The incomparable fascination of comparative physiology: 40 years with animals in the field and laboratory.

This paper is not meant to be a review article. Instead, it gives an overview of the major research projects that the author, together with his students, colleagues and collaborators, has worked on. Although the main focus of the author's work has always been the fish lateral line, this paper is mainly about all the other research projects he did or that were done in his laboratory. These include studies on fishing spiders, weakly electric fish, seals, water rats, bottom dwelling sharks, freshwater rays, venomous snakes, birds of prey, fire loving beetles and backswimmers. The reasons for this diversity of research projects? Simple. The authors's lifelong enthusiasm for animals, and nature's ingenuity in inventing new biological solutions. Indeed, this most certainly was a principal reason why Karl von Frisch and Alfred Kühn founded the Zeitschrift für vergleichende Physiologie (now Journal of Comparative Physiology A) 100 years ago.

Hydrodynamic reception in the Australian water rat, Hydromys chrysogaster.

The Australian water rat, Hydromys chrysogaster, preys on a wide variety of aquatic and semiaquatic arthropods and vertebrates, including fish. A frequently observed predatory strategy of Hydromys is sitting in wait at the water's edge with parts of its vibrissae submersed. Here we show that Hydromys can detect water motions with its whiskers. Behavioural thresholds range from 1.0 to 9.4 mm s-1 water velocity, based on maximal horizontal water velocity in the area covered by the whiskers. This high sensitivity to water motions would enable Hydromys to detect fishes passing by. No responses to surface waves generated by a vibrating rod and resembling the surface waves caused by struggling insects were found.

Responses of medullary lateral line units of the rudd, Scardinius erythrophthalmus, and the nase, Chondrostoma nasus, to vortex streets.

Fish use their mechanosensory lateral line amongst others for the detection of vortices shed by an upstream object and/or for the detection of vortices caused by the tail fin movements of another fish. Thus, vortices are one type of hydrodynamic stimuli to which fish are exposed in their natural environment. We investigated the responses of medullary lateral line units of common rudd, Scardinius erythrophthalmus, and common nase, Chondrostoma nasus (Cyprinidae), to water flow (9.5-13.3 cm-1) that contained vortices (a Kármán vortex street) shed by an upstream cylinder (diameter 2 cm). The distance between the cylinder and the tip of the fish's snout varied between 8 and 24 cm. 21 out of 42 units (S. erythrophthalmus), respectively, 9 out of 39 units (Chondrostoma nasus) responded to the vortices shed by the cylinder. Up to a cylinder distance of 24 cm, interburst intervals revealed the vortex shedding frequency, i.e., burst frequency was similar to or identical with the vortex shedding frequency.

The peregrine falcon's rapid dive: on the adaptedness of the arm skeleton and shoulder girdle.

During a dive, peregrine falcons (Falco peregrinus) can reach a velocity of up to 320 km h- 1. Our computational fluid dynamics simulations show that the forces that pull on the wings of a diving peregrine can reach up to three times the falcon's body mass at a stoop velocity of 80 m s- 1 (288 km h- 1). Since the bones of the wings and the shoulder girdle of a diving peregrine falcon experience large mechanical forces, we investigated these bones. For comparison, we also investigated the corresponding bones in European kestrels (Falco tinnunculus), sparrow hawks (Accipiter nisus) and pigeons (Columba livia domestica). The normalized bone mass of the entire arm skeleton and the shoulder girdle (coracoid, scapula, furcula) was significantly higher in F. peregrinus than in the other three species investigated. The midshaft cross section of the humerus of F. peregrinus had the highest second moment of area. The mineral densities of the humerus, radius, ulna, and sternum were highest in F. peregrinus, indicating again a larger overall stability of these bones. Furthermore, the bones of the arm and shoulder girdle were strongest in peregrine falcons.

The muscle activity of trout exposed to unsteady flow.

In running water trout seek out special regions for station holding. Trout exposed to flow fluctuations caused by a cylinder hold station immediately upstream of the cylinder (bow wake region), adjacent to the cylinder (entraining region) or downstream of the cylinder (Kármán gait). In addition it was shown that the activity of the axial red swimming muscles is reduced during Kármán gaiting. Up to now only the two-dimensional (horizontal) extensions of the above regions have been examined. We determined both, the horizontal and vertical extension of the Kármán gait, entraining and bow wake region by continuously recording the position (spatial resolution 1 cm3) of trout for 3 h. In addition we continuously recorded the trunk muscle activity. The Kármán gate region had the smallest vertical extension (13 cm, water level 28-29 cm, length of the submerged cylinder 27 cm), followed by the entraining (21 cm) and bow wake region (25 cm). A fourth so far unknown region used for station holding was immediately below a stationary surface wave which, at flow velocities ≥36 cm s- 1, developed slightly downstream of the cylinder. While in any of the above regions the activity of the axial swimming muscles was significantly reduced.

Responses of infrared-sensitive tectal units of the pit viper Crotalus atrox to moving objects.

Rattlesnakes perceive IR radiation with their pit organs. This enables them to detect and strike towards warm-blooded prey even in the dark. In addition, the IR sense allows rattlesnakes to find places for thermoregulation. Animate objects (e.g., prey) tend to move and thus cause moving IR images across the pit membrane. Even when an object is stationary, scanning head movements of rattlesnakes will result in moving IR images across the pit membrane. We recorded the neuronal activity of IR-sensitive tectal neurons of the rattlesnake Crotalus atrox while stimulating the snakes with an IR source that moved horizontally at various velocities. As long as object velocity was low (angular velocity of ~5°/s) IR-sensitive tectal neurons hardly showed any responses. With increasing object velocity though, neuronal activity reached a maximum at ~50°/s. A further increase in object velocity up to ~120°/s resulted in a slight decrease of neuronal activity. Our results demonstrate the importance of moving stimuli for the snake's IR detection abilities: in contrast to fast moving objects, stationary or slowly moving objects will not be detected when the snake is motionless, but might be detected by scanning head movements.

Smart Mechanical Dipole: a device for the measurement of sphere motion in behavioral and neurophysiological experiments.

Fluid motion and pressure fields induced by vibrating spheres are frequently used to investigate the function of biological mechanosensory systems and artificial sensors. The calibration of the sphere motion amplitude (displacement, velocity, acceleration), time course and vibration direction often demands expensive equipment. To mitigate this requirement, we have developed a high-quality, low-cost device that we term a 'Smart Mechanical Dipole'. It provides real-time measurement of sphere acceleration along three axes and can be used to obtain an accurate stimulation trace. We applied digital filtering to equalize the frequency response of the vibrating sphere, which also reduced unwanted amplitude and frequency changes in the hydrodynamic signal. In addition, we show that the angular orientation of the rod to which the sphere was attached, i.e. axial versus transverse, but not the immersion depth of the sphere affected sphere vibration behavior.

Medullary lateral line units of rudd, Scardinius erythrophthalmus, are sensitive to Kármán vortex streets.

We investigated the responses of medullary lateral line units of the rudd, Scardinius erythrophthalmus, to bulk water flow (7 cm s(-1)) and to water flow that contained vortices shed by an upstream half cylinder (diameter 1, 2, and 3 cm). Thirty-five percent of the medullary units either increased or decreased their discharge rate with the increasing cylinder diameter. In some units, the spike patterns revealed the vortex shedding frequency, i.e., in these units the amplitude of spike train frequency spectra was similar or identical to the vortex shedding frequency.

The shark Chiloscyllium griseum can orient using turn responses before and after partial telencephalon ablation.

This study assessed spatial memory and orientation strategies in Chiloscyllium griseum. In the presence of visual landmarks, six sharks were trained in a fixed turn response. Group 1 started from two possible compartments approaching two goal locations, while group 2 started from and approached only one location, respectively. The learning criterion was reached within 9 ± 5.29 (group 1) and 8.3 ± 3.51 sessions (group 2). Transfer tests revealed that sharks had applied a direction strategy, possibly in combination with some form of place learning. Without visual cues, sharks relied solely on the former. To identify the underlying neural substrate(s), telencephalic were lesioned and performance compared before and after surgery. Ablation of the dorsal and medial pallia only had an effect on one shark (group 1), indicating that the acquisition and retention of previously gained knowledge were unaffected in the remaining four individuals. Nonetheless, the shark re-learned the task. In summary, C. griseum can utilize fixed turn responses to navigate to a goal; there is also some evidence for the use of external visual landmarks while orienting. Probably, strategies can be used alone or in combination. Neither the dorsal nor medial pallium seems to be responsible for the acquisition and processing of egocentric information.

Place learning prior to and after telencephalon ablation in bamboo and coral cat sharks (Chiloscyllium griseum and Atelomycterus marmoratus).

This study assessed complex spatial learning and memory in two species of shark, the grey bamboo shark (Chiloscyllium griseum) and the coral cat shark (Atelomycterus marmoratus). It was hypothesized that sharks can learn and apply an allocentric orientation strategy. Eight out of ten sharks successfully completed the initial training phase (by locating a fixed goal position in a diamond maze from two possible start points) within 14.9 ± 7.6 sessions and proceeded to seven sets of transfer tests, in which sharks had to perform under altered environmental conditions. Transfer tests revealed that sharks had oriented and solved the tasks visually, using all of the provided environmental cues. Unintentional cueing did not occur. Results correspond to earlier studies on spatial memory and cognitive mapping in other vertebrates. Future experiments should investigate whether sharks possess a cognitive spatial mapping system as has already been found in several teleosts and stingrays. Following the completion of transfer tests, sharks were subjected to ablation of most of the pallium, which compromised their previously acquired place learning abilities. These results indicate that the telencephalon plays a crucial role in the processing of information on place learning and allocentric orientation strategies.

Interaction of barn owl leading edge serrations with freestream turbulence.

The silent flight of barn owls is associated with wing and feather specialisations. Three special features are known: a serrated leading edge that is formed by free-standing barb tips which appears as a comb-like structure, a soft dorsal surface, and a fringed trailing edge. We used a model of the leading edge comb with 3D-curved serrations that was designed based on 3D micro-scans of rows of barbs from selected barn-owl feathers. The interaction of the flow with the serrations was measured with Particle-Image-Velocimetry in a flow channel at uniform steady inflow and was compared to the situation of inflow with freestream turbulence, generated from the turbulent wake of a cylinder placed upstream. In steady uniform flow, the serrations caused regular velocity streaks and a flow turning effect. When vortices of different size impacted the serrations, the serrations reduced the flow fluctuations downstream in each case, exemplified by a decreased root-mean-square value of the fluctuations in the wake of the serrations. This attenuation effect was stronger for the spanwise velocity component, leading to an overall flow homogenization. Our findings suggest that the serrations of the barn owl provide a passive flow control leading to reduced leading-edge noise when flying in turbulent environments.

Avoidance conditioning in bamboo sharks (Chiloscyllium griseum and C. punctatum): behavioral and neuroanatomical aspects.

Animals face different threats; to survive, they have to anticipate how to react or how to avoid these. It has already been shown in teleosts that selected regions in the telencephalon, i.e., the medial pallium, are involved in avoidance learning strategies. No such study exists for any chondrichthyan. In nature, an avoidance reaction may vary, ranging from a 'freeze' reaction to a startling response and quick escape. This study investigated whether elasmobranchs (Chiloscylliumgriseum and C. punctatum) can be conditioned in an aversive classical conditioning paradigm. Upon successful conditioning, the dorsal, medial and lateral pallium were removed (group 1) and performance tested again. In a second group, the same operation was performed prior to training. While conditioning was successful in individuals of both groups, no escape responses were observed. Post-operative performance was assessed and compared between individual and groups to reveal if the neural substrates governing avoidance behavior or tasks learned in a classical conditioning paradigm are located within the telencephalon, as has been shown for teleosts such as goldfish.

The functional significance of lateral line canal morphology on the trunk of the marine teleost Xiphister atropurpureus (Stichaeidae).

We investigated the filter properties of the highly branched trunk lateral lines of the stichaeid Xiphister atropurpureus and compared them to the filter properties of simple lateral line canals. For this purpose artificial canals were constructed, some of which were fitted with artificial neuromasts. In still water, the response of a simple canal versus two types of Xiphister-like canals to a vibrating sphere stimulus were similar, as was the decrease in the responses as a function of sphere distance. Also comparable was the mechanical coupling between neighboring parts of the main canal. However, compared to the simple canal, the Xiphister-like canals showed a lower spatial resolution. Equipping artificial lateral line canals with artificial neuromasts revealed that Xiphister-like canals, i.e., lateral lines canals with tubuli that contained widely spaced pores, improve the signal-to-noise ratio in a highly turbulent environment. Even though a reduced spatial resolution is the price for this improvement, Xiphister may compensate for this compromise by having four instead of the usual single trunk lateral line canal. We suggest that lateral line canals with tubuli that contain widely spaced pores and multiple lateral line canals on each body side are an adaptation to a highly turbulent aquatic environment.

Potential targets aimed at by spitting cobras when deterring predators from attacking.

When threatened, spitting cobras eject venom towards the face of an aggressor. To uncover the relevant cues used by cobras for face recognition we determined how often artificial targets equipped with or without eyes elicited spitting behavior. In addition, we measured whether and how target shape and size influenced the spitting behavior of cobras. Results show that oval- and round-shaped targets were most effective, while triangles with the same surface area as oval 'face like' targets hardly elicited spitting. The likelihood of spitting depended on neither the presence, the spatial arrangement (horizontal or vertical) nor the surface texture (shiny or matt) of glass eyes. Most likely, cobras do not specifically aim at the eyes of an offender but at the center of the body part closest to them. As this is usually the face of an animal, this strategy will result in at least one eye of the offender being hit most of the time.

Toral lateral line units of goldfish, Carassius auratus, are sensitive to the position and vibration direction of a vibrating sphere.

We recorded the responses of lateral line units in the midbrain torus semicircularis of goldfish, Carassius auratus, to a 50-Hz vibrating sphere and determined the unit's spatial receptive fields for various distances between fish and sphere and for different directions of sphere vibration. All but one unit responded to the vibrating sphere with an increase in discharge rate. Only a proportion (25%) of the units exhibited phase-locked responses. Receptive fields were narrow or broad and contained one, two or more areas of increased discharge rate. The data show that the receptive fields of toral lateral line units are in many respects similar to those of brainstem units but differ from those of afferent nerve fibres. The responses of primary afferents represent the pressure gradient pattern generated by a vibrating sphere and provide information about sphere location and vibration direction. Across the array of lateral line neuromasts, the fish brain in principle can derive this information. Nevertheless, toral units tuned to a distinct sphere location or sensitive to a distinct sphere vibration direction were not found. Therefore, it is conceivable that the torus semicircularis uses a population code to determine spatial location and vibration direction of a vibrating sphere.

Coping with flow: behavior, neurophysiology and modeling of the fish lateral line system.

With the mechanosensory lateral line fish perceive water motions relative to their body surface and local pressure gradients. The lateral line plays an important role in many fish behaviors including the detection and localization of dipole sources and the tracking of prey fish. The sensory units of the lateral line are the neuromasts which are distributed across the surface of the animal. Water motions are received and transduced into neuronal signals by the neuromasts. These signals are conveyed by afferent nerve fibers to the fish brain and processed by lateral line neurons in parts of the brainstem, cerebellum, midbrain, and forebrain. In the cerebellum, midbrain, and forebrain, lateral line information is integrated with sensory information from other modalities. The present review introduces the peripheral morphology of the lateral line, and describes our understanding of lateral line physiology and behavior. It focuses on recent studies that have investigated: how fish behave in unsteady flow; what kind of sensory information is provided by flow; and how fish use and process this information. Finally, it reports new theoretical and biomimetic approaches to understand lateral line function.

Two-dimensional receptive fields of midbrain lateral line units in the goldfish, Carassius auratus.

We determined the receptive fields of midbrain lateral line units in goldfish, Carassius auratus, with a 50 Hz vibrating sphere placed at various azimuths and elevations alongside the fish and studied how responses were affected by different directions of sphere vibration. The receptive fields of toral lateral line units, in contrast to those of primary afferent nerve fibers, did not represent the pressure gradient pattern generated by a vibrating sphere. Thus, unlike primary afferents, single toral lateral line units did not code for source location in their spatial discharge patterns. The two-dimensional receptive fields were round, horizontally or vertically stretched, or complex. While some toral lateral line units were sensitive to the direction of sphere vibration others were not.

Responses of brainstem lateral line units to different stimulus source locations and vibration directions.

We recorded responses of lateral line units in the medial octavolateralis nucleus in the brainstem of goldfish, Carassius auratus, to a 50 Hz vibrating sphere and studied how responses were affected by placing the sphere at various locations alongside the fish and by different directions of vibration. In most units (88%), stimulation with the sphere from one or more spatial locations caused an increase and/or decrease in discharge rate. In few units (10%), discharge rate was increased by stimulation from one location and decreased by stimulation from an adjacent location in space. In a minority of the units (2%), changing sphere location did not affect discharge rates but caused a change in phase coupling. Units sensitive to a distinct sphere vibration direction were not found. The data also show that the responses of most brainstem units differ from those of primary afferent nerve fibers. Whereas primary afferents represent the pressure gradient pattern generated by the sphere and thus encode location and vibration direction of a vibrating sphere, most brainstem units do not. This information may be represented in the brainstem by a population code or in higher centers of the ascending lateral line pathway.

Target tracking during venom 'spitting' by cobras.

Spitting cobras, which defend themselves by streaming venom towards the face and/or eyes of a predator, must be highly accurate because the venom they spit is only an effective deterrent if it lands on the predator's cornea. Several factors make this level of accuracy difficult to achieve; the target is moving, is frequently >1 m away from the snake and the venom stream is released in approximately 50 ms. In the present study we show that spitting cobras can accurately track the movements of a potentially threatening vertebrate, and by anticipating its subsequent (short-term) movements direct their venom to maximize the likelihood of striking the target's eye. Unlike other animals that project material, in spitting cobras the discharge orifice (the fang) is relatively fixed so directing the venom stream requires rapid movements of the entire head. The cobra's ability to track and anticipate the target's movement, and to perform rapid cephalic oscillations that coordinate with the target's movements suggest a level of neural processing that has not been attributed to snakes, or other reptiles, previously.

Entraining in trout: a behavioural and hydrodynamic analysis.

Rheophilic fish commonly experience unsteady flows and hydrodynamic perturbations. Instead of avoiding turbulent zones though, rheophilic fish often seek out these zones for station holding. A behaviour associated with station holding in running water is called entraining. We investigated the entraining behaviour of rainbow trout swimming in the wake of a D-shaped cylinder or sideways of a semi-infinite flat plate displaying a rounded leading edge. Entraining trout moved into specific positions close to and sideways of the submerged objects, where they often maintained their position without corrective body and/or fin motions. To identify the hydrodynamic mechanism of entraining, the flow characteristics around an artificial trout placed at the position preferred by entraining trout were analysed. Numerical simulations of the 3-D unsteady flow field were performed to obtain the unsteady pressure forces. Our results suggest that entraining trout minimise their energy expenditure during station holding by tilting their body into the mean flow direction at an angle, where the resulting lift force and wake suction force cancel out the drag. Small motions of the caudal and/or pectoral fins provide an efficient way to correct the angle, such that an equilibrium is even reached in case of unsteadiness imposed by the wake of an object.