Search

Karyotyping is a basic method to investigate chromosomal evolution and genomic rearrangements. Sixteen genera within the basal teleost order Osteoglossiformes are currently described cytogenetically. Our study adds information to this chromosomal dataset by determining the karyotype of Campylomormyrus compressirostris, a genus of African weakly electric fish that has not been previously examined. Our results indicate a diploid chromosome number of 2n=48 (4sm+26m+18a) with a fundamental number of FN=72. This chromosome number is identical to the number documented for the sister taxon of the genus Campylomormyrus, i.e., Gnathonemus petersii (2n=48). These results support the close relationship of Campylomormyrus and Gnathonemus. However, the karyotype formula of C. compressirostris is different from Gnathonemus petersii, thereby confirming the high variability of karyotype formulae within the Mormyridae. We infer that the differences in chromosome number and formula of Campylomormyrus relative to other mormyrids may be caused by Robertsonian fusion and pericentric inversion. In addition to the karyotype description and classification of Campylomormyrus, a ChromEvol analysis was used to determine the ancestral haploid chromosome number of osteoglossiform taxa. Our results indicate a relatively conservative haploid chromosome number of n=24 for the most recent common ancestor of Osteoglossiformes and for most of the internal nodes of osteoglossiform phylogeny. Hence, we presume that the high chromosome variability evolved recently on multiple independent occasions. Furthermore, we suggest that the most likely ancestral chromosome number of Mormyridae is either n=24 or n=25. To the best of our knowledge this is the first attempt to determine and classify the karyotype of the weakly electric fish genus Campylomormyrus and to analyze chromosomal evolution within the Osteoglossiformes based on Maximum Likelihood and Bayesian Inference analyses.

The Electrosensory Lateral Line lobe (ELL) is the first central target where the electrosensory information encoded in the spatiotemporal pattern electroreceptor afferent discharges is processed. These afferents encode the minute amplitude changes of the basal electric field through both a change in latency and discharge rate. In the ELL the time and rate-coded input pattern of the sensory periphery goes through the granular cell layer before reaching the main efferent cells of the network: large fusiform (LF) and large ganglion (LG) cells. The evidence until now shows that granular cells are inhibitory. Given that large fusiform cells are excited by the sensory input, it remains a mystery how the afferent input produce excitation through a layer composed by only inhibitory cells. We addressed this problem by modeling how the known circuitry of the ELL could produce excitation in LF cells with only inhibitory granular cells. Alternatively we show that a network composed of a mix of excitatory and inhibitory granular cell not only performs better, as expected, carrying excitation to LF cells but it does so robustly and at higher sensitivity by enhancing the contrast of the electric image between the periphery and the ELLs output. We then show with refined histological methods that a subpopulation of the granular cells indeed are excitatory, providing the necessary input for this contrast enhancing mechanism. Copyright 2016 Elsevier Ltd. All rights reserved.

African weakly electric mormyrid fish show a high diversity of their electric organ discharge (EOD) both across and within genera. Thanks to a recently developed technique of artificial reproduction in mormyrid fish, we were able to perform hybridizations between different genera and within one genus (Campylomormyrus). The hybrids of intergenus hybridizations exhibited different degrees of reduced survival related to the phylogenetic distance of the parent species: hybrids of the crosses between C. rhynchophorus and its sister genus Gnathonemus survived and developed normally. Hybrids between C. rhynchophorus and a Mormyrus species (a more basal clade compared to Campylomormyrus s) survived up to 42days and developed many malformations, e.g., at the level of the unpaired fins. Hybrids between C. numenius and Hippopotamyrus pictus (a derived clade, only distantly related to Campylomormyrus) only survived for two days during embryological development. Eight different hybrid combinations among five Campylomormyrus species (C. tamandua, C. compressirostris, C. tshokwe, C. rhynchophorus, C. numenius) were performed. The aim of the hybridizations was to combine species with (1) either caudal or rostral position of the main stalk innervating the electrocytes in the electric organ and (2) short, median or long duration of their EOD. The hybrids, though they are still juveniles, show very interesting features concerning electrocyte geometry as well as EOD form and duration: the caudal position of the stalk is prevailing over the rostral position, and the penetration of the stalk is dominant over the non-penetrating feature (in the Campylomormyrus hybrids); in the hybrid between C. rhynchophorus and Gnathonemus petersii it is the opposite. When crossing species with long and short EODs, it is always the long duration EOD that is expressed in the hybrids. The F1-Hybrids of the cross C. tamandua×C. compressirostris are fertile: viable F2-fish could be obtained with artificial reproduction.

Weakly electric fish orient, hunt and communicate by emitting electrical pulses, enabling them to discriminate objects, conspecifics and prey. In addition to the electrosensory modality - although dominating in importance in these fishes - other modalities, like vision, play important roles for survival. The visual system of Gnathonemus petersii, a member of the family mormyridae living in West African blackwater streams shows remarkable specializations: Cone photoreceptors are grouped in bundles within a light reflecting tapetum lucidum, while the rods are also bundled but located at the back within a light-scattering guanine layer. Such an organization does not improve light sensitivity nor does it provide high spatial resolution. Thus, the function of the grouped retinal arrangement for the visual performance of the fish remains unclear. Here we investigated the contrast sensitivity of the temporal transfer properties of the visual system of Gnathonemus. To do so, we analyzed visual evoked potentials in the optic tectum and tested the critical flicker fusion frequency in a behavioral paradigm. Results obtained in Gnathonemus are compared to results obtained with goldfish (Carassius auratus), revealing differences in the filter characteristics of their visual systems: While goldfish responds best to low frequencies, Gnathonemus responds best at higher frequencies. The visual system of goldfish shows characteristics of a low-pass filter while the visual system of Gnathonemus has characteristics of a band-pass filter. Furthermore we show that the visual system of Gnathonemus is more robust towards contrast reduction as compared to the goldfish. The grouped retina might enable Gnathonemus to see large, fast moving objects even under low contrast conditions. Due to the fact that the electric sense is a modality of limited range, it is tempting to speculate that the retinal specialization of Gnathonemus petersii might be advantageous for predator avoidance even when brightness differences are small.

Summary Many sighted animals use color as a salient and reliable cue [ 1 ] to identify conspecifics [ 2 , 3 , 4 ], predators, or food [ 5 , 6 , 7 ]. Similarly, nocturnal, weakly electric fish Gnathonemus petersii might rely on “electric colors” [ 8 ] for unambiguous, critical object recognitions. These fish identify nearby targets by emitting electric signals and by sensing the object-evoked signal modulations in amplitude and waveform with two types of epidermal electroreceptors (active electrolocation) [ 9 , 10 , 11 , 12 ]. Electrical capacitive objects (animals, plants) modulate both parameters; resistive targets (e.g., rocks) modulate only the signal’s amplitude [ 11 , 12 ]. Ambiguities of electrosensory inputs arise when object size, distance, or position vary. While previous reports suggest electrosensory disambiguations when both modulations are combined as electric colors [ 8 , 13 , 14 ], this concept has never been demonstrated in a natural, behaviorally relevant context. Here, we assessed electric-color perception (1) by recording object-evoked signal modulations and (2) by testing the fishes’ behavioral responses to these objects during foraging. We found that modulations caused by aquatic animals or plants provided electric colors when combined as a ratio. Individual electric colors designated crucial targets (electric fish, prey insect larvae, or others) irrespective of their size, distance, or position. In behavioral tests, electrolocating fish reliably identified prey insect larvae of varying sizes from different distances and did not differentiate between artificial prey items generating similar electric colors. Our results indicate a color-like perceptual cue during active electrolocation, the computation [ 15 ], reliability, and use of which resemble those of color in vision. This suggests “color” perception as a sensory concept beyond vision and passive sensing.

The weaklyelectric fish Gnathonemus petersii detects, localizes, and analyzes objects during active electrolocation even in complete darkness. This enables these fish to lead a nocturnal life and find and identifytheir prey(small insect larvae) on the ground of their freshwater habitat. During active electrolocation, fish produce a series of brief electric signals, electric organ discharges (EOD), with an electric organ in their tail. Each EOD builds up a stable electric field around the fish, which is distorted onlyby nearbyobjects. Field distortions lead to changes of the transepidermal electric current flow at a region of the fish’s electroreceptive skin surface called the ‘electric image’. Within the electric image, locallyperceived EODs can be either altered in amplitude or waveform byan object. Fish measure both parameters to assess object properties, such as the capacitive and resistive components of the object’s complex impedance, the object’s size and shape, and its distance from the fish. None of these object properties can be evaluated in isolation, but have to be inferred during parallel processing of electric image spatial and qualitative parameters. Two anterior skin regions of G. petersii appear to possess particular properties for special electrolocation tasks and we therefore refer to them as ‘foveal’ regions. Because of its high electroreceptor density, the electric field geometry around it, and its behavioral use, the ‘nasal region’ between the nares and the mouth at the head of the fish is suggested to be a fovea for long-range guidance and object detection. We propose that the ‘Schnauzenorgan’, a long and flexible chin appendix covered denselywith electroreceptor organs, is a second electroreceptive fovea associated with a short-range (food) identification system. Together, these two electric foveae constitute an effective preydetection and identification sy stem. # 2003 Elsevier Ltd. All rights reserved.

We review modelling and experimental work dealing with the mechanisms of generation of electric image. We discuss: (1) the concept of electric image in the context of the reafference principle; (2) how waveform codes an impedance related qualia of the object image, referred to as “electric colour”; (3) that some characteristics of the spatial profiles generated by pre-receptor mechanisms are suitable for edge detection; (4) which parameters of the spatial profiles provide information for distance discrimination; (5) that electric images are distributed representations of the scene.

The weakly electric fish Gnathonemus petersii uses active electrolocation for non-visual orientation and object detection. Fish emit brief electric signals, which build up an electric field around their body. Nearby objects project electric images onto the fish's skin, which are perceived by epidermal electroreceptors, enabling the animal to discriminate between objects and to detect their electric properties. To test up to what distance object recognition is possible and whether complex backgrounds influence the discrimination performance, animals were trained to detect various objects of different sizes and shapes in front of different stagnant or moving backgrounds. Our results show that G. petersii is able to distinguish between objects up to a distance of about 3 cm even in complete darkness. Moving backgrounds improve the discrimination distance up to 4 cm. Using electric fish as models, underwater robots can be given an electric sense, which will enable them to navigate in dark or turbid environments. Our bionic approach will allow the robots to detect and analyse objects and recognize different object properties such as material composition, size and shape. Object recognition based on active electrolocation will still be possible in opaque surroundings where visual orientation is impossible.

Both routine electron microscopy and immunocytochemistry with protein A–gold were used to identify the cell types within the islet organs of four species of teleosts ( Osteoglossum bicirrhosum, Pantodon buchholzi, Notopterus chitala, and Gnathonemus petersii ) within Osteoglossomorpha, a subdivision with an ancient lineage. Four primary endocrine cell types, A, B, D, and F, were identified within the islets of the four species examined. The B- and D-cells were located mainly in the central core of the islet in the four species. In general, the A-cells were located at the islet periphery in all of the four species but in P. buchholzi and N. chitala they were also differently distributed toward the islet core. F-cells were present only at the islet periphery. Granules of B-cells in three species had a relatively homogeneous shape of the matrix core, but in O. bicirrhosum, the shape varied greatly. Variation in matrix shape of B-cell granules may indicate a different conformation of insulin molecules among at least some species of osteoglossomorphs, and this observation may have some taxonomic significance. Two somatostatin-containing (SST) D-cell types (D1 and DX) with granules of different shape were observed in all four species of osteoglossomorphs. The granules of the two D-cells immunostained either with anti-SST-25 and anti-SST-14 (D1-cells) or with anti-SST-34 (DX-cells). Immunocytochemistry confirmed that A-cells, containing glucagon-family peptides, and F-cells, containing peptides of the pancreatic polypeptide family, had granules of different shape. The cells of the islet organs of these osteoglossomorphs are more similar to those in more derived teleosts than they are to those of nonteleost actinopterygians.

In part I (Shuai et al. 1998. Biophys. J . 75:1712–1726), we presented a cellular model of the A- and B-electroreceptors of the weakly electric fish Gnathonemus petersii. The model made clear the cellular origin of the differences in the response functions of A- and B-receptors, which sensitively code the intensity of the fish's own electric organ discharge (EOD) and the variations in the EOD waveform, respectively. The main purpose of the present paper is to clarify the cellular origin of the inverse waveform tuning of the B-receptors by using the receptor model. Inverse waveform tuning means that B-receptors respond more sensitively to the 180° inverted EOD than to undistorted or less distorted EODs. We investigated how the A- and B-receptor models respond to EODs with various waveforms, which are the phase-shifted EODs, whose shift angle is varied from −1° to −180°, and single-period sine wave stimuli of various frequencies. We show that the tuning properties of the B-receptors arise mainly from the combination of two attributes: 1) The waveform of the stimuli ( B stim ) effectively sensed by the B-receptor cells. This consists of a first smaller and a second larger positive peak, even though in the original phase-shifted EOD stimuli, the amplitudes of the two positive peaks are reversed. 2) The effective time constant of dynamical response of the receptor cells. It is on the order of the duration of a single EOD pulse. We also calculated the response properties of the A- and B-receptor models when stimulated with natural EODs distorted by various capacitive and resistive objects. Furthermore, we investigated the effect of EOD amplitude on the receptor responses to capacitive and resistive objects. The models presented can systematically reproduce the experimentally observed response properties of natural A- and B-receptor cells. The mechanism producing these properties can be reasonably explained by the variation in the stimulus waveforms effectively sensed by the A- and B-receptor cells and by time constants.

Active electroreception is an ancestral trait found in many aquatic vertebrates and has evolved independently in two teleost lineages, the Gymnotiformes and the Mormyriformes. Unique to these so-called weakly electric fish is their ability to actively generate electrical currents in the water and sense the electrical properties of the environment. How natural behavior contributes to this sensory system has been of interest to neuroethologists since the pioneering works of Lissmann. Here we report on a mutual modeling and experimental study of the stimuli available during active electrolocation of Gnathonemus petersii (Mormyridae). We show the validity of the model (I) by demonstrating that localized spatial patterns of object induced modulations in the electric field (electric images) are comparable to experimentally mapped 2-dimensional electric images and (II) by replicating earlier key findings showing that a normalized metric of electric image width provides an unambiguous cue for distance estimation. We then show that electric images and the distance metric vary systematically when an object is moved along the trunk. These potential ambiguities with regard to localization lead us to a spatiotemporal analysis of electric images. We introduce a new temporal metric for distance estimation that is based on the normalized spatial properties of electrical images. Finally, based on a survey of exploratory behavior, we show how objects situated at the tail, a region previously neglected, cast global electric images that extend over the whole sensory epithelium of the animals.

Abstract. Newly imported, adult weakly discharging electric fish Gnathonemus petersii , were examined for sex differences in the waveform of their electric organ discharge and in external morphology. The discharge waveform of gonadally ripe subjects are quantitatively described at three stages corresponding to the Nigerian pre-rainy, rainy breeding, and post-rainy season in 1988. A sex difference was found in external morphology, with males displaying a distinct indentation at the dorsal margin of the anal fin, regardless of season. Pronounced sex differences in electric organ discharges were found only in fish imported during the Nigerian rainy season, and only in two phase of the tetraphasic discharge, with males exhibiting longer durations of phases 2 and 3. These sex differences were reflected in the Fourier transformation, with male discharges exhibiting lower peak power spectral frequencies than those of females. No sex differences were found in total duration of the electric discharge. Correlations between the electric discharge measures and fish length, as well as between the discharge measures themselves, revealed relationships which varied in both direction and degree as a function of sex and/or season. These results indicate that the duration of individual phases, rather than the entire duration of the electric organ discharge, function in the communication of sex differences in this mormyrid species. Previous work has shown that phase durations of the electric discharge are steroid-dependent and therefore indicative of physiological states of the animal. Thus, the kinds of information conveyed during breeding and non-breeding seasons may be quite different, given great differences in gonadal steroid levels. These findings explain previously unsuccessful attempts to find sex differences as well as a report of a sex difference in a direction opposite to that shown here in G. petersii.

The present paper describes the location, morphology, ultrastructure and immunocytochemical properties of neurons in the brain of the mormyrid fish Gnathonemus petersii, that appear to be unspecifically immunoreactive to a number of secondary or tertiary antibodies used in immunohistochemical procedures, including rabbit-anti-mouse immunoglobulins (IGGs), rabbit peroxidase-anti-peroxidase IGGs, and rabbit-anti-sheep or sheep peroxidase-anti-peroxidase IGGs. Unspecifically immunoreactive (UI) cells have typically neuronal morphological and ultrastructural characteristics, and occur at four specific locations in the mormyrid brain. A small rhombencephalic group is located rostrolateral to the efferent octaval nucleus, between the fasciculus longitudinalis medialis and the decussation of the lateral lemniscus. A mesencephalic cluster of cells is located in the dorsal midbrain tegmentum against the tractus telencephalo-mesencephalicus. In addition, dispersed UI neurons were observed in the nucleus lateralis of the torus semicircularis and in the preoptic region above the optic chiasm. Remarkably, UI cells are clearly present in a substantial number of brains investigated, but not detectable in others. The present findings point to a curious pitfall in chemical neuroanatomy, the functional significance of which is unknown at present. In several previous studies using the brain of G. petersii, UI cells were abusively included in the description of monoaminergic cell groups. Similar cells have until now not been reported in other vertebrate brains.

1. l. Specimens of Gnathonemus petersii were restrained for recording ECG, EOD, cerebellar and telencephalic EEG responses to experimental, environmental stimuli. 2. 2. All novel stimuli induced a bradycardia and increase in EOD discharge rate. 3. 3. Subsequently alumina was topically applied to the brain, inducing electroencephalographic seizures after 5 hr. 4. 4. Seizures were most evident in the cerebellar EEG where they followed a large rise in the rate of EODs, which then ceased. 5. 5. In most animals seizures were terminal. EOD rates may be useful quantitative measures of arousal and seizures.

Weakly electric fish can serve as model systems for active sensing because they actively emit electric signals into the environment, which they also perceive with more than 2000 electroreceptor organs (mormyromasts) distributed over almost their entire skin surface. In a process called active electrolocation, animals are able to detect and analyse objects in their environment, which allows them to perceive a detailed electrical picture of their surroundings even in complete darkness. The African mormyrid fish Gnathonemus petersii can not only detect nearby objects, but in addition can perceive other properties such as their distance, their complex electrical impedance, and their three-dimensional shape. Because most of the sensory signals the fish perceive during their nightly activity period are self-produced, evolution has shaped and adapted the mechanisms for signal production, signal perception and signal analysis by the brain. Like in many other sensory systems, so-called prereceptor mechanisms exist, which passively improve the sensory signals in such a way that the signal carrier is optimized for the extraction of relevant sensory information. In G. petersii prereceptor mechanisms include properties of the animal's skin and internal tissue and the shape of the fish's body. These lead to a specific design of the signal carrier at different skin regions of the fish, preparing them to perform certain detection tasks. Prereceptor mechanisms also ensure that the moveable skin appendix of G. petersii, the 'Schnauzenorgan', receives an optimal sensory signal during all stages of its movement. Another important aspect of active sensing in G. petersii concerns the locomotor strategies during electrolocation. When foraging, the animals adopt a particular position with the body slanted forward bringing the so-called 'nasal region' in a position to examine the environment in front of and at the side of the fish. Simultaneously, the Schnauzenorgan performs rhythmic left-right searching movements. When an object of interest is encountered, the Schnauzenorgan is brought in a twitching movement towards the object and is moved over it for further exploration. The densities of electroreceptor organs is extraordinary high at the Schnauzenorgan and, to a lesser extend, at the nasal region. In these so-called foveal regions, the mormyromasts have a different morphology compared to other parts of the electroreceptive skin. Our results on mormyromast density and morphology, prereceptor mechanisms and electric images, central processing of electroreceptive information, and on behavioural strategies of G. petersii lead us to formulate the hypothesis that these fish possess two separate electric foveae, each of which is specialized for certain perceptional tasks.

Substance P has been proposed as a candidate neurotransmitter or neuromodulator in the nociceptive system. Using a light microscopial immunohistochemical peroxidase-anti-peroxidase technique we have detected high substance P-like immunoreactivity (SPLI) in several types of sensory organs of 4 species of gymnotiform teleost fish: olfactory epithelium, vestibular, lateral line and electrosensory organs. The olfactory nerve and its endings within the olfactory bulb and the telencephalon were also strongly labelled. At these sites no SPLI was revealed in other teleosts (Carassius auratus, Gnathonemus petersii). The findings suggest that substance P may be involved in neurotransmission or neuromodulation in these specific sensory systems of these species.

The cutaneous ampullary organs (type A) of the specific lateral line system in Gnathonemus petersii (mormyrids) have been investigated with ultrastructural methods. The stratified epithelium of the ampulla contains sensory cells which bear microvilli on their luminal surface and many sensorineural junctions on their basal cytoplasm. All nerve terminals for a single organ are issued from a single lateral nerve. The epidermal basal membrane only is in contact with the accessory cells. The three types of accessory cells show excretory functions. They fill up the ampulla and the intraepidermic canal with morphologically different mucopolysaccharides. The accessory cells of the type III appear to be similar to “goblet” cells of the intestinal epithelium.

The rhombencephalic posterior lateral line lobe nucleus (nLLL) and its connections were investigated in the mormyrid fish Gnathonemus petersii, at light and electron microscopical levels using HRP tracing. The nLLL, constituted on each side of about 1500 large, round shaped, adendritic cells, is located in the intermediate cell and fibre layer exclusively, in the ventrolateral zone of the posterior lateral line lobe. The cells show a complex synaptology: boutons with chemical synapses cover the largest part of the soma and the long initial segment of the axons. In addition each nLLL cell bears generally two club endings which form gap junctions with the postsynaptic membrane. On the unmyelinated portion of the club ending, a particular synaptic complex (= serial synaptic connections) was observed; large endings, bearing boutons with chemical synapses, contact the club ending with gap junctions. HRP injection into either lateral line nerve showed that the club endings represent the peripheral input of the nLLL. This input is exclusively ipsilateral; the anterior and posterior lateral line nerve projection does not seem to overlay at this level. Retrograde labelling of nLLL cells after HRP injection into the anterior mesencephalic exterolateral nucleus (nELa) confirms the electrophysiological results according to which the nLLL projects directly and bilaterally to the nELa. This rhombo-mesencephalic connection is established through club endings of the nLLL axons which form gap junctions with the large cells of the nELa. About two thirds of the nLLL axons form a crossed and one third an uncrossed bundle within the lateral lemniscal pathway. Anterograde transport in the axon collaterals shows that some of the nLLL cells project simultaneously to the ipsi- and to the contralateral nELa permitting this system a high degree of electrosensory information processing.

A technically simple fish monitor was developed using the weakly electric fish, Gnathonemus petersi, as the test organism. The test system consists of a plexiglass test chamber (21., 27°C) with a flow-through of 4–5 times the chamber volume per hour. The spontaneous electric organ discharges (EOD) are recorded as activity signals. EODs, which have a potential of about 500 mV, last 0.3 ms and they can be converted to computer-compatible signals by a simple amplifier. The EOD-activity is analyzed by a low-cost desk-top computer. Two different modes of analysis which represent two levels of response sensitivity, are possible either alternatively or simultaneously. The toxicity response of Gnathonemus petersi was measured in experiments with the toxicants Hg2+, Cu2+, NaAsO2 and CN−. The sensitivity to acutely toxic concentrations was found to be as high as that of trout or minnows.

The connections of a cerebellar structure, the eminentia granularis ( e.gr. ) in the mormyrid fish Gnathonemus petersii were demonstrated by the HRP method. Injections of horseradish peroxydase in this area and in several cranial nerves enabled us to distinguish two regions in this extensive granular structure: the pars anterior ( p.a. ) which receives root fibres from the lateral line and vestibular nerves, and the pars posterior (p.p.) towards which converge afferent fibres from the spinal cord, the medulla oblongata and the mesencephalon. In the light of these results the former is considered as a primary centre for specific sensory impulses and will be called the eminentia granularis, sensu strictiori ; in contrast the latter, which may be considered as the auricula , represents an integration centre for secondary impulses arriving from the common (cutaneous, proprioceptive) sensory system.

The sensory basis of group cohesion in the weak-electric fish Gnathonemus petersii was investigated in a circular tank with groups of four fish each, interacting through a wide-meshed plastic screen with intact or operated conspecifics, or with other stimulus objects. We confined these stimuli to one or two peripheral holding compartments. The response measures were obtained from the free swimming fish and included (1) the time the fish spent together as a group, (2) the time they spent in front of the holding compartments, (3) the circular distribution of the fish's positions, and (4) the mean nearest neighbour distances. Under empty compartment conditions, four different groups were tested, consisting of either (1) intact, electrically active fish, or (2) electrically ‘silent’ fish (with their electric organ surgically rendered inoperative), or (3) blind, or (4) ‘silent’ and blind animals. The loss of either sensory modality, vision or feedback from electric organ discharge, led to changes of comparable size, decreasing the time spent as a group and increasing the mean nearest neighbour distance. In fish lacking both modalities, group cohesion was further impaired. With stimuli present in one or both holding compartments, the strength of social attraction depended on the nature of the stimulus: the more intact stimulus conspecifics were present, the more densely did the fish group in front of the stimulus compartment. ‘Wired-in’ electric organ discharges (simulating waveform and intensity) and electrically ‘silent’ fish were equally attractive, but only half as attractive as intact fish. Blind free swimming fish aggregated with intact and also with ‘silent’ conspecifics. Under dim light conditions, group cohesion was predominantly, though not exclusively, affected by electrosensory feedback from the electric organ discharge and visual input. Mechanical and olfactory cues may also be involved.

We investigated group cohesion in four species of African weakly electric fish (Brienomyrus niger, Gnathonemus petersii, Marcusenius cyprinoides and Pollimyrus isidori). The attraction among members of the same and different species served as the criterion for species recognition. To identify possible mechanisms underlying this ability, conspecifics were allowed to interact through a wide-meshed plastic screen with either a pair of confined conspecifics or a pair from a related species. Members of each of the four test species were optimally attracted to their own kind, but also responded selectively to the presence of the other species. These interspecific interactions ranged from attraction to avoidance. The difference in the fish's preference to aggregate in inter- and intraspecific interactions pointed to species-specific social cues that facilitate group cohesion in mormyrid fish. Since all four species respond to each other's electric organ discharge with changes in their own discharge rate, species recognition cannot be merely a function of electroreceptor characteristics but must involve the integration of electroreceptive and other sensory cues.

In the presence of a novel stimulus such as an aluminum or a plastic rod mormyrid fish (Marcusenius cyprinoides and Gnathonemus petersii) exhibit characteristic motor probing acts (PMAs): 'chin probing', 'radial' and 'lateral va-et-vient', 'lateral probing', 'tangential probing', and 'stationary probing'. During the display of these PMAs the fish maintain characteristic probing distances from the object which were 4.5 cm for the metal and 2.5-3.7 cm for the plastic stimulus. G. petersii with their electric organ rendered inoperative ('silent fish') no longer exhibited 'radial' and 'lateral va-et-vient'. Regardless of the nature of the stimulus the probing distances were shorter in 'silent' fish and ranged from 1.8 to 2.6 cm. During the display of PMAs intact fish changed their variable electric organ discharge rate to a unique and stable, regularized rate, with the interdischarge interval maintained at 28-30 ms. The fish's electric and non-electric (motor) probing behavior in the presence of novel objects (and following their removal during phantom PMAs) is discussed in light of theories on exploratory behavior.

The electric organ discharges (EODs) of pairs of weakly electric fish, Gnathonemus petersii, were simultaneously recorded to study the significance of the EODs as communication signals. In a 400-litre tank a larger fish (12 to 15 cm) was passively moved within a shelter tube toward a smaller specimen (6 to 9 cm), either in steps or a continuous move. The movement was stopped at that distance when at least one fish significantly lowered or ceased its EOD activity. From this ‘threshold interfish distance’ the spatial range of a ‘communication field’ was found to extend about 30 cm from the fish. At threshold distances an EOD frequency increase caused a temporary EOD activity cessation in the second fish. The spontaneous irregular EOD pattern of the fish displaying the increased EOD rate changed into a regular one with almost equal time intervals between fish pulses.

The weakly electric fish, Gathonemus niger, discharged with a frequency of 4 to 8 Hz during the day and 10 to 16 Hz during the night. The frequency of superimposed burst discharges (32 to 56 Hz) was independent of diurnal factors. The variation of the electric organ discharge frequency during the day was investigated in response to controlled electric stimulus patterns: (a) A free running stimulus frequency of 4 Hz, simulating the resting frequency of another fish, and different stimulus intensities, simulating different distances between two fish. (b) Free running frequencies of 4, 8, 16, …, 128 Hz and two particular stimulus intensities. (c) Discharge coupled stimuli (each discharge triggered an electric stimulus with a fixed delay) and different stimulus intensities. All three kinds of stimuli elicited defined and predictable response discharge patterns supporting the assumption that an electric fish would respond to a particular discharge pattern of another fish also in a similar and predictable manner. Low stimulus intensities (0·04 to 0·2 mV per cm) caused cessation of the discharge activity, a ‘hiding’ or ‘listening’ response. The discharge rate increased linearly with the logarithm of the stimulus intensity. The fish was particularly sensitive to stimulus frequencies which simulated its burst activity (32 to 56 Hz). Discharge coupled stimuli showed that the fish responded to about eight times lower stimulus intensities if the stimulus occurred between two discharges (15 to 30 m-s after the fish's discharge) than if the stimulus occurred within or immediately after the discharge. All suprathreshold stimuli elicited a typical discharge pattern: The irregular resting discharge activity became significantly regular. The degree of regularity was even improved during maintained stimulation. The regularisation of the discharge activity is thought to be involved in the fish's electrolocating system whereas frequency variations are considered as being involved in both the locating system and as communication signals among weakly electric fish.

Light and electron microscopic observations on the type B organ (mormyromast) of the lateral line complex of mormyrids reveal the existence of two types of sensory cells each with distinct separate innervation. The type A cell is specific for this organ whereas the type B cell is almost identical with that of the tuberous organs of mormyrids. The type A cells are located at the base of an intradermal jelly sphere. The type B cells are found within a separate sensory chamber connected to the jelly sphere with a small duct. The type B censory cells are attached to a highly differentiated accessory platform in which three types of accessory cells can be distinguished. The morphological data suggest that the type B organ can provide, by two channels, sensory informations concerned with the same external event, i.e., can detect two different parameters of the same external stimulus.

In Mormyrid fish the electric organ appears to discharge as a single unit, and the frequency of discharge is a simply quantifiable behavioural variable. Following a study showing that acceleration in discharge rate can be conditioned classically, demonstration of operant conditioning was undertaken using a yoked control procedure. For each pair of fish six to twelve series of trials were run, and in successive series the avoidance and control contingencies were reversed. The procedure permitted comparison of the contingencies successively in the same fish and simultaneously between fish. The results from eight Gnathonemus indicated that avoidance conditioning can control electric organ activity. In naive subjects avoidance conditioning produced significantly more responding than yoked classical conditioning (P > 0·02). The occurrence of avoidance conditioning was confirmed by the effects of subsequent reversals of avoidance and control schedules.

Mormyrids, African weakly electric fish, display a low frequency electric discharge generated by the caudally located electric organ between the anal, dorsal and caudal fins. Each electric pulse has a diphasic form and a duration of 0.5 msec. The properties of sensory impulses recorded in single lateral line fibers of Gnathonemus petersi and stanleyanus were studied. Many fibers in the dorsal branch of the lateral line nerve responded with 1–2 or several (up to five) impulses to each electric organ pulse. Regardless of the number of sensory impulses, a shortening or lengthening of their latency with respect to the electric organ pulse (latency change) was observed, when a conductive (silverplate) or non-conductive (plastic plate) object was placed at the level of the receptor. The range of the latency change appeared to be larger (7–8 msec) in fibers which displayed a maximal discharge of one or two impulses than in those which at maximum displayed several impulses. In the latter, a variation of latency occurs along with a change in number of impulses. Conductive and non-conductive objects have opposite effects. An imposed electric current pulse in the water elicited fiber responses similar to those evoked by the electric organs pulse. Increasing pulse strength produced a smooth latency shortening of sensory impulses.