak (Brown and Marshall 1978; Longrie et al. 2013). For Herotilapia multispinosa , sounds were enumerated during reproductive and agonistic behaviors during different stages of the reproductive period. Spawning is reported to be restricted to daylight hours (Brown and Marshall 1978), but no monitoring of sounds was done nocturnally. In Oreochromis niloticus , direct acoustic sampling of fishes over a 24-h period found peak diurnal calling activity (Longrie et al. 2013). This is an aspect of future research well worth examining in more detail. Early pre-dawn calling by males could be a stimulus for spawning. The courtship sound produced by male Oreochromis mossambicus was reported to facilitate oviposition in females (Marshall 1972). It is possible that some cichlids may be sonically active before daylight. Smith (2007) recorded cichlid-like sounds in the field in areas where cichlid species were present and found nocturnal peaks in the production of these sounds. On coral reefs, damselfishes (Pomacentridae) were once thought to be only acoustic ally active during daytime until 24-h recordings revealed pre-dawn peaks in chorus activity (Mann and Lobel 1995). It makes sense that acoustic signals would be used in low light periods, possibly as an early cue to stimulate females to initiate egg hydration prior to spawning. The technology and software for underwater acoustic s have greatly advanced in recent years and allows for documenting the natural cycle of sound production for sonic fishes in the wild as well as assessing the impacts from noise pollution (Lindseth and Lobel 2018).

For species whose social behavioral context for sound production are known (N = 39; Table 2) as of the year 2019, 20 species produced sounds during agonistic intraspecific encounters and 30 spp. in reproductive display context (Table 2). For studies of reproductive behaviors, in all species the callers were male when sex was identified in all except Pelmatolapia mariae where the caller was female. Acoustic male courtship displays are known or proposed in 21 (55%) of these (Table 2). For reproduction-associated contexts monitored with a hydrophone, 30 species (79%) produce sounds in reproductive context during various aspects of nest visiting male calling in association with nest, or male courtship display to female (n = 20, 53%) or courtship encounters with unspecified sex (n = 5, 13%). Female calling in Pelmatolapia mariae was recorded in the vicinity of the nest area (n = 1, 3%). Of these, the males of four species (11%) were observed to engage in courtship with a visiting female but no sounds were observed before the recordings began which were limited to passive recording of sounds (no video) from the focal male’s nest after the field observations. Both male and female cichlids have been observed to produce sounds in some species (see Table 2) although sound production by both sexes may not be universal. It should be noted that acoustic studies in fishes have largely focused on male behaviors even though in many fishes a mechanism of known sound production is present in males and females of the species (Ladich 2015). Among those cichlids where males and females have been noted to produce sounds in the context of aggression: 50% of species with specifically identified sex of caller in agonism were female and 70% were male (N = 20 species).

Table 2 Behavioral context of sound production in cichlids
Full size table

The pattern of sound production with reproductive behavior has not always been clear and more research is needed to define if such sounds are meaningful to the fishes. For example, in Oreochromis niloticus females and males produced agonistic sounds but no sounds were detected during courtship from either sex (Longrie et al. 2013). Males of H. multispinosa produced abundant sounds in both agonistic and reproductive contexts (Brown and Marshall 1978). In this species, males and females produced sounds throughout the egg, larval, and free-swimming young stages of offspring development during agonistic encounters with intruders. There are only five studies that statistically describe the temporal characteristics of female sounds and associated behavior in detail. These include pulsed sounds while mouthbrooding embryos for Oreochromis niloticus (Longrie et al. 2013); broadband frequency 2-pulsed sound in agonism for Neolamprologus pulcher (Spinks et al. 2017), interpulsed sounds in agonism for Maylandia zebra (Simoes et al. 2008b); and single-pulse sounds in the vicinity of nest with males for Pelmatolapia mariae (Kottege et al. 2015). Myrberg et al. (1965) described the pulse rate for Hemichromis bimaculatus female agonistic sounds during the parental care period.

Agonistic sounds between females were also described for Maylandia zebra (Simoes et al. 2008b). Female sounds were shorter in duration and consisted of fewer pulses than males (Simoes et al. 2008b). Males produce sounds during courtship-associated displays to females, and females seem to be silent in this context for video documented studies (e.g., Ripley and Lobel 2004). M. zebra juveniles are active sound producers during agonistic interactions, producing isolated pulses early in development (Bertucci et al. 2012a, b). There were two species (6%) where the callers also included a female although their sounds were observed anecdotally and not digitally recorded (n = 22 species; sex of caller specifically examined). Females of Tramitichromis intermedius , Archocentrus centrarchus , Oreochromi mossambicus , and Astatotilapia burtoni were observed as silent (Lanzing 1974; Ripley and Lobel 2004; Maruska et al. 2012). It is important to note that the sex of the caller is often difficult to determine exactly when recording fishes in groups and with omnidirectional hydrophones. Thus, this aspect deserves closer study.

Sounds of Juvenile Cichlids

The ontogeny of sound production in cichlids has only received limited study to date. The Malawi cichlid , M. zebra juveniles are active sound producers during agonistic interactions, producing isolated pulses early in development (Bertucci et al. 2012a, b). Young Oreochromis niloticus , also produce sounds, supporting the hypothesis that sound production is not learned in cichlids (Longrie et al. 2008). The onset of vocal ability in young males of Tramitochromis intermedius was detected at about 7 months of age (Ripley and Lobel 2004).

3 Sound Producing Mechanism

Definitive experiments have not yet demonstrated which muscles and bones are used for sound production by cichlid fishes. The pharyngeal bones of some fishes have been implicated as the sound-producing instrument (Darwin 1874; Marshall 1962) and the same has been proposed for cichlids (Lanzing 1974; Rice et al. 2001; Rice and Lobel 2002, 2003; Marshall 1962; Nelissen 1977). Sounds produced by the pharyngeal apparatus are presumably amplified by the swimbladder (Rowland 1978; Marshall 1962). Cichlids produce a clear and distinct sound while feeding and grinding food in the pharyngeal mill, as well as during courtship and agonistic behavior (Lobel 2001; Lanzing 1974; Nelissen 1977; Rodman 1966). However, an alternative mechanism called the “buckling method” (an indirect swimbladder associated muscle mechanism) has been proposed by Longrie et al. (2009), and is discussed below. Not all sounds may be volitional, and any behavioral significance has not yet been demonstrated. For example, the agonistic “thump” described for Herotilapia multispinosa by Brown and Marshall (1978) coincides with a head jerking upward and downward but also coincided with well-established agonistic frontal displaying suggesting a possible volitional nature to the sounds. Sounds were also produced during jaw snapping during agonistic behaviors (Baylis 1974; Brown and Marshall 1978). Brown and Marshall (1978) described non-pulsed “thump” sounds as most likely incidental to behaviors and not communicative. Jaw snapping is proposed as a mechanism for the reproductive season of sound production in females for Tilapia mariae (Kottege et al. 2015). These sounds could be incidental to body movements or intentional acoustic components of displays. The specific context for these sounds was less ritualized and less tightly linked to a specific body posture or behavioral display than is typically found in other cichlids.

If cichlid pharyngeal jaws are involved in sound production , as preliminary evidence indicates, then perhaps it is possible that different pharyngeal tooth morphologies may produce different types of sounds (i.e., sounds with different individual pulse waveforms, e.g., Lobel 2001). Of course, the key experiment would be to determine if such different pulse waveforms produce different sound qualities that are directly detectable by a fish. It is possible that the simplest type of acoustic signal produced is the specific sound of individual pulses, which are then frequently repeated (Fig. 2). It may be the perceived sound of the pulse in combination with pulse rate timing that may contribute to species recognition. Prior studies in other fishes, especially the damselfishes (Pomacentridae), show that pulse number and/or repetition rate is closely correlated with the behavioral context of sounds (Hawkins 1986; Myrberg 1981, 1997; Olivier et al. 2015; Spanier 1979). Pomacentrids that produce similar sounds to cichlids have had their sound production mechanism described as including jaw element movements similar to the feeding process (Olivier et al. 2015; Parmentier and Fine 2016).

Fig. 2
figure 2

Example of the waveform patterns from three cichlids in Lake Malawi. (a) Maylandia zebra , (b) Mchenga conophorus , and (c) Tramitochromis intermedius . The simplest type of acoustic signal produced is the specific sound of individual pulses, which are then frequently repeated. Field recordings made underwater of free-living fishes in their natural habitat (depth 3–8 m) at Cape McClear, Lake Malawi National Park offshore of the World Wildlife Fund Education Center and at Otter Point, August 5–27, 1990 (Lobel 1998)

Full size image

Pharyngeal jaws in cichlids are well known for their advanced capability for complex muscle modulation and bone movement (Liem 1973, 1991). Four cichlid species (Tables 3 and 4) produce sounds that exhibit more complex amplitude modulation patterns and pulse repetition rates than ones produced by many other fish families with simpler and probably less manipulatable pharyngeal apparatuses. For example, the morphology of the pharyngeal jaw complex suggested a close relationship between pomacentrids (damselfishes) and cichlids within Labroidei (Wainwright et al. 2012; Lauder and Liem 1983), although this close relationship has been overturned in recent molecular phylogenies (e.g., Betancur-R et al. 2017). Cichlids differ significantly from other fishes in the muscle and bone architecture of their pharyngeal apparatus (Liem and Greenwood 1981; Kaufman and Liem 1982; Seehausen et al. 1999; Stiassny and Jensen 1987).

Table 3 Acoustic traits of cichlid species sounds associated with agonistic and reproductive behavior. Frequency range was reported if no other frequency data were available
Full size table
Table 4 Bioacoustic research topics
Full size table

If this difference in pharyngeal morphology is a basis for sound production, then presumably pomacentrids may not be able to forcibly occlude or grind the pharyngeal jaws in the way that the cichlids do. The pomacentrid sound production mechanism has been described as resulting from jaw element movements similar to the feeding process (Parmentier et al. 2007; Parmentier and Fine 2016; Olivier et al. 2017). Sound production in pomacentrids has been extensively described (Lobel and Kerr 1999; Lobel and Mann 1995; Myrberg et al. 1978, 1986; Myrberg and Riggio 1985; Mann and Lobel 1995, 1997, 1998; Spanier 1979). However, the mechanism of sound production still needs to be studied across species.

3.1 Functional Morphology of the Cichlid Sound Mechanisms

Despite the recent research describing cichlid sounds and behavior, the morphological mechanism(s) responsible for these sounds still requires detailed study. In many sonic fish species, there are clear morphological adaptations and specializations dedicated to sound production (e.g., Bass and Ladich 2008; Kasumyan 2008). These sonic mechanisms involve the swimbladder (e.g., Alexander 1966; Fange 1966; Parmentier and Diogo 2006), different bones or fins (Fine et al. 1996; Parmentier et al. 2010), connective tissues (Parmentier et al. 2007), or in some cases multiple mechanisms (Kaatz 2002; Kaatz et al. 2017; Sörensen 1894–1895). In some species of fish, such as the rockfishes (Family Scorpaenidae), their sonic ability was initially hypothesized or inferred based on the identification of well-developed sonic muscles (Hallacher 1974), and later confirmed through acoustic recordings (Širović and Demer 2009). Even though it is now recognized that the occurrence of sounds in cichlids is widespread, the exact identity of the sonic mechanism is elusive with a few different hypotheses proposed or initially tested in the literature (Rice and Lobel 2003; Parmentier and Fine 2016).

The acoustic properties of fish sounds can be suggestive of the sonic mechanism: muscle-driven sounds typically have harmonic frequency structure, and variable call durations, while stridulatory sounds (produced by the grinding of hard surfaces) are broadband and shorter duration (Demski et al. 1973). However, these spectral and temporal properties of sounds are likely endpoints along a continuum, as there are many examples of muscle driven sounds that produce short duration pulses with little spectral structure, such as in red drum (Family Sciaenidae) (Guest and Lasswell 1978). Since most fishes are ectothermic, the frequency in sounds produced by muscles has a temperature-dependent response: warmer water temperatures increase the contraction speed of sonic muscles, which increases the frequency of the call (Connaughton et al. 2002; McKibben and Bass 1998). In the case of toadfishes (Family Batrachoididae), the contraction rate of the sonic muscle sets the fundamental frequency of the call (McKibben and Bass 1998).

The cichlid sonic mechanism is difficult to identify based only upon the acoustic properties of the cichlid sounds. The majority of cichlid sounds are relatively short duration (duration <1 s) and low frequency (dominant frequencies <1000 Hz). Based on the conceptual framework of Demski et al. (1973), these types of sound could be produced by either muscular or osteological mechanisms. Furthermore, there are aspects of the sounds, such as harmonic content (reviewed in Rice and Lobel 2003) and stereotypic acoustic features (Lobel 1998; Rice and Lobel 2003) that make the case even more ambiguous. Despite the number of publications documenting sounds produced by different cichlid species, comparatively few have examined the functional morphology . To date, two morphological systems have been hypothesized to be the source of cichlid sounds: pharyngeal teeth with associated musculature, which is then amplified by the swimbladder , and the buckling method involving abdominal musculature impinging directly on the swimbladder (Longrie et al. 2009).

3.2 Pharyngeal Teeth

The earliest hypothesis for the sonic mechanism in cichlids focused on the pharyngeal teeth (Lanzing 1974; Lanzing and Higginbotham 1976). The cichlid pharyngeal jaws have been suggested as a critical evolutionary innovation responsible for driving the ecological success and diversification of the family (Liem 1973, 1978, 1979; Liem and Greenwood 1981; Liem and Osse 1975). Early observations of cichlid noted that sounds associated with feeding behavior showed similar acoustic properties to communicative sounds (Lanzing 1974); however, some species show differences in the acoustic properties between these sound types (Longrie et al. 2009).

The shapes of cichlid pharyngeal teeth are intricate and elaborate (Casciotta and Arratia 1993; Lanzing and Higginbotham 1976), and show a diet-dependent phenotypic plasticity (Huysseune 1995; Smits et al. 1996a; Smits et al. 1996b; Trapani 2003). Many of the upper pharyngeal jaw teeth have “hooks,” whereas the bottom teeth are more of flat surface (Lanzing and Higginbotham 1976); for example, see Fig. 3. Lanzing and Higginbotham (1976) suggested that the divergent morphology of upper and lower pharyngeal jaw teeth form a “hook and anvil arrangement” and could be involved in the production of sounds. Sounds created by occluding teeth would presumably produce short duration, broadband sounds, with relatively higher frequencies. A similar sonic mechanism has been suggested in jacks (Carangidae; Taylor and Mansueti 1960).

Fig. 3
figure 3

Detail of the pharyngeal teeth from a Tramitochromis intermedius male. The upper and lower pharyngeal jaw teeth form a “hook and anvil arrangement.” The magnification on the lower PJ is 15x and the magnification on the upper PJ is 19x. SEMs taken by A. Rice using a JEOL JSM-840 SEM

Full size image

3.3 Pharyngeal Jaw Complex

One of the hypothesized evolutionary innovations and keys to the cichlid adaptive radiations in both the African Rift Lakes and the Neotropics is their highly developed pharyngeal jaw apparatus (e.g., Burress 2016; Liem 1973; Wainwright et al. 2012). The jaws are suspended and controlled by a suite of pharyngeal jaw muscles that are responsible for jaw occlusion, lateral, and anterior-posterior movement (Claes and De Vree 1991; Liem 1973). Intriguingly, the pharyngeal jaw muscles of cichlid fish have different physiological properties (Rice et al. 2001), which are also sexually dimorphic (Rice and Lobel 2002). This finding suggests the hypothesis that one sex is more morphologically adapted to produce sounds than is the other sex. Other sonic species of fish, where primarily the males produce the sounds, show a sexual dimorphism in sonic muscle morphology or physiology (Brantley et al. 1993; Walsh et al. 1987; Walsh et al. 1989) or brain regions controlling sound production (e.g., Bass and Andersen 1991; Bass and Baker 1990; Knapp et al. 1999). Thus, the corroborating evidence of physiological difference in pharyngeal jaw muscle performance, sexual dimorphisms in muscle physiology, along with the intricate movements of pharyngeal jaws sets the stage for in vivo physiological experiments to demonstrate the degree to which the pharyngeal jaw complex is involved in cichlid sound production (Rice and Lobel 2003).

Examples of sounds from one male Tramitichromis intermedius support the notion that the pharyngeal jaws are involved in sound production and that such sounds are amplified by the swimbladder . A typical courtship sound is illustrated in Fig. 4a showing both the series of pulses in a single call (Fig. 4a) and the detailed waveform structure of two pulses (Fig. 4b). A similar analysis involved sounds recorded while the same fish consumed small sand-dwelling prey. In the audio–video recording, it was easily observable that the fish was manipulating and grinding a hard-bodied prey in its buccal cavity. The series of food grinding sounds are shown with corresponding oscillographic patterns (Fig. 4c). Detailed waveform structure of two grinding sound pulses is shown in Fig. 4d. Courtship sound s are distinguished by distinct and consistently repetitive pulsed patterns with defined temporal meristics and repeatable pulse waveforms. Grinding sounds are highly irregular with inconsistent properties and appear much like simple noise type signals . Significantly, the two types of sounds from the same individual fish displayed the identical dominant frequency at 517 Hz indicating that amplification and resonance by the swimbladder were involved.

Fig. 4
figure 4

Examples of courtship and grinding food sounds from one male Tramitichromis intermedius and aggressive sounds from Copadichromis conophorus. A typical courtship sound showing both the series of pulses in a single call composed of 10 pulses and overall duration of about 195.5 ms (a) and the detailed waveform structure of two pulses (b), pulse durations 6.69 and 7.33 ms separated by an interpulse interval of 8.54 ms. The same fish grinding consumed small sand-dwelling prey (c, d). In the audio–video recording, it was easily observable that the fish was manipulating and grinding a hard-bodied prey in its buccal cavity. Oscillographic of the entire food grinding sequence of 227 ms duration. Courtship call and food grinding produce the same dominant frequency at 517 Hz. (d) Detail of the individual grinding sounds waveform, indicated by arrows. The first pulse is 8.17 ms duration; interpulse interval 5.74 ms, second pulse is 10.32 ms followed by 10.75 ms interpulse interval and third pulse is 7.17 ms. (e, f, g) Two bursts of an “aggressive sound” for comparison that was made by Mchenga (Copadichromis) conophorus: The first (left) is 351.2 ms duration and the second (right) is 168.7 ms

Full size image

3.4 Fin-Girdle Muscles (Buckling Mechanism)

Research on Oreochromis niloticus raises the possibility of another sonic mechanism in cichlids termed “buckling” (Longrie et al. 2009). O. niloticus produces short duration, pulsed sounds in agonistic contexts (Longrie et al. 2009). This sound production coincided with posterior movement of the pectoral and pelvic girdles (Longrie et al. 2009). Electrical stimulation of the lateral body in anesthetized fish also resulted in sound production , with a decreased amplitude in fish with a deflated swimbladder (Longrie et al. 2009). Using the kinematic and electrostimulation data combined with morphological inspection, Longrie et al. (2009) suggest that the vesica longitudinalis muscle, which originates on the post-cleithrum, runs ventral to the swimbladder and inserts on the second anal fin pterygiophore is the candidate sonic mechanism (Longrie et al. 2009). This work raises interesting questions as to whether the vesica longitudinalis is involved in sound production more broadly across cichlid species, or whether it is an independent evolutionary innovation in tilapiine fish. It also raises the question of whether silent cichlids would represent a secondary loss of a vocal mechanism or independent origins of different vocal mechanisms.

3.5 Evolutionary Implications

If cichlid pharyngeal jaws are involved in sound production , as preliminary evidence indicates, then we can expect that different pharyngeal jaw morphologies may produce different types of sounds. It remains to be determined if such different pulse waveforms reflect different types of sounds that are directly detectable by a fish. It is possible that the simplest type of acoustic signal produced is the specific sound of individual pulses, which are then frequently repeated. It may be the perceived sound of the pulse in combination with pulse rate timing that may contribute to species recognition. It is clear from prior studies that pulse number and/or repetition rate is closely correlated with the behavioral context of sounds.

Cichlids have attracted a great deal of attention as an adaptive radiation, particularly within the African rift lakes, with a spectacular diversity of behavior, ecology, and morphology (Albertson et al. 1999; Galis and Metz 1998; Kornfield and Smith 2000). It is no surprise then, that this lineage would show a widespread variety in sound production, and potentially sonic mechanisms. It is fascinating to investigate what is the cichlid sound producing mechanism(s). It has the potential to illustrate the evolution of acoustic diversity, especially among sympatric species. This research could potentially reveal the genetic mechanisms that are foundational to morphological specializations (Streelman et al. 2007), and has been proposed for the diversity of cichlid head and jaw morphologies (Albertson and Kocher 2006; Albertson et al. 2003).

4 Sound Characteristics

Many, but not all, cichlids produce a soft “purr” or “drumming” sound while performing courtship behavior, often accompanying the “quiver” display (Rowland 1978; Lobel 1998; Ripley and Lobel 2004; Brown and Marshall 1978). Fish sounds can be broadly categorized along a spectrum from continuous “long” sound to a series of short pulses or notes (Lobel et al. 2010). For cichlid species whose temporal patterns have been analyzed to date, most species consist of a series of regular pulses or notes with regular interpulse intervals (Table 3). Exceptions include in H. multispinosa, agonistic sounds are continuous, whereas in M. conophoros agonistic sounds are comprised of pulses with irregular interpulse intervals. Grunting and clicking species distinguished by Van Staaden and Smith (2011), varied in interpulse interval suggesting some species may have more continuous call patterns than others but the noise level of the recording environment precludes this conclusion. In the genus Maylandia , males have been documented to produce low amplitude, frequency-modulated sounds in addition to higher amplitude pulsed sounds (Simoes et al. 2008b). Hemichromis bimaculatus produced two different sound types in the same display (Rowland 1978). Single-pulse sound types were additionally produced by Hemichromis bimaculatus , Herotilapia multispinosa , Maylandia zebra juveniles, and Oreochromis mossambicus. However, some examples of sounds from these fishes appear to include occasional pulsed sections. An acoustic repertoire of a cichlid can range from one to four sound types per species. Typically, a male will display and produce sounds in the presence of a female pre-reproductively as a female approaches the spawning site. Examples of such sonic behavior while courting are shown for Mchenga (Copadichromis) conophorus (Fig. 5) and T. intermedius (Fig. 6). These same fish are shown in their natural habitat (Figs. 7 and 8).

Fig. 5
figure 5

Courtship sequence shown for Mchenga (Copadichromis) conophorus (Aug 23, 1990; 08:18:48–08:19:05 Hrs). Female enters bower and spawns then leaves the arena. See Fig. 7 for the fish recorded

Full size image
Fig. 6
figure 6

Courtship sequence shown for Tramitichromis intermedius (Aug 22, 1990; 07:51–07:51:30 Hrs). A single male courting a single female as she swims casually passing-by, but she does not stop. See Fig. 8 for the fish recorded

Full size image
Fig. 7
figure 7

(a), A male Mchenga (Copadichromis) conophorus defending his bower from an intruder male. (b) Male M. conophorus and (c) Female M. conophorus shown holding embryos in her mouth. Photographed at Cape McClear, Lake Malawi National Park offshore (depth 3–8 m) of the World Wildlife Fund Education Center and at Otter Point, August 5–27, 1990 (photos by P. S. Lobel)

Full size image
Fig. 8
figure 8

(a) A male Tramitichromis intermedius with several females in background. (b) A male patrolling the rim of his bower. (c) A male in courtship colors. (d) A female. Photographed at Cape McClear, Lake Malawi National Park offshore (depth 3–8 m) of the World Wildlife Fund Education Center and at Otter Point, August 5–27, 1990 (photos by P. S. Lobel)

Full size image

4.1 Acoustic Patterns

The spectral and temporal sound measurements for cichlids are summarized in Table 3. Of the 39 cichlid species whose sounds have been acoustic ally detected with a hydrophone or analyzed to some extent, the three cichlids with few data are: Amphilophus citrinellus , Chindongo elongatus , and Sarotherodon galilaeus. Among the other 35 cichlid species whose sounds have been analyzed to some extent, six species have had pulse rate described and seven species have had call rate described. Additional acoustic traits have been reported for 30 species (Table 3) but not every species had all characteristics measured. Pterophyllum sp. has only had pulse rate (6 pulses/sec) and frequency described; Amatitlania nigrofasciata has only had rate (20 pulses/sec) and frequency described; two species, Labidochromis caeruleus and Cichlasoma portalegrense , have only had frequency described. Pelmatolapia mariae had sounds described in general but without temporal or frequency data listed (Table 3).

4.2 Typical Sound Characteristics

Pulsed sound types best characterize cichlid sounds and it is intriguing to contemplate whether this temporal information is meaningful in communication (e.g., Kihslinger and Klimley 2002). The number of species for which pulsed sounds (excluding chewing sounds) were statistically described were: duration (n = 26); pulse number (n = 26); pulse duration (n = 19); pulse period (n = 10); and interpulse interval (n = 4). Temporal sound traits for these sounds produced during agonism or courtship with statistical descriptions available for species including ranges (that exclude species with high frequency chewing sound types and single pulse sounds) for minimum to maximum range values are: call duration (ms) 277 ± 160 SD to 1319 ± 789 SD (51–4000; n = 28; 16 species; pulse number 6 ± 3 SD to 24 ± 19 SD (2–92, n = 28, 15 species); pulse duration (ms) 9 ± 5 SD to 21 ± 19 SD (3–70, n = 17, 13 species); pulse period (ms) 57 ± 24 SD to 118 ± 45 SD (7–250); n = 17; 6 species); and interpulse interval (ms) 21± 31 SD to 138 ± 122 SD (5–316, n = 5, four species). These values characterize the described sounds of cichlids that are produced in pulse series at a lower frequency range.

Most cichlids produce one sound type per call, which consists of a series of short pulses (Table 3). Not all contexts have been acoustically surveyed in all known acoustic cichlids therefore a statistic for call repertoire would be hypothetical. However, for the species with data, the reported maximum repertoire size for social contexts the conservative estimate for sounds of a certain origin or numerically described ranges from 1 to 4; n = 35 species. Herotilapia multispinosa had the largest repertoire (Brown and Marshall 1978). Non-Pulsed (single or continuous) are known from ten species (Tables 3 and 4) but have only been quantitatively described for three species for natural sounds with comparable range data for H. multispinosa , O. mossambicus , and O. tilapia . These sounds appear to be shorter in duration than pulsed calls. Their temporal duration ranges of minimum to maximum are 119 ± 94 SD (n = 4, three species) to 437 ± 237 SD (n = 4, three species).

A recent review has provided a lexicon for what could constitute different types of complex acoustic behaviors including variation across populations, across or within individuals, across or within contexts (Bolgan et al. 2018). Complex patterning within a call has been documented in two species of cichlids (Table 4). H. bimaculatus agonistic sounds include two sounds that overlapped on spectrograms, a pulsed sound and “thump” which is not sufficiently expanded to determine the pattern, although it lacks pulses on the spectrograms available (Rowland 1978). Male courtship sound s in three other species ( Maylandia emmiltos , M. fainzilberi and a hybrid of Pseudotropheus “zebra gold’ × M. zebra ) consisted of a continuous waveform sometimes frequency-modulated sound followed by a pulsed sound. Increased complexity in vertebrate calls is well established to play a range of significant roles in both intra- and interspecific communication in frogs (Zelick et al. 1999).

Call rate per minute (aka calling rate) has been described for ten species: Archocentrus centrarchus ( Hemichromis bimaculatus (Myrberg et al. 1965; Rowland 1978); Herotilapia multispinosa (Brown and Marshall 1978); Maylandia callainos , Pseudotropheus “zebra gold,” Pseudotropheus emmiltos , Maylandia fainzilberi , and Maylandia zebra (Amorim et al. 2008); Oreochromis mossambicus (Amorim et al. 2003) and Oreochromis niloticus (Longrie et al. 2013). Sounds per minute is a readily comparable sound characteristic that can be compared across species and has been calculated for male courtship calls in six species. Among Maylandia callainos ; M. emmiltos ; M. fainzilberi ; and M. zebra ; Pseudotropheus “zebra gold” (Amorim et al. 2008); M. emmiltos had the highest call rate with 1.4 sounds/min and was statistically higher than the other species (Table 5). Oreochromis mossambicus call rate in courtship was 1.1 ± 07 sounds/min.

Table 5 Acoustic comparisons between sympatric cichlid species
Full size table

Temporal Data

The number of species for which temporal data was reported from reproductive and agonistic calls are: call duration n = 26; pulse number n = 26; pulse duration n = 19; pulse period n = 10; and between pulse interval n = 4. For these sounds minimum and maximum values for temporal traits (all data pooled) are: sound duration (ms) 255 ± 165 SD to 1226 ± 803 SD (27–4000; n = 31; 16 species); pulse number 6 ± 3SD to 23 ± 19SD (2–92; n = 28; 15 species); pulse duration (ms) 9 ± 5SD to 21 ± 19 SD (3–70; n = 17; 13 species); pulse period (ms) 57 ± 24 SD to 118 ± 45 SD (7–250; n = 17; 6 species); and interpulse interval (ms) 21 ± 31 SD to 138 ± 122 SD (5–316; n = 5; four species). The details are given in Table 3.

Frequency Data

Frequency measurements describing cichlid sounds differ across publications with range, fundamental, maximum, center, or dominant (dominant synonyms are peak, principle, primary, and main) frequency values (Table 3). Frequency range measurements are subject to different methods, vary as a function of FFT used in the spectrogram, and these details are not always reported in cichlids. But, since many cichlids sounds are pulsed and broadband, frequency may not be the best way to describe these sounds. The number of species for which social context reproductive or agonistic sound frequency data was reported are: peak frequency n = 29 and frequency range n = 18. Frequency measurements describing cichlid sounds differ across publications with range, maximum, middle, or dominant (peak, principle, main) frequency values published for 31 species (Table 3). Frequency range measurements are subject to different methods and are not widely reported. Some examples for social contexts for minimum and maximum range values are: 95 Hz ±55 to 1385 Hz ± 1202 (25–5000 Hz; n = 16; 10 species). The minimum to maximum dominant frequency was: 493 Hz ± 1095 and the maximum was 1487 Hz ± 3555 (50–16,000, n = 36; 18, species).

Frequency range is low in Oreochromis nilotucus and O. mossambicus compared to some other cichlids. The sounds produced were described as very low, with most of the energy below 200 Hz (Longrie et al. 2008, 2009, 2013; McPherson 2012) and a mean dominant frequency below 68 Hz (46–68 Hz;) compared to other cichlids (Amorim et al. 2003; Longrie et al. 2013). The tonal sounds of several additional species, Maylandia emmiltos , M. fainzilberi , and Pseudotropheus “zebra gold” × M. zebra , are also very low among cichlids (Range 22–86 Hz Peak frequency), which have been recorded to date.

Frequency range is high with peak or main frequency well above 1000 Hz, for Neolamprologus pulcher (agonistic sounds, Spinks et al. 2017), Pterophyllum (scalare?) (Myrberg et al. 1965), and O. mossambicus (Lanzing 1974). High frequency (greater than 1000 Hz) sounds have also been reported from Tropheus moorii , T. duboisi , T. brichardi , Simochromis diagramma , and Astatotilapia burtoni (Nelissen 1978). These pulsed high-frequency range sounds seem to be different sound types from the typical pulsed lower frequency sounds of cichlids. Six of these species had sound ranges described for several sound characteristics and these are summarized by (mean ± SD): sound duration (ms) 370 ± 399 to 3660 ± 1141 (80–5000; n = 5; five species); pulse number 2 ± 1 to 8 ± 3 (1–11; n = 6; six species); pulse duration (ms) 42 ± 8 to 124 + 38 (30–170; n = 5; five species); and dominant frequency (Hz) 3667± 2021 to 12,833 ± 1768 (2500–16,000; n = 3; three species). We excluded the higher frequency chewing call types from our final characterization of cichlid sound call parameters analysis below. It is important to note that while understudied feeding sounds even if incidental can have a significant impact on social behavior suggesting their use as foraging cues as observed in a group foraging cyprinid (Scholz and Ladich 2006).

Fish Size and Dominant Frequency

In general, among teleosts with simple swimbladders, larger fish produce lower frequency sounds than do smaller fish (Demski et al. 1973; Fine et al. 1977; Myrberg and Riggio 1985). This applies to cichlids (Myrberg et al. 1965; Rowland 1978) and also to pomacentrids, which have similar swimbladder structure (Myrberg et al. 1993; Lobel and Mann 1995). The hypothesis is that sound quality in terms of amplitude and frequency is in large part a morphologically determined signal related to swimbladder and body size. It is a basic principle of underwater physics that larger swimbladders resonate at lower frequencies than smaller ones (Clay and Medwin 1977; Urick 1983). This association could provide a reliable indicator of male size to the listening female fish that is selecting among potential mates.

Body size is an important variable affecting ultimate reproductive success of male cichlids and many other fishes (Bisazza and Marin 1991; Hert 1990; Magnhagen and Kvarnema 1989; Myrberg et al. 1986; Noonan 1983). Two studies have demonstrated that, in fact, the dominant frequency of the courtship sound from two pomacentrid species does decrease with increasing male size (Myrberg et al. 1993; Mann and Lobel 1995). A preference for females to mate with larger males, producing lower frequency courtship calls, has been demonstrated experimentally for the bicolor damselfish, Stegastes partitus . Female choice in this pomacentrid is also strongly influenced by the vigorousness of male courtship activity (Myrberg et al. 1986; Knapp and Warner 1991). Vigor is defined by the rate at which males display courtship-swimming motions and simultaneously produce courtship sound s although Knapp and Warner (1991) did not include bioacoustics . An individual cichlid’s sounds differ in call parameters relating to body size (Amorim et al. 2004; Simoes et al. 2008b; Bertucci et al. 2012a, b; Verzijden et al. 2010; Maruska et al. 2012). Frequency patterns could contribute to individual sound differences (Bertucci et al. 2012a, b). However, the frequency may not vary significantly between species (Amorim et al. 2008). It is still an open question as to whether the specific acoustic structure conveys specific information or whether it is merely the occurrence and rapidity of sound production , but not details, that matter for fish mate selection. The simplest information that might be communicated by a fish’s sonic behavior may be an individual male’s location, readiness to spawn, and or his overall vigor (Lobel 2001).

4.3 Aggression and Courtship

The typical sound produced by a cichlid during aggressive and courting behavior is a series of rapid low frequency (typically 200–600 Hz) pulses. The “grunting” and “clicking” cichlid sound types described by Van Staaden and Smith (2011) were distinguished by discernibility of the duration of the interpulse interval. These acoustic traits could provide signature differences among individuals (Bertucci et al. 2012a, b). The territorial cichlid males responded to conspecific sounds but not temporal coding modifications (Bertucci et al. 2013). These preliminary data support the hypothesis that cichlid sounds are temporally structured in a way that contains information that could be used in the mate selection process. This type of acoustic coding has long been documented and is common in insects and amphibians and marine mammals, and has been proposed for fishes as well (Myrberg et al. 1978; Kihslinger and Klimley 2002; Spanier 1979). For example, pulse coding of species-specific sounds is a well-known feature of the mating behavior of Drosophila and other arthropods (Ewing 1989). Male field crickets with longer calling-bout durations are preferred by females (Hedrick 1986). Acoustic communication is a demonstrated feature in amphibian sexual selection (Ryan 1985; Zelick et al. 1999). It seems reasonable to consider that if insects and amphibians are using pulsed sounds in mate selection then the same is possible, even likely, for some fishes.

The following example illustrates typical courtship associated sounds for three cichlid species and emphasizes the comparative details of interspecific variation in single pulse patterns. It is possible that specific waveform structure of single pulses is a product of the sound producing mechanism. Waveforms of courtship sounds for Maylandia zebra (formerly Pseudotropheus zebra , Stauffer et al. 1997), Mchenga (Copadichromis) conophoros , and Tramitochromis intermedius are shown in Fig. 2.

Individual fish can vary the type of sound produced in different behavioral contexts. Courtship and aggressive sounds are hard to distinguish acoustically by listening when played to humans through a loudspeaker. Sounds differ between agonistic and courtship contexts for Maylandia species (Simoes et al. 2008b). However, in one other cichlid, Pundamilia nyererei , sounds produced in different contexts were the same (Verzijden et al. 2010). Acoustic contests in a laboratory setting between males did not show a correlation between male size, increased fighting intensity, and call number (Amorim and Almada 2005). Studies of O. niloticus demonstrated that in male–male contests winners were quick to court, courtship duration was extended and their calls were greater in number with longer pulse durations and lower peak frequencies than previous contest losers (Amorim and Almada 2005), traits that correlate with increased fitness in other vertebrate species. Dyadic contests elicited agonistic sound production in P. nyererei (Verzijden et al. 2010). These sounds are associated with aggressive behavior between males.

Mchenga (Copadichromis) conophoros aggressive sounds (Fig. 4e, f and g) are composed of irregular and longer duration pulses, which are clearly distinguishable from the regular and distinct pulses that are seen in typical courtship sound patterns.

5 Playback Experiments

Early evidence demonstrated the ability for individual male recognition by female H. bimaculatus (Nobel and Curtis 1939). For future research, playback of sounds will be crucial to discern if sounds elicit a behavioral response, which would support the notion that the sounds have specific temporal encoding information. There have been, so far, a limited number of experimental studies on cichlids, which are summarized in Table 4. Playback experiments have demonstrated that a cichlid will respond to conspecific sounds (Rowland 1978). Rowland (1978) conducted behavioral studies to observe the reaction of cichlids to physical models and to the playback of those sound recordings during repeat model experiments. He presented dummy fish with and without playback of the fish’s courtship-display-like and aggressive-display-like sounds as well as presentations with white noise and silence. There was no difference to fish response between white noise and silent dummies, but fish showed a significant response to dummies playing back conspecific sounds over silent dummies. Cichlid sound playback in another study resulted in listeners becoming less aggressive (Schwarz 1974b). Schwartz (1974a, b) found that male Archocentrus centrarchus responded to conspecific sounds and that in this species acoustic cues inhibited aggression. This species does not apparently produce a courtship-associated sound (Schwarz 1980). A preliminary study of neighboring O. mossambicus nest territory holding males were found to produce agonistic “thump” sounds which are not the typically multi-pulsed call (McPherson 2012) and observations of playback suggest a possible chorusing interaction by the lead caller of played-back sound. Bertucci et al. (2013) found that male cichlids during agonistic encounters responded to sounds in general, but not to modification of variation in temporal patterns. Auditory plasticity influences female cichlid hearing and there is overlap between female hearing in reproductive condition and male courtship songs indicating the importance of sound in reproductive communication (Maruska et al. 2012). Two studies found that females preferred males with natural calls compared to males with (a) noise with male (Maruska et al. 2012) and (b) silent with male in playback experiments (Verzijden et al. 2010). Estramil et al. (2014) experimentally demonstrated the multimodal nature of cichlid sounds with females only responding to sounds if live males are present.

6 Species-Specific Sounds

Simple comparison of the pulse rate in the courtship calls of two sympatric species (but different genera) that live close together in Lake Malawi, Mchenga (Copadichromis) conophoros , and Tramitichromis intermedius , revealed statistically significant differences in pulse rate and pulse duration but not pulse number, call duration or inter-pulse duration (Lobel 1998). Lake Victoria cichlids (Verzijden et al. 2010) produce species-specific sounds. Maylandia spp. male courtship sound s differ in temporal acoustic traits (Amorim et al. 2008; Danley et al. 2012) and peak frequency (Amorim et al. 2004). Male courtship sound s of Lake Malawi cichlid s differed among four congeneric and heterogeneric species with a more significant difference between congeners (Van Staaden and Smith 2011) and similar sounds for one pair of heterogeneric species. Male courtship calls of Lake Victoria cichlids, congeners, and heterogeneric species, are species specific (Verzijden et al. 2010). The importance of male calls to female listeners was shown experimentally to the female cichlid ear, with peak sensitivity occurring when females are in reproductive condition prior to the mouthbrooding phase (Maruska et al. 2012) in response to changing hormones demonstrating a peak in call sensitivity during reproductive conditioning. Yet, to date, no playback studies have examined the role of specific temporal or frequency trait differences among different species calls to the female listener. Studies examining contrasting differences in sound characteristics between the males of cichlid species are summarized in Table 5.

6.1 Sonic Diversity and Color Patterns

Nelissen (1975, 1977, 1978) provided a detailed analysis of the sounds of some Lake Tanganyika cichlids and found that several species displayed different acoustic repertoires. The number of sonic displays produced by a species was negatively correlated with the number of color pattern s that species displayed with its behavior (Nelissen 1978). Species could be substituting between visual acoustic diversity and sonic diversity. The most sonically diverse species ( Tropheus spp.) displayed the fewest color changes. The sonically least active species, Simochromis spp. maybe using color pattern displays to communicate. Another species, Astatotilapia burtoni, was intermediate. Increased sound activity with behavior was found for Tropheus spp., which is mainly nocturnal, compared to the diurnally active Simochromis (Nelissen 1978). A similar pattern is well known in coral reef fish communities where many nocturnal species (e.g., holocentrids, pempherids, and sciaenids) are also among the most sonically active. More recently, Amorim et al. (2008) found that among five Lake Malawi cichlid fish species of the Maylandia (Pseudotropheus) zebra complex, that the male courtship acoustic signals differed significantly in the number of pulses and in pulse period. They reported that the largest differences in acoustic variables were found between the two sympatric Mphanga Rocks species P. emmiltos and P. fainzilberi , that, in contrast to the other three species, look (to us at least) very similar to each other in color and pattern.

6.2 Silent Behavior and/or Non-acoustic Cichlids

Although many cichlids appear to make sounds with specific behavior , there have been several reported as being silent during certain behavioral interactions. This is an intriguing phenomenon and further study will be interesting. Ladich and Popper (2001) refer to fish species that are not known to produce volitional sound as non-vocal. “Silent ” is applied to fishes that have been examined for acoustic ability and found to lack it (Kaatz et al. 2010). Although many cichlids appear to make sounds with specific behavior, there have been 11 species reported as being silent during certain behavioral interactions (Table 6). Neolamprologus pulcher did not produce sounds during acoustic ally monitored social interactions (Pisanski et al. 2015, but see Spinks et al. 2017) although it did produce sounds when males were exposed to their image in a mirror (Spinks et al. 2017). The only evidence for possibly silent cichlid species is an anecdotal report found in Spinks et al. (2017) for Astatoreochromis allaudi , based on aquarium studies. Cyphotilapia frontosa lacked an acoustic response to electrical stimulation that did stimulate sounds in one other tested cichlid known to be a sound producer (Longrie et al. 2009). However, it should be emphasized that no behavioral studies of living social groups have yet been conducted for these species. The occurrence of these silent cichlid species raises the question as to whether this represents a secondary loss of this ability, or whether sounds in cichlids are independently derived. This exciting research is in its beginning phase of discovering which fish species may be making behaviorally relevant sounds . This is an intriguing phenomenon for further study.

Table 6 Cichlids displaying behaviors unaccompanied by sounds
Full size table

Contexts that were well monitored and found to lack sound for all individuals tested in a species with other acoustic contexts included are listed in Table 6. Some examples are: Hemichromis bimaculatus courtship and spawning, brood care (Myrberg et al. 1965); A. nigrofasciata reproduction and male fighting (Myrberg et al. 1965); courtship, spawning, and guarding of fry in O. niloticus (Longrie et al. 2009); juveniles schooling and female during agonism, spawning, brooding, and fry release for Tramitichromis cf. intermedius (Ripley and Lobel 2004); Pterophyllum (scalare sp?) for schooling adults and parents on nest or with eggs, larvae, or fry without intruders (Myrberg et al. 1965) courtship associated A. centrarchus during spawning for H. multispinosa (Brown and Marshall 1978). Neolamprologus pulcher did not produce sounds during acoustic ally monitored social interactions including male–male agonism and courtship (Pisanski et al. 2015, but see Spinks et al. 2017). Female Tilapia mariae did not produce sounds during male–female and male–male agonistic interactions (Kottege et al. 2015). Courted males of Oreochromis mossambicus were silent when displayed by vocalizing males (Amorim et al. 2003). Individuals of the subpopulation Maylandia zebra "katale" only produced incidental patterned sounds on the territory while clearing gravel, not during visual courtship displays (Smith and Van Staaden 2009; Van Staaden and Smith 2011). There are additional unpublished data reported by Spinks et al. (2017) supporting silence in another cichlid species, Astatoreochromis allaudi , based on aquarium studies. It is possible that there is individual variation in vocal activity (Van Staaden and Smith 2011); therefore, a statistically adequate number of individuals should be investigated before concluding the sonic status of a species or a specific behavioral context.Relatively low-intensity acoustic activity has, so far, been observed in three cichlid species (Table 4). Calling with visual courtship displays was rare or absent in some individuals of two tested species ( Maylandia zebra “katale” and Melanochromis auratus ) compared to other sympatric species (Van Staaden and Smith 2011). However, reproductive behaviors were also uncommon in the small aquarium population tested (Smith and Van Staaden 2009). Intraspecific agonistic interactions of non-territorial males and females in O. mossambicus were rarely associated with inter-pulsed sound production (Amorim et al. 2003) while territoriality was the condition under which males began calling in greater abundance. In another study, territorial O. mossambicus males were found to produce distinct agonistic “thump” and pulsed courtship sound s on nest territories in the field (McPherson 2012). More studies are required to explore this aspect of cichlid bioacoustics .

7 Sound Detection and Hearing

The hearing abilities have not been measured in many cichlid species, compared to other taxa, and most of the work on the auditory system has examined ear or otolith morphology. The cichlid Sarotherodon macrocephala responds to underwater sound stimulation from 50 to 900 Hz with maximum sensitivity at 100 Hz (Fay and Popper 1975). This range of hearing sensitivity is typical for percoid type fishes without special morphological adaptations for hearing (e.g., ostariophysians, clupeids, Fay 1988; Popper and Fay 1973, 1993). The ears in both Old World Sarotherodon macrocephala and New World Aequidens pulcher are very similar to one another (Popper 1977). The otoliths of Cichlidae are diagnostic for the family (Gaemers 1984; Gaemers and Crapon de Crapona 1986). Sexual dimorphism in otoliths occurs in those haplochromine species that have a maximum male size greater than maximum female size (Gaemers and Crapon de Crapona 1986). The auditory sensitivity of the Malawi cichlid , Tramitichromis intermedius is correlated with the frequency range of sounds it produces during aggression and courtship (Ripley et al. 2002). The acoustic cichlid Astatotilapia burtoni ’s hearing is sensitive to its own call acoustic traits (Maruska et al. 2012) as is the hearing in Labidochromis caeruleus (Higgs et al. 2011). The peak hearing sensitivity of Tramitochromis intermedius corresponds with the dominant frequency of its courtship calls (Ripley et al. 2002).

Auditory psychophysiological data support the notion that fishes, in general, have peak auditory sensitivity below 1000 Hz (Fay 1988; Popper and Fay 1993). The fish’s central (brain) and peripheral (ear) auditory systems are fully capable of processing: l) sound amplitude fluctuation with respect to both sensitivity and pattern discrimination and 2) sound source localization in azimuth and elevation (Fay 1988; Popper and Fay 1993). The response by females to pulsed courtship calls for mate selection has been demonstrated experimentally for a pomacentrid and is implicated in other fishes (Gerald 1971; Fine et al. 1977; Myrberg et al. 1978, 1986; Myrberg and Spires 1980; Schwarz 1985; Spanier 1979). That a cichlid can detect and learn response to sound has been demonstrated in Oreochromis aureus, which was behaviorally conditioned to associate a buzzer sound to being fed. Acoustic interaction between sexes is also suggested by results, which indicated that the courtship sound produced by male Oreochromis mossambicus (syn. Sarotherodon , Tilapia mossambica ) facilitated oviposition in females (Marshall 1972). Auditory plasticity in male and female Astatotilapia burtoni has been examined and could influence how social sounds are detected through the reproductive cycle (Maruska and Fernald 2010).

Lateral line systems of fishes are sensitive to water displacement as well as near field and likely low-frequency sound. Both the lateral line and the ear are innervated by the VIIIth cranial nerve (see Webb et al. 2021). Various cichlid genera possess at least eight trunk canal patterns (Webb 1990). Another potential sensory adaptation , which is very poorly understood, may involve enlarged sensory pores in the mandible. These pores are apparently linked to the lateral line and presumably, may have a role associated with enhanced sound detection (Konings 1995). This feature has been identified in the genera Aulonocara , Tramitichromis , and related species (Konings 1995; Turner 1996). How these lateral line and sensory pore patterns may be related to a cichlid's response to sounds is not known. The suggestion is that these mandibular pores function to detect sounds or vibrations produced by invertebrate prey moving beneath the sand (Janssen 1990; Konings 1995). The role of lateral line cues in fish communication is an emerging topic of study (Butler and Maruska 2015, 2016; Weeg and Bass 2000; Weeg et al. 2005).

7.1 Background Noise Interference

Background noises due to pumps, filters, lights, and echoes in aquarium settings make it difficult to conduct critical experiments in small aquaria on how cichlid fishes produce and respond to sounds (Nelissen 1991). Acoustic recordings and playbacks made in an aquarium may be potentially misleading when interpreting specific acoustic characteristics of fish sounds because of reflections and reverberations against the walls (Akamatsu et al. 2002; Hawkins 1973). The type of aquarium wall materials may also have an impact on sound recordings (Parmentier et al. 2014).

Background noises (e.g., aquarium pumps and filters) sometimes appear as dark horizontal bands in sonograms (Nelissen 1978) and can mask part or all of a fish’s sound, making analysis difficult (Rowland 1978). The effects of constant aquarium noise on a cichlid’s development and behavior may also be a concern. Damage to fishes’ hearing can result from excessive noise (Banner and Hyatt 1973; Cox et al. 1987; Ha 1985; Popper and Clarke 1976). Experiments have shown that environmental noise reduces behavioral activity levels in a cichlid and could therefore also impact its acoustic display activity (Slabbekoorn et al. 2012).

The influence of aquarium noises on fish hearing and communication is important to determine from three perspectives. First, how such noise affects acoustic recordings is an analytical issue. Second, it is a potential problem when interpreting behavior in an aquarium because we do not know how this noise may directly interfere by masking fish communication . Third, the direct effect on hearing and the development of hearing as fish are raised in captivity is an obvious concern. It is possible that fishes housed in high noise aquaculture and aquarium systems may have their hearing adversely impacted.

8 Future Directions

Two alternative (but mutually compatible) hypotheses have been proposed to explain rapid and extensive speciation of cichlids, one based on morphological plasticity and the other on sexual selection by female mate choice (Greenwood 1991). Morphological plasticity is inherent in cichlids and enables rapid anatomical changes, especially in the feeding mechanism of which the pharyngeal jaws are a key component (Liem 1973, 1991; Greenwood 1973, 1991). Alternatively , sexual selection is based upon female choice between variants in male features, most notably coloration (Dominey 1984; Seehausen et al. 1999) and bower (nest) height and shape (Stauffer and Kellogg 1996). The hypothesis debate is based, in part, upon the premise that anatomical changes in the pharyngeal jaws are not affected by sexual selection (Greenwood 1991). If, however, the pharyngeal apparatus is important for sound production as well as feeding, then some features of the morphology may be linked to sexual selection (as related to sound production) as well as trophic specialization. For female choice to function on this basis, different pharyngeal morphologies would be expected to produce sounds with different qualities. Preliminary observations of individual pulse waveforms among cichlids and pomacentrids reveal such differences. The evidence available to date indicates that sound production is an integral part of cichlid behavior, including courtship. However, before its role in the speciation process can be evaluated, the role of sound communication in mate selection needs to be determined. Darwin (1874; pp. 366–367) argued, “It is almost certain that in [fishes] sound producing instruments have, at least in some cases, been developed through sexual selection as a means for bringing the sexes together.” This hypothesis remains to be rigorously tested and although data is accumulating that suggests sounds do play an essential role in mate selection among cichlids, however, acoustic s may not be the prime driver alone. Coincident signals of colors, fins and swimming, and odors may also be required to reinforce a female’s choice of a mate. Most simply, sound with behavior could serve to be an initial attention attracting display without any deeper content to the acoustic signal . These hypotheses are ready for experimental testing.

Acoustic signals are involved in much of the social and reproductive behavior of many animals, and the behavioral influence of sounds has been extensively studied in these terrestrial groups. In contrast, research, to date, on the acoustic communication of fishes has primarily focused on the description of sounds, sonic morphology, and definition of the contexts in which sounds are produced. Progress in the field of fish bioacoustics has lagged behind that of terrestrial systems mainly due to limitations of technology and the logistical difficulty of research in the underwater environment. Recent advancements have provided better tools to successfully conduct such studies and the number of fishes discovered making sounds has grown enormously in the last few decades (Rosenthal and Lobel 2006; Webb et al. 2008). Although it is now known that many fishes make sounds, only a limited number of studies have experimentally examined the behavioral significance of these sounds by using audio playback (Ladich 1997; Luczkovich and Keusenkothen 2007; Lugli 1997; Lugli et al. 2004; McKibben and Bass 1998; Myrberg et al. 1986; Myrberg and Riggio 1985; Rollo and Higgs 2008; Winn 1972; Yan and Popper 1991).

There is now ample evidence that cichlids, like many fishes, make distinct sounds associated with specific behaviors. The field of fish bioacoustics is making progress. However, to advance closer to the level of understanding that has been achieved in the studies of terrestrial animals, the next phase of research for fishes needs to focus on experiments using playback trials to determine on how sounds combined with visual and chemical cues are behaviorally relevant.