The Mechanosensory Lateral Line

The mechanosensory lateral line system can be most easily distinguished from its electrosensory counterpart by the morphology of its end organ, the neuromast, which consists of hair cells and associated support cells (see Jørgensen, Chapter 6, for further details). Historically, at least four classes of end organs, distinguished primarily by surrounding support structures, have been classified as mechanosensory lateral line organs: (1) superficial neuromasts found on the skin surface in cartilaginous and bony fishes and some aquatic and semiterrestrial amphibians; (2) canal neuromasts enclosed in tubes formed by cartilage in elasmobranch fishes and by bone or scale in most bony fishes; (3) spiracular organs housed in diverticula of the hyoid pouch and found in most nonteleost bony fishes; and (4) vesicles of Savi, enclosed pouches containing a cluster of neuromasts and found in some elasmobranchs. The majority of the work contained in this book focuses on what is known about the first two classes of neuromasts (see Dijkgraaf 1963 and Coombs et al. 1988 for further reviews on morphological variation within these classes), but the chapter by Barry and Bennett is devoted exclusively to spiracular organs and the vesicles of Savi.

When the organizers of this conference asked me to give a short personal historical review of the development of research work on the lateral line system, I felt of course honored and gladly accepted, be it not without a certain timidity, because among a group of mostly young or middle-aged conference attendants still engaged in active research, I felt somewhat like a living fossil. Therefore I have confined my review to the origin and first results of physiological research.

Phylogenetic studies can be undertaken with a number of aims: (1) to describe the history of traits; (2) to discern form-function relationships that may not be evident from examining the traits in a single species; (3) to generate hypotheses regarding the evolutionary mechanisms responsible for the history of the traits; and (4) to detect gaps in our knowledge of the traits (Gans 1985; Northcutt 1985a, 1986a). The following analysis involves each of these aims but is confined to the distribution and innervation of mechanoreceptive neuromasts. It does not attempt to deal with related receptors and organs or the phylogeny of the octavolateralis system as a whole, although I have discussed electroreceptors and the inner ear in previous publications (Northcutt 1980, 1986a,b).

Evolution has been described as the result of two independent and sequential processes: the origin of variation and natural selection. The phenotypic variation upon which natural selection acts is, in turn, a result of processes that generate as well as those that limit variation (Alberch 1980, 1982b). The processes that generate variation (mutation, recombination) have been the mainstay of evolutionary thought for the past century (Mayr 1982), but many processes limit and channel variation to produce “morphological gaps” (Alberch 1980,1982b). These gaps or “potential morphospaces” remain unoccupied, not because of the absence of appropriate selection pressures, but because historical (phylogenetic) and ontogenetic constraints on morphological change limit morphological variation (Liem and Wake 1985). Consideration of the contributions of these constraints in morphological evolution (e.g., Gould and Lewontin 1979; Alberch 1980) challenges the adaptationist program that has prevailed in the field of evolutionary biology for the past century.

Amphibians possess, at least as larvae, a system of placodally derived secondary sensory cells aggregated into organs on the head and the trunk. Similar to many fishes (Northcutt 1986), two amphibian orders, urodeles and gymnophionans, possess two types of organs: mechanoreceptive neuromasts (the lateral line) and electroreceptive ampullary organs (the electrosensory system). The third order, the anurans, possess only neuromasts (Fritzsch and Münz 1986; Fig. 5. 1a). Present in all aquatic larvae and some in utero embryos, these organs are lost in many, but not all, amphibians during metamorphosis (Wahnschaffe et al. 1987). Moreover, in some amphibians, without free-living larvae, none of these organs seem to be formed, and they are absent in all amniotic vertebrates (Escher 1925).

According to the acousticolateralis hypothesis (Ayers 1892; Kingsbury 1895; van Bergeijk 1967), the vertebrate inner ear has its evolutionary origin in the lateral line system. The arguments for the hypothesis were based on (1) the origin of the lateralis nerves from the immediate vicinity of the eighth nerve to the inner ear, (2) the general resemblance of the sensory cells of the two systems, (3) the tendency of the canal neuromasts and inner ear to sink below the surface during ontogenetic development, and (4) the ontogenetic development of both systems from a common epibranchial thickening, or placode.

Early studies of the developing lateral lines of amphibians demonstrated that a variety of well-orchestrated cellular behaviors give rise to a rather simple morphological pattern in the embryo. In 1904, Harrison first demonstrated that neuromasts of the posterior lateral lines of frogs (Rana palustris, R. Sylvatica) are derived from cells that migrate from the head region, and that the posterior lateral line nerve connects the posterior lateral line ganglion with the migrating primordium. Later, in studies of the salamander (Ambystoma punctatum) Stone (1922) demonstrated that the posterior lateral line ganglion cells and the primordial cells that form neuromasts both arise from a common postotic placode. Cells in the rostra1 part of the placode become the sensory neurons, and cells in the caudal part form the migratory primordium. As the primordium migrates caudally along the midbody line, clusters of cells are deposited at intervals from the trailing end of the primordium. These cell clusters subsequently differentiate to become the first, or “primary”, neuromasts of the midbody line. Thus, the development of the posterior lateral line involves a variety of fundamentally important but poorly understood cellular activities, such as the formation and differentiation of a placode, initiation and guidance of cell migration, and axon outgrowth and guidance. The posterior lateral line system may prove to be a good model system in which to study these phenomena because of its relative simplicity (being a simple linear array of receptors in the embryo) and accessibility (the receptors are located superficially within the epidermis) and because of the considerable background of information that is already available regarding the development of this system.

In aquatic amphibians, removal of a patch of skin that contains lateral line neuromast organs evokes the production of new neuromasts that differentiate in the approximate location of the organs that were lost. The whole process can be completed in less than 2 weeks, with the number of replacement neuromasts roughly matching the number lost. This chapter attempts to explain some of the cellular mechanisms that lead to the regenerative replacement of lost neuromasts. It also discusses the potential generality of those mechanisms in regard to regeneration of cochlear hair cells, a recently discovered phenomenon that appears to result in recovery of function following profound sensorineural hearing loss in birds.

In detecting moving objects, the fish’s lateral line and inner ear support different, but closely related, hydrodynamic functions, suggesting early developments in the vertebrate sense of hearing. In an effort to elucidate the functional evolution of the two sensory systems, this chapter examines the hydrodynamic and acoustic fields in nature, the physics and physiology of the detection process, the evaluation of the sensory data, and the ensuing behavioral responses.

It is known that certain fish that have been blinded are still able to avoid collision with stationary objects such as the aquarium wall (Dijkgraaf 1962). Congenitally blind fish, like the blind cave fish Anopthichthys jordani, can swim around in their environment as surely as fish with vision.

There are good reasons for measuring stimuli to lateral-line systems as pressure gradients (Denton and Gray 1988), the most likely useful natural stimuli to lateral line sense organs being local pressure gradients produced by a fish’s own swimming movements and local pressure gradients produced by external sources. In what follows we shall assume that (1) the effective stimulus for a canal neuromast will be one that produces a movement of the cupula relative to the wall of the canal, (2) the force moving the cupula must be transmitted to the cupula through the liquid that the canal contains, and (3) displacements of such liquid and the cupula of the neuromast will arise from net pressure gradients along the short stretch of canal carrying the neuromast. This net gradient can be measured in terms of the local accelerations of the medium adjacent to the canal relative to the surface of the fish (Denton and Gray 1983). It follows that if the medium adjacent to the lateral line moves in exactly the same way as the surface of the fish that carries a particular “section” of lateral line, the neuromast in this section will not be stimulated. When considering an external source, the stimuli to lateral lines will often be proportional to the accelerations of the sonic surface (Gray 1984; Kalmijn 1988, Chapter 9)

In the transmission and processing of mechanically coded information by organs of the acousticolateralis system of vertebrates, the transfer of vibrations via fluid to accessory structures of hair cells plays an important role (Pumphrey 1950). In the lateral line system, the hydrodynamic coupling between the cupula and the surrounding water together with the mechanical properties of the cupula filters the mechanical information that reaches the sensory hair cells.

Sensory hair cells are specialized epithelial cells, which function as mechanoelectrical transducers in the acousticolateralis organs of vertebrates. Since early in the 1950s (Jielof et al. 1952), the lateral line has been extensively used as a model for studying the fundamental processes of hair cell physiology (for reviews see Dijkgraaf 1963; Flock 1971; Russell 1976; Sand 1984). In fact, most of the functional properties of hair cells, such as their directional sensitivity (Flock and Weräll 1962; Harris et al. 1970) and their nonlinear input output function (Flock1965), were discovered as a result of lateral line studies.

In the lateral line, as in any other sensory system, the function of the system depends on the peripheral receptive structures. These receptive structures, which are the neuromasts in the mechanosensory lateral line system, determine the character of the information available to higher brain centers. Parameters related to the functional organization of the lateral line periphery include the spatial arrangement and the innervation of neuromasts and the presence of different morphological types of neuromasts which may have different physiological properties. Additionally, the centrifugal control of the periphery mediated by the efferent innervation has to be considered.

Although there is considerable information on the peripheral anatomy of lateral line systems (see Dijkgraaf 1962; Russell 1976; Coombs et al. 1988 for reviews), relatively little is known about how sensory information is encoded by the nervous system, either at the periphery or more centrally. To answer questions of this nature, it is necessary to have information not only on the peripheral anatomy of the system and its connections to and within the central nervous system but also on the response properties of neurons at various levels of the nervous system. This kind of information must then be evaluated in terms of the response of the entire system-i.e., the behavioral output or sensory capabilities of the animal.

The central neuronal pathways and connections required to process mechanosensory lateral line information probably arose, concomitantly with the peripheral receptors, in the first vertebrates. The cartilaginous fishes (Chondrichthyes) share the mechanosensory component of the lateral line system with all other anamniotes, although there are vast differences in the central nervous systems of those vertebrates that possess lateral line sensory capabilities. Even among the elasmobranchs, there are different levels of brain organization and profound variations within each level (Northcutt 1978). It is important to identify mechanosensory lateral line centers and their fiber relations with other octavolateralis centers among those elasmobranchs that possess different patterns of brain organization, in order to establish homologies and to better understand the evolution of mechanoreception in chondrichthyans. However, our knowledge of the central mechanosensory lateral line pathways is limited to only a few elasmobranchs, e.g., the batoids (thornback guitarfish Platyrhinoidis triseriata, clearnose skate Raja eglanteria, and the little skate Raja erinacea), the squalomorph sharks (spiny dogfish Squalus acanthias), and the galeomorph sharks (carpet shark Cephaloscyllium isabella). Central lateral line pathways have been studied mostly in skates that are considered advanced elasmobranchs in terms of brain organization (Northcutt 1978), but some information is available in the spiny dogfish and carpet shark which, in contrast to skates, are more primitive in terms of brain organization. The purpose of this chapter is to delineate some of the central mechanosensory lateral line centers and their connections at each level of the neuraxis and to establish homologies in these few species of cartilaginous fishes.

The lateral line mechanosensory system originated early in vertebrate history. It was apparently present in at least some extinct agnathous forms and inherited from these vertebrates by the early gnathostome fishes. So far as is known, the system is present in petromyzontid agnathans, in all extant cartilaginous and bony fishes, and in most larval and many postmetamorphic amphibians (Boord and Montgomery Chapter 16; Fritzsch Chapter 5; Northcutt Chapter 3). By all indications, the lateral line mechanosensory system evolved in early agnathans and was retained in subsequent radiations; it is thus by definition homologous among those vertebrates that possess it. This homology is reflected both in the anatomical similarity of the neuromasts across groups and in the similarity in the overall pattern of central connections of the system, at least among gnathostomes. This chapter will summarize our current understanding of these central anatomical pathways in bony fish, pointing out variations within this group and similarities to cartilaginous fishes and amphibians.

In recent years neurobiological research on the lateral line system has produced an impressive body of knowledge not the least of which is evidenced by the present volume. In this chapter, present data on the central organization of the amphibian lateral line system will be reviewed. Commencing with the central termination of lateral line afferents, it will successively consider higher nuclear levels, taxonomic differences being discussed at each level. Yet, already at the medullary level, comparative neuroanatomical material exists mainly for only urodeles and anurans. As for the third amphibian order, gymnophionans, several studies have dealt with the peripheral lateral line system (Taylor 1970; Hethrington and Wake 1979; Fritzsch et al. 1985; Wahnschaffe et al. 1985). Knowledge about central connections, however, is still limited to a very few data. Furthermore, there are hardly any data on the physiology of the lateral line at the medullary level. Neuroanatomical data on higher projections, except for the projection to the mesencephalon in urodeles and gymnophionans, are available only for anurans. However, far from being complete, they are limited to the mesencephalon and its connections.

Fishes and aquatic amphibians use the mechanoreceptive lateral line to detect weak water currents (Dijkgraaf 1963; Bleckmann 1986; Kalmijn 1988) and water surface waves (Schwartz 1971; Bleckmann 1988). Electrophysiological studies have shown that the pattern of impulses carried by primary lateral line afferents encodes information about the nature of the peripheral stimulus with respect to duration, amplitude, frequency, and phase (see Münz Chapter 14). If the activity of several neuromasts, which may differ with respect to the alignment of their most sensitive axis, is integrated over time and space, the additional information of stimulus direction and, perhaps, stimulus distance may be obtained. Thus the peripheral lateral line provides the brain with all cues necessary to evaluate a complex wave stimulus with respect to stimulus origin, stimulus duration, and stimulus type.

Some predatory fishes and aquatic amphibians are able to use surface waves to find prey, such as struggling insects trapped on the water surface (for review, see Bleckmann 1986, 1987; Bleckmann et al. Chapter 25). The sensory system mainly responsible for the detection of surface waves is the mechanosensory lateral line system. The sensory organs of the lateral line system, the neuromasts, are stimulated by the water motion induced by the propagating surface waves. Sensory processing in the lateral line system provides the animals with information about the direction and, at least in some surface feeding fish, the distance to the source of a surface wave (Bleckmann and Schwartz 1982).

It is well established that the mesencephalic torus semicircularis of fish obtains input from the octavus, lateral line, and visual systems. In electroreceptive teleosts and elasmobranchs, electroreceptive information reaches the torus as well. The modalities mentioned are processed separately not only in unimodally sensitive units but also in multimodal neurons.

Our understanding of lateral line operation cannot be complete until the significance of the efferent innervation is understood. This innervation is a fundamental component of hair cell function: it is found not only in canal and superficial neuromasts of the lateral line but also in the inner ear. This efferent system, in common with the efferent innervation to the retina and muscle spindle, provides the opportunity for a dynamic control of sense organ properties, but how this is used and organized, in the case of octavolateralis efferent neurons, is unknown.

The Mauthner cells(M cells) are a pair of neurons found in the hindbrain of most fishes and amphibians (Stefanelli 1951; Zottoli 1978; Will 1986).These exceptionally large cells (up to 100 μm in cross-sectional diameter) are identifiable by their distinctive morphology, their position within the medulla oblongata, and their electrophysiological response characteristics. All of these features combine to make M cells an easily accessible “model system” for studying a number of synaptic phenomena, including electrotonic transmission (Furshpan 1964;Faber et al.1980), synaptic integration (Bartelmez 1915; Bodian 1937;Chang et al.1987;Faber and Korn 1978; Korn and Faber 1975a; Lin et al.1983; Zottoli and Faber 1979;Zottoli et al. 1987a), and transmitter release mechanisms (Faber and Korn 1982;Korn et al.1982, 1984, 1986; Korn and Faber 1987).

Most species of marine teleosts have very small transparent larvae at hatching, with total lengths from about 1.5 to 8 mm. The eyes may or may not be pigmented at this time, but they always become functional when the larvae commence feeding a few hours to a few days after hatching.Feeding seems to be mainly a visual process, and the larvae of only a few species, such as Dover sole, Solea solea, are known to feed in darkness(Blaxter 1969).All species examined have free neuromast organs distributed over the head and body as superficial hillocks. These hillocks, which are very large relative to the body of the larva, have gelatinous cupulae projecting into the surrounding water.Larval neuromasts are well described by Iwai (1980) in several species including the goldfish(Carassius auratus), sea bass (Lateolabrar japonicus), black porgy (Acanthopagrus schlegeli), and right-eye flounder (Kareius bicoloratus).Disler(1971) followed the changes in number and distribution of the free neuromasts during the development of the sturgeon (Acipenser stellatus), chum salmon(Oncorhynchus keta), and several freshwater percids and cyprinids.Later work on gadoids (Fridgeirsson 1978), northern anchovy (Engraulis mordar)(O’Connell 1981), Atlantic herring (Clupea harengus)(Blaxter et al.1983a), Atlantic halibut(Hip-poglossus hippoglossus) (Blaxter et al.1983b), and plaice(Pleuronectesplatessa) and turbot (Scophthalmus marimus)(Neave 1986)confirmed the earlier findings, showing the increase in numbers with age and also the modifications that occur during the metamorphosis of flatfish.

Fishes live in almost every type of aquatic habitat. Not surprisingly, many fishes show striking morphological, physiological, and behavioral adaptations. Some species, belonging to the families Cyprinodontidae, Hemirhamphidae, Gasteropelecidae, and Pantodonitidae, are specialized in foraging at the water surface. Schwartz (1965, 1971) has shown that surface-feeding fish detect and localize part of their prey, terrestrial insects fallen into the water, by means of capillary surface waves elicited by the prey’s struggling. The receptive structures involved in prey detection are lateral line organs located on the fish’s head and back (Schwartz 1970). Behavioral studies indicate that surface-feeding fish distinguish between different wave types. In addition, these fish determine the direction and the distance to a wave source under open loop conditions-i.e., even if only a single, short-lasting wave train is presented. In this chapter we review the literature on surface-feeding fish. In addition, we include recent unpublished anatomical, electrophysiological, and behavioral results. For more general reviews on water surface wave reception, the reader is referred to Bleckmann (1985a, 1988) and Wilcox (1988).

Perception of water movement on the body by means of lateral line organs is lacking in humans. Therefore, to understand stimulus analysis with the lateral line system, behavioral analyses that reveal the system’s sensory capacities are particularly important. The results of psychophysical testing provide insight into what kind of information about the aquatic environment can be obtained with the system. They are further prerequisites for adequate electrophysiological and theoretical analyses of how the system’s properties are accomplished. In this chapter, the results of behavioral analyses of the mechanoreceptive lateral line system (subsequently referred to as lateral line system) in amphibians are reviewed, and some parallels to stimulus analysis in the electroreceptive lateral line and auditory systems are pointed out.

Since the experiments of Kramer (1933) and Dijkgraaf (1947), it has been known that the clawed toad (Xenopus sp.) can localize moving objects in water or on the water surface. Kramer demonstrated that the lateral line system was responsible for the toad’s ability to turn toward the source of disturbance. Ablation of all lateral line stitches on one side of the animal abolished its response to a hand-held vibrating sphere and to a weak water current delivered by a pipette on the side. Responses to stimuli delivered on the intact side remained unaffected. Dijkgraaf has investigated the reactions of the toad to surface waves created by dipping a small rod into the water. The toad responded by turning toward the wave center even when it rested on the substrate, 10 cm below the water surface. As it turned out, surface waves are well suited for a detailed study of lateral line function in Xenopus, since they elicit unconditioned orienting responses which do not habituate over the course of more than 140 successive stimulations (Görner 1976).

This chapter is in three main sections. The first reviews the available evidence that some fish and aquatic amphibians are capable of using the lateral line to feed on zooplankton. This ability enables them to feed at night (and hence avoid diurnal predators or utilize a food resource enriched by demersal zooplankton) or to feed through the long periods of darkness at high latitude. Evidence is also presented that lateral line information is likely to be of importance in complementing vision in other species of plankton-feeding fish.

The lateral line, comprising the canal organs and free neuromasts, is generally regarded as a sensory system for the detection of local water currents (Hofer 1908; Dijkgraaf 1934) and surface waves (Schwartz 1965; Bleckmann et al., Chapter 25). However, whether the lateral line responds to low-frequency sound as well has long remained an issue of debate. Dijkgraaf (1963, Chapter 2) strongly argued against an acoustic function of the lateral line, citing a lack of compelling behavioral evidence. Sand (1981, 1984) explained that the operation of the lateral line in free-moving fish is physically restricted to the immediate vicinity of the source. Kalmijn (1988a, Chapter 9) subsequently focused attention on the function of the inner ear in detecting the local flow fields of moving objects at distances beyond the limited range of the lateral line. The lowfrequency nature of the two sensory systems is consistent with the results of earlier physiological studies (Suckling and Suckling 1950; Harris and van Bergeijk 1962; Enger 1966; Kalmijn 1988a).

The mechanoreceptive lateral line system shows considerable intra-and interspecific diversity in morphology (Coombs et al. 1988), but the functional ramifications of this diversity are poorly understood. The two receptor systems discussed in this chapter, the spiracular organs and the vesicles of Savi, are specialized lateral line organs, whose physiology has been studied in elasmobranchs.

Water currents are among the most basic events in the environment for almost all aquatic invertebrates. Since these animals are predominantly active at night or have no or only poorly developed eyes to localize a moving prey, predator, or conspecific, water currents–with or without chemical signals–are probably the most important source of information about events in their environment. Thus, in aquatic invertebrates one should expect a widespread occurrence of receptor systems that are sensitive to some kind of water movement.

The inner ear and lateral line form the octavolateralis system in aquatic vertebrates. This system provides an array of electrosensory and mechanosensory inputs that are integrated with chemosensory and visual information to produce behavioral responses appropriate for the organism (see Blaxter 1988). As several chapters in this volume note (Enger et al. Chapter 29; Kalmijn Chapter 9), there are structural and functional parallels between the inner ear and the mechanosensory lateral line systems. Direct comparisons between these systems are useful with regard to the phylogeny, ontogeny, micromechanics, transduction processes, coding, central connections, and behavioral use of these organs. Just as research on the lateral line has advanced our understanding of inner ear function (see Flock 1974), insights gained from vestibular and auditory research might help to guide questions regarding lateral line function. This chapter addresses some parallels between the mechanosensory lateral line and the ear of fishes and considers how investigators of these two systems might learn from one another.

A volume on the lateral line would be incomplete without a consideration of its closely allied sense, electroreception. Electroreception is usually considered a specialized lateral line sense (Dijkgraaf 1962; Lissmann 1967; Szabo 1974). The marked similarities between mechanosensory lateral line (hereafter referred to as lateral line) and electrosensory systems, including embryology and morphology of receptors, receptor innervation and distribution, and CNS anatomy, are generally appreciated and held by most to reflect a close evolutionary relationship between these so-called lateralis senses. Since the discovery of electroreception 30 years ago in weakly electric teleost fishes (Lissmann 1958; Bennett and Grundfest 1959; Bullock et al. 1961; Fessard and Szabo 1961), the favored hypothesis has been that electroreceptors evolved as specialized derivatives of preexisting lateral line neuromasts (Lissmann 1958, 1967; Mullinger 1964; Bennett 1971). Major support for this hypothesis was the phylogenetic distribution of the two senses as it was then known. In the last volume on the lateral line, published 20 years ago, Lissmann (1967) stated the case simply: “The ordinary neuromast is the rule, the ampullary organ [electroreceptor] the exception.”

Lateral lines are difficult to understand, since we have no comparable system, and the presumed adequate stimuli overlap with those of other systems. Sensory biology in lower vertebrates is skewed in respect to well-developed and poorly developed areas. This volume shows substantial progress in some facets of lateral line research, such as anatomical and peripheral physiological aspects, at the same time that it reflects slower advance in others, such as behavioral and central physiological aspects. The following remarks can only be one man’s personal view of the prospects and opportunities ahead, since it would take multiple standpoints to assure that some important challenge is not left out. I will confine these comments to three areas: evolutionary, general neurobiological, and neuroethological.