Ernst-August Seyfarth

Idiothetic orientation of a wandering spider: Compensation of detours and estimates of goal distance

SummaryThe wandering spider Cupiennius salei Keys uses idiothetic orientation, i.e., memorized information about its own previous movements, to retrieve lost prey. Spiders, having been chased away from a prey fly, return to the capture site (the goal) over a distance of more than 75 cm even though all external orientation cues were precluded. This behavior and its sensory basis were examined by varying the proprioceptive and ‘motor command’ inputs to the memory and by ablating particular lyriform slit sense organs on the legs of the spider.The success rate of returns to the goal after rectilinear chases over 6 discrete distances ranging from 20 cm to>41 cm declines with increasing distances. At distances>41 cm, more than 50% of the performances of intact spiders are nevertheless ‘successful’, in that the animals approach the capture site as close as 5 cm (or less).Animals that have been operated on (lyriform organs on all femora destroyed) are much less successful even at short distances. The mean starting angles of the returns by intact spiders and by those operated on do not differ signficantly. ‘Walking error en’ for each segment of the entire return path shows that intact animals deviate little from the ideal return route and correctly estimate the distance to the goal. The operated spiders tend to drift off the ideal return route, while their distance estimates remain largely accurate.Returns after curvilinear chases through a semicircular corridor do not retrace the curved path; instead the spiders take a shortcut. Of all performances by intact and by control spiders (with sham operations) 85% are successful. By contrast, most of the 8 groups with sensory ablations have a success rate of less than 50%.Compensation for the semicircular detours is not quite complete: the mean starting directions of returns are biased, pointing to the corridor, and the shape of many return paths reflects the curved corridor shape. Spiders with unilateral ablations of their femoral lyriform organs show low success rates only if the operated legs are on the inner curve perimeter during the chase, while their return parameters resemble those of the intact group in the reverse situation (operated legs on outer perimeter). These side-specific ablation effects, which are correlated with the geometrical situation existing while idiothetic information is gathered and memorized, suggest that the idiothetic memory depends at least partly on input from proprioceptors.

Spiders of the genus Cupiennius Simon 1891 (Araneae, Ctenidae)

SummaryCupiennius is a genus of hunting spiders with seven established species. One of these (C. salei) has been used in laboratory research for many years. Here we report on the geographic distribution of the genus and some characteristics of its habitat. (1) The genus is Central American. Its range is from the state of Veracruz in Mexico in the north to Panama in the south. Five of the seven species are known to occur in the Canal Area, Panama. Sympatry is best documented for C. getazi and C. coccineus and is likely to occur in other species. (2) All known species of Cupiennius are closely associated with particular plants on which they hide during the day and prey, court, and moult at night. The most typical dwelling plant such as a bromeliad or a banana plant is a monocotyledon with mechanically strong and unbranched leaves that provide retreats at their bases. On plants not providing “ready-made” shelters, such as ginger or members of the Araceae, several species of Cupiennius have been observed to build retreats. (3) Average monthly rainfall and temperature data are given for six locations where we have recently observed C. coccineus, C. getazi, C. panamensis, and C. salei. According to measurements taken in the field the microclimate within a typical retreat differs considerably from the external environment: during the day the retreat space shows lower aver-age water evaporation rates and higher relative air humidity.

From stress and strain to spikes: mechanotransduction in spider slit sensilla

Abstract. This review focuses on the structure and function of a single mechanoreceptor organ in the cuticle of spiders. Knowledge emerging from the study of this organ promises to yield general principles that can be applied to mechanosensation in a wide range of animal systems. The lyriform slit sense organ on the antero-lateral leg patella of the spider Cupiennius salei is unusual in possessing large sensory neurons, whose cell bodies are close to the sites of sensory transduction, and accessible to intracellular recording during mechanotransduction. This situation, combined with recent technical developments, has made it possible to observe and experiment with all the major stages of mechanosensation. Important findings include the approximate size, number and ionic selectivity of the ion channels responsible for mechanotransduction, the types of voltage-activated ion channels responsible for action potential encoding, and the mechanisms controlling the dynamic properties of transduction and encoding. Most recently, a complex efferent system for peripheral modulation of mechanosensation has been discovered and partially characterized. Much remains to be learned about mechanosensation, but the lyriform slit sense organ system continues to offer important opportunities to advance our understanding of this crucial sense.

Peripheral GABAergic inhibition of spider mechanosensory afferents

Spider mechanosensory neurons receive an extensive network of efferent synapses onto their sensory dendrites, somata and distal axonal regions. The function of these synapses is unknown. Peripheral synapses are also found on crustacean stretch‐receptor neurons but not on mechanosensory afferents of other species, although inhibitory GABAergic synapses are a common feature of centrally located axon terminals. Here we investigated the effects of GABA receptor agonists and antagonists on one group of spider mechanosensory neurons, the slit sense organ VS‐3, which are accessible to current‐ and voltage‐clamp recordings. Bath application of GABA activated an inward current that depolarized the membrane and increased the membrane conductance leading to impulse inhibition. VS‐3 neuron GABA receptors were activated by muscimol and inhibited by picrotoxin but not bicuculline, and their dose–response relationship had an EC50 of 103.4 µm, features typical for insect ionotropic GABA receptors. Voltage‐ and current‐clamp analysis confirmed that, while the Na+ channel inhibition resulting from depolarization can lead to impulse inhibition, the increase in membrane conductance (i.e. ‘shunting’) completely inhibited impulse propagation. This result argues against previous findings from other preparations that GABA‐mediated inhibition is caused by a depolarization that inactivates Na+ conductance, and it supports those findings that assign this role to membrane shunting. Our results show that GABA can rapidly and selectively inhibit specific mechanoreceptors in the periphery. This type of peripheral inhibition may provide spiders with a mechanism for distinguishing between signals from potential prey, predators or mates, and responding with appropriate behaviour to each signal.

PERIPHERAL SYNAPSES AT IDENTIFIABLE MECHANOSENSORY NEURONS IN THE SPIDER CUPIENNIUS SALEI : SYNAPSIN-LIKE IMMUNOREACTIVITY

Abstract Indirect immunocytochemical tests were used at the light- and electron-microscopic levels to investigate peripheral chemical synapses in identified sensory neurons of two types of cuticular mechanosensors in the spider Cupiennius salei Keys.: (1) in the lyriform slit-sense organ VS-3 (comprising 7–8 cuticular slits, each innervated by 2 bipolar sensory neurons) and (2) in tactile hair sensilla (each supplied with 3 bipolar sensory cells). All these neurons are mechanosensitive. Application of a monoclonal antibody against Drosophila synapsin revealed clear punctate immunofluorescence in whole-mount preparations of both mechanoreceptor types. The size and overall distribution of immunoreactive puncta suggested that these were labeled presynaptic sites. Immunofluorescent puncta were 0.5–6.8 μm long and located 0.5–6.6 μm apart from each other. They were concentrated at the initial axon segments of the sensory neurons, while the somata and the dendritic regions showed fewer puncta. Western blot analysis with the same synapsin antibody against samples of spider sensory hypodermis and against samples from the central nervous system revealed a characteristic doublet band at 72 kDa and 75 kDa, corresponding to the apparent molecular mass of synapsin in Drosophila and in mammals. Conventional transmissionelectron-microscopic staining demonstrated that numerous chemical synapses (with at least 2 vesicle types) were present at these mechanosensory neurons and their surrounding glial sheath. The distribution of these synapses corresponded to our immunofluorescence results.Ultrastructural examination of anti-synapsin-stained neurons confirmed that reaction product was associated with synaptic vesicles. We assume that the peripheral synaptic contacts originate from efferents that could exert a complex modulatory influence on mechanosensory activity.

Spider Proprioception: Receptors, Reflexes, and Control of Locomotion

Historically, the question of how animals control their own movements has been a major concern of behavioral physiology. Much of the early and recent work has focused upon the recurring debate about central versus peripheral control of repetitive behavior such as locomotion. While many rhythmic behaviors such as walking, flying, or breathing can be maintained in the absence of patterned sensory input, there is now apparent consent among neurobiologists that peripheral feedback by way of sensory receptors (such as proprioceptors) is essential for fine control of certain behavior patterns and for reacting to the environment as the situation may demand (Delcomyn 1980; Selverston 1980).

Lyriform slit sense organs and muscle reflexes in the spider leg

Summary1.Muscle reflexes were studied in spider legs (Cupiennius salei Keys.) by electromyographic recordings during imposed sinusoidal movements of the patella/tibia and the tibia/metatarsus joints (Fig. 1). Compound slit sense (“lyriform”) organs and hair sensilla near the joints are stimulated by the induced movements. Their possible involvement in leg reflexes was examined with ablation experiments.2.Induced deflections of either joint in its main plane of movement elicit resistance reflexes: passive promotion of the tibia activates the remotor tibiae muscle and passive remotion excites the promotor (Fig. 3). Imposed elevation of the metatarsus against the tibia activates both the flexor metatarsi longus and the fl. met. bilobatus (Fig. 7). Neither the destruction of lyriform organs nor the removal of hair sensilla at these joints causes changes in the strength and temporal pattern of muscle activity which exceed the variability of reflex discharges in intact animals (Figs. 5, 6, 9, 10).3.Smalllateral deflections of the metatarsus activate patellar muscles (i.e., extrinsic to the stimulated joint), which act in synergy with the imposed movements so as to augment them: forward deflection excites the promotor tibiae, backward movement the remotor. The destruction of particular lyriform organs on the tibia, stimulated by the induced movements (Fig. 12), causes complete failure of these extrinsic, synergic leg reflexes (Figs. 13, 14).4.Both these resistance and synergic reflex responses are restricted to muscles of the leg which is stimulated.

Tactile hairs and the adjustment of body height in wandering spiders: Behavior, leg reflexes, and afferent projections in the leg ganglia

SummaryIn spiders, stimulation of cuticular tactile hairs on the ventral aspects of proximal leg parts elicits reflex activity in certain leg muscles; contraction of these muscles raises the body. InCupiennius salei, using video film analysis, electromyography, and sensory recordings we studied the body-raising behavior and associated leg reflexes (i) in freely walking spiders, (ii) in animals tethered above a treadmill device, and (iii) in completely immobilized preparations.1.Freely walking spiders abruptly (within 160 ms) raise their bodies by extending the legs when they touch a ventral flexible wire obstacle. The same animals collide with the obstacle when all tactile hairs on the sternum and on the ventral sides of all proximal leg segments have been removed (Fig. 1).2.Body raising is also readily induced in tetheredCupiennius that are stimulated ventrally while standing on an air-suspended styrofoam sphere. Deflection of just one ventral hair can induce coordinated extension of all 8 legs (Fig. 2). Electromyograms reveal transient activity in several muscles that occurs almost simultaneously in all walking legs (Fig. 3).3.Unlike the coordinated response of all legs in spiders tested on the treadmill, tactile reflex activity in spiders immobilized on their backs is confined to muscles of the particular leg being stimulated (‘local reflexes’). Again, deflection of a single tactile hair elicits activity in muscles such as the promotor/adductor of the leg coxa (muscle c2; Figs. 4, 7), which is involved in body raising.4.The 3 bipolar, mechanosensory neurons innervating each tactile hair differ both in their rates of adaptation to maintained, step-like deflections of their hair shafts (Fig. 8) and in their frequency response to single and to repeated sinusoidal stimulus cycles. Tests with trains of sinusoidal hair deflections demonstrate that activity of one sensory-hair unit alone (unit ‘3’ in Fig. 8) is sufficient to drive the local tactile reflex in muscle c2.5.Anterograde cobalt fillings of afferents from tactile hairs on the antero-ventral coxa (nerve ‘aCo’) into the fused subesophageal leg ganglia reveals a ventral group of fibers with ipsilateral, intrasegmental endings and a dorsal group with ipsilateral, plurisegmental arborizations in central neuropil (Fig. 9). A companion paper (Milde and Seyfarth 1988) identifies central neuronal correlates of the reflex activity described here.

Electrical and mechanical stimulation of a spider slit sensillum: Outward current excites

Summary1.When current is passed between a surface electrode on the lyriform organ of the spider and a reference electrode in the hemolymph, the spike rates of the slit-sensillum sensory cells are modulated. Outward current (surface electrode negative) excites and inward current abolishes spontaneous activity (Fig. 2). This electrical response is the opposite of that reported in other arthropod mechano- and chemoreceptors. It is, however, compatible with a distal site of the spike-initiating region, possibly near the dendrite tip. Such an arrangement in the spider parallels the finding of Rick et al. (1976) that the lymph space surrounding the apical dendrite appears (unlike the situation in the insects examined) to have a high concentration of Na+.2.Spikes recorded at the surface of this mechanoreceptor during compression of the slit do not differ appreciably in shape from those elicited by outward current (Fig. 3). Both have a negative leading edge; again, the polarity is the opposite of that measured in most insect epithelial-receptor spikes.3.Responses to electrical and mechanical stimuli can be superimposed (Fig. 4), so that electrical stimuli can be used in behavioral experiments to modulate the response to mechanical input.4.The spike rate elicited by maintained steps of outward current does not decline (Fig. 5). Hence the rapid adaptation to mechanical stimuli is not a property of the spike-initiating process that is driven by imposed current. On the other hand, responses to electrical test stimuli do sample some slowly recovering aftereffect of a period of adaptation to a mechanical stimulus (Fig. 6).5.Although the distributions of capacitance and resistance near these sensilla are unknown, we discuss trial explanations of the negative spikes measured, by qualitative comparison with the volume conductor analyses of Lorente de Nó.

Tactile hairs and leg reflexes in wandering spiders: physiological and anatomical correlates of reflex activity in the leg ganglia

SummaryIntracellular recordings and Lucifer-yellow fillings were used in a wandering spider,Cupiennius salei Keys., to identify central neuronal correlates of local reflex activity in muscle c2, which inserts on the leg coxa. Here we describe related neuronal elements in the hindleg neuromere of the fused, subesophageal-ganglion complex:1.Projectionsof primary sensory axons excited by hair deflection are confined to ventral parts of the ipsilateral leg-neuromere (Fig. 1); their central terminals end near longitudinal, interganglionic tracts.2.Two identified excitatorymotor neurons for muscle c2 (which is a promotor/adductor of the coxa) are also confined to the ipsilateral (hindleg) ganglion. The dendritic branches and the efferent axonal segment extend in regions well dorsal to the sensory projections (Fig. 2); we found neither morphological nor electrophysiological evidence for direct synaptic contacts between hair afferents and motor neurons (Fig. 3).3.Various types of identifiedinterneurons give responses correlated with the reflex. We classified them, by anatomical criteria, aslocal interneurons confined to the ipsilateral hindleg neuromere (Figs. 4, 5) and asplurisegmental interneurons arborizing in more than one neuromere (Figs. 6, 7, 8). Although detailed electrophysiological tests of functional connections are not available for all these elements, we discuss how the various interneurons identified here may be involved in the local reflex response and in the coordinated, intersegmental reflex behavior that is observed when the unrestrained spider uses all 8 legs to raise its body (see the companion paper by Eckweiler and Seyfarth 1988).