Comparative brain morphology of cleaning and sponge-dwelling Elacatinus gobies.

Introduction Comparative studies of brain anatomy between closely related species have been very useful in demonstrating selective changes in brain structure. Within-species comparisons can be particularly useful for identifying changes in brain structure caused by contrasting environmental selection pressures. Here, we aimed to understand whether differences within and between species in habitat use and foraging behaviour influence brain morphology, on both ecological and evolutionary time scales. Methods We used as a study model three species of the Elacatinus genus that differ in their habitat-foraging mode. The obligatory cleaning goby Elacatinus evelynae inhabits mainly corals and feeds mostly on ectoparasites removed from larger fish during cleaning interactions. In contrast, the obligatory sponge-dwelling goby Elacatinus chancei inhabits tubular sponges and feeds on microinvertebrates buried in the sponges' tissues. Finally, in the facultatively cleaning goby Elacatinus prochilos, individuals can adopt either phenotype, the cleaning or the sponge-dwelling habitat-foraging mode. By comparing the brains of the facultative goby phenotypes to the brains of the obligatory species we can test whether brain morphology is better predicted by phylogenetic relatedness or the habitat-foraging modes (cleaning x sponge-dwelling). Results We found that E. prochilos brains from both types (cleaning and sponge-dwelling) were highly similar to each other. Their brains were in general more similar to the brains of the most closely related species, E. evelynae (obligatory cleaning species), than to the brains of E. chancei (sponge-dwelling species). In contrast, we found significant brain structure differences between the cleaning species (E. evelynae and E. prochilos) and the sponge-dwelling species (E. chancei). These differences revealed independent changes in functionally correlated brain areas that might be ecologically adaptive. E. evelynae and E. prochilos had a relatively larger visual input processing brain axis and a relatively smaller lateral line input processing brain axis than E. chancei. Conclusion The similar brain morphology of the two types of E. prochilos corroborates other studies showing that individuals of both types can be highly plastic in their social and foraging behaviours. Our results in the Elacatinus species suggest that morphological adaptations of the brain are likely to be found in specialists whereas species that are more flexible in their habitat may only show behavioral plasticity without showing anatomical differences.

Evolution of the visual system in ray-finned fishes.

The vertebrate eye allows to capture an enormous amount of detail about the surrounding world which can only be exploited with sophisticated central information processing. Furthermore, vision is an active process due to head and eye movements that enables the animal to change the gaze and actively select objects to investigate in detail. The entire system requires a coordinated coevolution of its parts to work properly. Ray-finned fishes offer a unique opportunity to study the evolution of the visual system due to the high diversity in all of its parts. Here, we are bringing together information on retinal specializations (fovea), central visual centers (brain morphology studies), and eye movements in a large number of ray-finned fishes in a cladistic framework. The nucleus glomerulosus-inferior lobe system is well developed only in Acanthopterygii. A fovea, independent eye movements, and an enlargement of the nucleus glomerulosus-inferior lobe system coevolved at least five times independently within Acanthopterygii. This suggests that the nucleus glomerulosus-inferior lobe system is involved in advanced object recognition which is especially well developed in association with a fovea and independent eye movements. None of the non-Acanthopterygii have a fovea (except for some deep sea fish) or independent eye movements and they also lack important parts of the glomerulosus-inferior lobe system. This suggests that structures for advanced visual object recognition evolved within ray-finned fishes independent of the ones in tetrapods and non-ray-finned fishes as a result of a coevolution of retinal, central, and oculomotor structures.

The Diversity of the Brains of Ray-Finned Fishes.

Brains are very plastic, both in response to phenotypic diversity and to larger evolutionary trends. Differences between taxa cannot be easily attributed to either factors. Comparative morphological data on higher taxonomic levels are scarce, especially in ray-finned fishes. Here we show the great diversity of brain areas of more than 150 species of ray-finned fishes by volumetric measurements using block-face imaging. We found that differences among families or orders are more likely due to environmental needs than to systematic position. Most notable changes are present in the brain areas processing sensory input (chemosenses and lateral line vs. visual system) between salt- and freshwater species due to fundamental differences in habitat properties. Further, some patterns of brain volumetry are linked to characteristics of body morphology. There is a positive correlation between cerebellum size and body depth, as well as the presence of a swim bladder. Since body morphology is linked to ecotypes and habitat selection, a complex character space of brain and body morphology and ecological factors together could explain better the differentiation of species into their ecological niches and may lead to a better understanding of how animals adapt to their environment.

Sex Differences in the Swordtail Xiphophorus hellerii Revealed by a New Method to Investigate Volumetric Data.

Comparing the relative volumes of body parts is a useful tool in morphology, but it is not trivial to do this in animals that differ in overall size. To account for scaling differences, a "reference size" has to be determined and the original absolute volumes have to be "corrected for" by this scaling reference. However, the outcome of a statistical analysis is greatly affected by this "reference size," and it is practically impossible to determine the "overall size" of a structure independent of the changes in the relative size of the parts of it. Here, a new method is introduced to compare the relative volumes of parts that does not need a scaling reference. The method transforms the absolute part volumes into a ratio matrix (volume ratio transformation, VRT). The VRT is free of any scaling factors and can be used to compare groups of animals. This paper also reviews various other errors made frequently when comparing brain morphology between animals. Finally, the VRT is applied to investigate sex differences in the swordtail fish (Xiphophorus hellerii), which show profound differences in the size of the valvula cerebelli.

Local differences in calretinin immunoreactivity in the optic tectum of the ocellated dragonet.

The optic tectum of the ocellated dragonet (Synchiropus ocellatus) was studied with immunohistochemistry. Antibodies raised against the calcium binding protein calretinin (CR) revealed a lamination similar to that already reported for other ray finned fish. Most immunoreactive fibers could be observed in those layers receiving retinal afferents and most immunoreactive cells occur in the stratum periventriculare. However, there are marked differences in the presence of other calretinin-positive cell types and immunoreactive lamina between the dorsomedial and ventrolateral parts of the tectum. Synchiropus is a bottom dwelling fish with strong functional subdivisions of the visual system into dorsal and lateral visual fields. The differences in calretinin-positive cell bodies and fibers may be a sensitive indicator of functional differences of tectal circuitry.

A direct primary afferent projection of the trigeminal nerve to the valvula cerebelli in the spiny eel Macrognathus zebrinus.

This study is a re-examination of the direct primary sensory input to the valvula cerebelli in spiny eel. The valvula in Macrognathus zebrinus receives a primary afferent projection from the trigeminal nerve as revealed by injections of biotinylated dextran amines into the rostrum. The descending trigeminal nucleus and nucleus of the tractus solitarius are innervated as well. Injections with DiI into the valvula labeled fibers in the descending trigeminal nucleus. The projection of tactile information from the rostrum to the valvula may be an adaptation of food search for spiny eels in regard to their highly mobile rostrum.

Forebrain organization in elasmobranchs.

It has long been known that many elasmobranch fishes have relatively large brains. The telencephalon, in particular, has increased in size in several groups, and as a percent of total brain weight, it is as large as in some mammals. Little is known, however, about the organization, connections, and functions of the telencephalon in elasmobranchs. Early experimental studies indicated that olfaction does not dominate the telencephalon and that other sensory modalities are represented, particularly in the pallium. We have investigated the intrinsic and extrinsic connections of the telencephalon in two elasmobranch species: the thornback guitarfish, Platyrhinoidis triseriata, and the spiny dogfish, Squalus acanthias. Tracers were injected into various parts of the forebrain and olfactory pathways were found to be extensive and were seen to involve the pallium. Injections into various parts of the pallium revealed a major input from the area basalis, which receives secondary and tertiary olfactory fibers. Nonolfactory input from the diencephalon appeared relatively minor and seemed to converge with olfactory information in the dorsal pallium and area superficialis basalis. Major descending projections were seen to originate in the dorsal pallium and terminate in the hypothalamus and - in the case of Platyrhinoidis - massively in the lateral mesencephalic nucleus. Descending pathways appeared mainly crossed in Platyrhinoidis, but not in Squalus. Our data indicate that the concept of the dorsal pallium as a nonolfactory area in elasmobranchs must be reconsidered, and we suggest that many telencephalic centers, including the dorsal pallium, are involved in olfactory orientation.

Two modes of information processing in the electrosensory system of the paddlefish (Polyodon spathula).

Paddlefish are uniquely adapted for the detection of their prey, small water fleas, by primarily using their passive electrosensory system. In a recent anatomical study, we found two populations of secondary neurons in the electrosensory hind brain area (dorsal octavolateral nucleus, DON). Cells in the anterior DON project to the contralateral tectum, whereas cells in the posterior DON project bilaterally to the torus semicircularis and lateral mesencephalic nucleus. In this study, we investigated the properties of both populations and found that they form two physiologically different populations. Cells in the posterior DON are about one order of magnitude more sensitive and respond better to stimuli with lower frequency content than anterior cells. The posterior cells are, therefore, better suited to detect distant prey represented by low-amplitude signals at the receptors, along with a lower frequency spectrum, whereas cells in the anterior DON may only be able to sense nearby prey. This suggests the existence of two distinct channels for electrosensory information processing: one for proximal signals via the anterior DON and one for distant stimuli via the posterior DON with the sensory input fed into the appropriate ascending channels based on the relative sensitivity of both cell populations.

Nonlinear dynamics of skin potentials in the electrosensory paddlefish.

It is known that steady skin potentials are present in fishes due to chloride pumps in the gills and in the skin. We have found previously that these skin potentials can fluctuate and oscillate in the electrosensory paddlefish. Here we show that larger, discharge like potentials can be triggered by applying external electric fields in the water surrounding the fish. These resemble action potentials in nerve cells, but have a longer time scale. Like action potentials, these discharges travel laterally in the skin. They start at the tip of the rostrum and propagate caudally to the tip of the gill covers. They follow the all-or-nothing rule and need some refractory period before they can be evoked again. This is the first time that such discharges, so strikingly similar to action potentials, have been described at the level of a whole organism.

Descending projections to the hindbrain and spinal cord in the paddlefish Polyodon spathula.

In vertebrates, almost all motor neurons innervating skeletal muscles are located in the hindbrain and spinal cord, and all brain centers that control behavior have descending projections into these parts of the central nervous system. With tracer injections into the spinal cord and hindbrain, we have studied cell groups with descending projections in the paddlefish. Spinal cord injections reveal retrogradely labeled cells in all reticular and raphe nuclei, as well as the nucleus of the medial longitudinal fascicle. Additional cell groups with projections to the spinal cord are the nucleus of the fasciculus solitarius, descending trigeminal nucleus, several octavolateral nuclei, the dorsal hypothalamic nucleus, and the pretectum. The only primary sensory fibers with descending projections are trigeminal fibers. Hindbrain injections reveal a number of additional cell groups in di- and mesencephalon. The most prominent source is the mesencephalic tectum. Other descending cells were found in the dorsal posterior thalamic nucleus, ventral thalamus, torus semicircularis, lateral mesencephalic nucleus, and the central gray of the mesencephalon. Our data show that descending spinal projections are comparable to those of other vertebrates and that the tectum is the most important motor control center projecting to the hindbrain. A surprising result was that the dorsal posterior thalamic nucleus also projects to the hindbrain. This nucleus is thought to be a center that relays sensory information to the telencephalon. Further studies are needed to determine the complete set of projections of the dorsal thalamus in paddlefish and other fishes to gain insights into its functional role.

Edge-detection filter improves spatial resolution in the electrosensory system of the paddlefish.

In many fishes, prey capture is guided primarily by vision. In the paddlefish, the electrosense can completely substitute for the visual system to detect tiny daphnia, their primary prey. Electroreceptors are distributed over the entire rostrum, head, and gill covers, and there are no accessory structures like a lens to form an image. To accurately locate planktonic prey in three-dimensional space, the poor spatial resolving power of peripheral receptors has to be improved by another mechanism. We have investigated information processing in the electrosensory system of the paddlefish at hind- and midbrain levels by recording single cells extracellularly. We stimulated with a linear array of electrodes that simulated a moving dipole field. In addition, global electric fields were applied to simulate the temporal component of a moving dipole only. Some stimulation were done with sinusoidal fields. The fire rate of cells in the hindbrain followed the first derivative of the stimulus wave form. In contrast, the response of tectal cells were similar to the third derivative. This improves spatial resolution and receptive fields of tectal units are much smaller than the ones of hind brain units. The principle is similar to a Laplacian of Gaussian filter that is commonly used in digital image processing. However, instead of working in the space domain, the paddlefish edge detection filter works in the time domain, thus eliminating the need for extensive interconnections in an array of topographically organized neurons.

Two parallel ascending pathways from the dorsal octavolateral nucleus to the midbrain in the paddlefish Polyodon spathula.

The paddlefish is a passive electrosensory ray-finned fish with a special rostral appendage that is covered with thousands of electroreceptors, which makes the fish extremely sensitive to electric fields produced by its primary prey, small water fleas. We reexamined the electrosensory pathways from the periphery to the midbrain by injecting the neuronal tracer BDA into different branches of the lateral line nerve and into different parts of the dorsal octavolateral nucleus (DON) and the tectum. Primary afferents from the anterior to posterior body axis terminate in different areas in the mediolateral axis of the DON, the first electrosensory processing station. Previous studies showed that DON neurons project to the tectum and two different areas in the tegmentum. Now, we have found differences between the anterior and the posterior DON. Fibers from the anterior DON project unilaterally to the contralateral tectum while its posterior neurons project bilaterally to two nuclei in the tegmentum, the torus semicircularis and the lateral mesencephalic nucleus. This study is the first to show that two different populations of ascending neurons project to two different targets in the midbrain. These two pathways are likely to have different functions and further investigations may reveal the functional significance of these two parallel ascending systems.

Measuring flow velocity and flow direction by spatial and temporal analysis of flow fluctuations.

If exposed to bulk water flow, fish lateral line afferents respond only to flow fluctuations (AC) and not to the steady (DC) component of the flow. Consequently, a single lateral line afferent can encode neither bulk flow direction nor velocity. It is possible, however, for a fish to obtain bulk flow information using multiple afferents that respond only to flow fluctuations. We show by means of particle image velocimetry that, if a flow contains fluctuations, these fluctuations propagate with the flow. A cross-correlation of water motion measured at an upstream point with that at a downstream point can then provide information about flow velocity and flow direction. In this study, we recorded from pairs of primary lateral line afferents while a fish was exposed to either bulk water flow, or to the water motion caused by a moving object. We confirm that lateral line afferents responded to the flow fluctuations and not to the DC component of the flow, and that responses of many fiber pairs were highly correlated, if they were time-shifted to correct for gross flow velocity and gross flow direction. To prove that a cross-correlation mechanism can be used to retrieve the information about gross flow velocity and direction, we measured the flow-induced bending motions of two flexible micropillars separated in a downstream direction. A cross-correlation of the bending motions of these micropillars did indeed produce an accurate estimate of the velocity vector along the direction of the micropillars.

Lateral line nerve fibers do not code bulk water flow direction in turbulent flow.

The discharges of anterior and posterior lateral line nerve afferents were recorded while stimulating goldfish, Carassius auratus, with bulk water flow. With increasing flow velocity lateral line afferents increased their discharge rates. However, an increased response to flow rates occurred even if flow direction was reversed. Thus, individual lateral line afferents did not encode the direction of running water. Frequency spectra of the water motions quantified with particle image velocimetry revealed flow fluctuations that increased with increasing flow velocity. Maximal spectral amplitudes of the flow fluctuations were below 5 Hz (bulk flow velocity 4-15 cms(-1)). The frequency spectra of the firing rates of lateral line afferents also showed an increase in amplitude when fish were exposed to running water. The maximal spectral amplitudes of the recorded data were in the frequency range 3-8 Hz. This suggests that the lateral line afferents mainly responded to the higher frequency fluctuations that developed under flow conditions, but not to the direct current flow or the lower frequency fluctuations. Although individual lateral line afferents encoded neither flow velocity nor flow direction we suggest that higher order lateral line neurons can do so by monitoring flow fluctuations as they move across the surface of the fish.

Organization of major telencephalic pathways in an elasmobranch, the thornback ray Platyrhinoidis triseriata.

The forebrain of elasmobranchs is well developed, and in some species the relative brain/body weight is comparable to that in mammals. However, little is known about the organization of major telencephalic pathways. We injected biotinylated dextran amines into the olfactory bulb, lateral pallium, dorsomedial pallium, and the forebrain bundles of the thornback ray, Platyrhinoidis triseriata. Secondary olfactory fibers from the bulb innervate the lateral pallium, the ventral division of the rostral telencephalon and area superficialis basalis. Retrogradely labeled cells were seen exclusively in the lateral periventricular area. The projections of the lateral pallium appeared basically similar to those of the olfactory bulb, but labeling was much denser in the superficial part of area basalis. Some fibers were also seen to innervate the posterior tuberal nucleus. Injections into the dorsomedial pallium revealed a major input from area basalis. Only a few cells were retrogradely labeled in the dorsal thalamus and posterior lateral thalamic nucleus. Major efferents of the dorsomedial pallium appear to reach the contralateral inferior lobe of the hypothalamus and the lateral mesencephalic nucleus. Tracer injections into the forebrain bundles retrogradely labeled many cells in the diencephalon and the mesencephalon and also revealed terminal fields in area superficialis basalis. In addition, a large number of cells were labeled in the dorsomedial pallium. Descending telencephalic fibers innervate heavily the inferior lobes and the lateral mesencephalic nucleus. Our results show that higher order olfactory pathway courses from the lateral pallium through area basalis to the dorsomedial pallium and that ascending non-olfactory input is integrated in area superficialis basalis and the dorsal pallium along with olfactory information, rather than being processed in separate, non-olfactory centers.

Response properties of electrosensory neurons in the lateral mesencephalic nucleus of the paddlefish.

Many fishes and amphibians are able to sense weak electric fields from prey animals or other sources. The response properties of primary afferent fibers innervating the electroreceptors and information processing at the level of the hindbrain is well investigated in a number of taxa. However, there are only a few studies in higher brain areas. We recorded from electrosensory neurons in the lateral mesencephalic nucleus (LMN) and from neurons in the dorsal octavolateral nucleus (DON) of the paddlefish. We stimulated with sine wave stimuli of different amplitudes and frequencies and with moving DC stimuli. During sinusoidal stimulation, DON units increased their firing rate during the negative cycle of the sine wave and decreased their firing rate to the positive cycle. Lateral mesencephalic nucleus units increased their rate for both half cycles of the sine wave. Lateral mesencephalic nucleus units are more sensitive than DON units, especially to small moving dipoles. Dorsal octavolateral nucleus units respond to a moving DC dipole with an increase followed by a decrease in spike rate or vice versa, depending on movement direction and dipole orientation. Lateral mesencephalic nucleus units, in contrast, increased their discharge rate for all stimuli. Any change in discharge rate of DON units is converted in the LMN to a discharge rate increase. Lateral mesencephalic nucleus units therefore appear to code the presence of a stimulus regardless of orientation and motion direction.

Resonant properties in the paddlefish electrosensory system caused by delayed feedback.

INTRODUCTION: The paddlefish electrosensory system consists of receptor cells in the skin that sense minute electric fields from their prey, small water fleas. The receptors thereby measure the difference of the voltage at the skin surface against the voltage inside the animal. Due to a high skin impedance, this internal voltage is considered to be relatively fixed. RESULTS: We found, however, that this internal voltage can fluctuate. It shows damped oscillations to a single short electric field pulse and changes, with some time delay, according to the previous history of stimulation, and shows resonance at a certain frequency. CONCLUSIONS: Computer simulations show that these phenomena can be explained by the presence of delayed feedback where the internal voltage is part of the feedback loop.

Kármán vortex street detection by the lateral line.

Fish use the lateral line system for prey detection, predator avoidance, schooling behavior, intraspecific communication and spatial orientation. In addition the lateral line may be important for station holding and for the detection of the hydrodynamic trails (vortex streets) generated by swimming fish. We investigated the responses of anterior lateral line nerve fibers of goldfish, Carassius auratus, to unidirectional water flow (10 cm s(-1)) and to running water that contained a Kármán vortex street. Compared to still water conditions, both unidirectional water flow and Kármán vortex streets caused a similar increase in the discharge rate of anterior lateral line nerve fibers. If exposed to a Kármán vortex street, the amplitude of spike train frequency spectra increased at the vortex shedding frequency. This increase was especially pronounced if the fish intercepted the edge of a Kármán vortex street. Our data show that the vortex shedding frequency can be retrieved from the responses of anterior lateral line nerve fibers.

Responses to dipole stimuli of anterior lateral line nerve fibres in goldfish, Carassius auratus, under still and running water conditions.

We investigated how fibres in the anterior lateral line nerve of goldfish, Carassius auratus, respond to sinusoidal water motions in a background of still or running water. Two types of fibres were distinguished: type I fibres, which most likely innervate superficial neuromasts, were stimulated by running water (10 cm s(-1)) while type II fibres, which most likely innervate canal neuromasts, were not stimulated by running water. The responses of type I fibres to sinusoidal water motions were masked in running water whereas responses of type II fibres were not masked. These findings are in agreement with previous data obtained from the posterior lateral line nerve of goldfish. Furthermore, we demonstrate here that for type I fibres the degree of response masking increased with increasing flow velocity. Finally, the ratio between responses that were masked in running water (type I) and those that were not masked (type II) increases with increasing flow velocity. Flow fluctuations that were generated by a cylinder in front of the fish did not affect ongoing activity in the flow, nor the dipole-evoked responses. The findings are discussed with respect to particle image velocimetry data of the water motions generated in the experiments.