Behavioural function and development of body-to-limb proportions and active movement ranges in three stick insect species.

Overall body proportions and relative limb length are highly characteristic for most insect taxa. In case of the legs, limb length has mostly been discussed with regard to parameters of locomotor performance and, in particular cases, as an adaptation to environmental factors or to the mating system. Here, we compare three species of stick and leaf insects (Phasmatodea) that differ strongly in the length ratio between antennae and walking legs, with the antennae of Medauroidea extradentata being much shorter than its legs, nearly equal length of antennae and legs in Carausius morosus, and considerably longer antennae than front legs in Aretaon asperrimus. We show that that relative limb length is directly related to the near-range exploration effort, with complementary function of the antennae and front legs irrespective of their length ratio. Assuming that these inter-species differences hold for both sexes and all developmental stages, we further explore how relative limb length differs between sexes and how it changes throughout postembryonic development. We show that the pattern of limb-to-body proportions is species-characteristic despite sexual dimorphism, and find that the change in sexual dimorphism is strongest during the last two moults. Finally, we show that antennal growth rate is consistently higher than that of front legs, but differs categorically between the species investigated. Whereas antennal growth rate is constant in Carausius, the antennae grow exponentially in Medauroidea and with a sudden boost during the last moult in Aretaon.

Active tactile exploration and tactually induced turning in tethered walking stick insects.

Many animals use their tactile sense for active exploration and tactually guided behaviors such as near-range orientation. In insects, tactile sensing is often intimately linked to locomotion, resulting in the orchestration of several concurrent active movements, including turning of the entire body, rotation of the head, and searching or sampling movements of the antennae. The present study aims at linking the sequence of tactile contact events to associated changes of all three kinds of these active movements (body, head and antennae). To do so, we chose the Indian stick insect Carausius morosus, an organism commonly used to study sensory control of locomotion. Methodologically, we combined recordings of walking speed, heading, whole-body kinematics and antennal contact sequences during stationary, tethered walking and controlled presentation of an 'artificial twig' for tactile exploration. Our results show that object presentation episodes as brief as 5 s are sufficient to allow for a systematic investigation of tactually induced turning behavior in walking stick insects. Animals began antennating the artificial twig within 0.5 s, and altered the beating fields of both antennae in a position-dependent manner. This change was mainly carried by a systematic shift of the head-scape joint movement and accompanied by associated changes in contact likelihood, contact location and sampling direction of the antennae. The turning tendency of the insect also depended on stimulus position, whereas the active, rhythmic head rotation remained unaffected by stimulus presentation. We conclude that the azimuth of contact location is a key parameter of active tactile exploration and tactually induced turning in stick insects.

Locomotor resilience through load-dependent modulation of muscle co-contraction.

Terrestrial locomotor behavior in variable environments requires resilience to sudden changes in substrate properties. For example, walking animals can adjust to substantial changes in slope and corresponding changes in load distribution among legs. In insects, slope-dependent adjustments have mainly been examined under steady-state conditions, whereas the transition dynamics have been largely neglected. In a previous study, we showed that steady-state adjustments of stick insects to ±45 deg slopes involve substantial changes in joint torques and muscle activity with only minor changes in leg kinematics. Here, we took a close look at the time course of these adjustments as stick insects compensate for various kinds of disturbances to load distribution. In particular, we tested whether the transition from one steady state to another involves distinct transition steps or follows a graded process. To resolve this, we combined simultaneous recordings of whole-body kinematics and hindleg muscle activity to elucidate how freely walking Carausius morosus negotiated a step-change in substrate slope. Step-by-step adjustments revealed that muscle activity changed in a graded manner as a function of body pitch relative to gravity. We further show analogous transient adjustment of muscle activity in response to destabilizing lift-off events of neighboring legs and the disappearance of antagonist co-activation during crawling episodes. Given these three examples of load-dependent regulation of antagonist muscle co-contraction, we conclude that stick insects respond to both transient and sustained changes in load distribution by regulating joint stiffness rather than through distinct transition steps.

When running is easier than walking: effects of experience and gait on human obstacle traversal in virtual reality.

Humans readily traverse obstacles irrespective of whether they walk or run, despite strong differences between these gaits. Assuming that the control of human obstacle traversal may be either gait-specific or gait-independent, the present study investigates whether previous experience in an obstacle traversal task transfers between the two gaits, and, if this was the case, whether transfer worked both ways. To this end, we conducted a within-group comparison of kinematic adjustments during human obstacle traversal in both walking and running, with distinct participant groups for the two gait sequences. Participants (n = 12/12 (f/m), avg. 25 yo) were motion captured as they traversed obstacles at walking and running speeds on a treadmill, surrounded by an immersive virtual reality (VR) environment. We find that kinematics recorded in our VR setup are consistent with that obtained in real-world experiments. Comparison of learning curves reveals that participants are able to utilize previous experience and transfer learned adjustments from one gait to another. However, this transfer is not symmetrical, with previous experience during running leading to increased success rate in walking, but not the other way round. From a range of step parameters we identified lacking toe height of the trailing leg as the main cause for this asymmetry.

Input of hair field afferents to a descending interneuron.

In insects the tactile sense is important for near-range orientation and is involved in various behaviors. Nocturnal insects, such as the stick insect Carausius morosus, continuously explore their surroundings by actively moving their antennae when walking. Upon antennal contact with objects, stick insects show a targeted front-leg movement. As this reaction occurs within 40 ms, descending transfer of information from the brain to the thorax needs to be fast. So far, a number of descending interneurons have been described that may be involved in this reach-to-grasp behavior. One of these is the contralateral ON-type velocity-sensitive neuron (cONv). cONv was found to encode antennal joint-angle velocity during passive movement. Here, we characterize the transient response properties of cONv, including its dependence on joint angle range and direction. As antennal hair field afferent terminals were shown to arborize close to cONv dendrites, we test whether antennal hair fields contribute to the joint-angle velocity encoding of cONv. To do so, we conducted bilateral extracellular recordings of both cONv interneurons per animal before and after hair field ablations. Our results show that cONv responses are highly transient, with velocity-dependent differences in delay and response magnitude. As yet, the steady state activity level was maintained until the stop of antennal movement, irrespective of movement velocity. Hair field ablation caused a moderate but significant reduction of movement-induced cONv firing rate by up to 40%. We conclude that antennal proprioceptive hair fields contribute to the velocity-tuning of cONv, though further antennal mechanoreceptors must be involved, too.NEW & NOTEWORTHY Active tactile exploration and tactually induced behaviors are important for many animals. They require descending information transfer about tactile sensor movement to thoracic networks. Here, we investigate response properties and afferent input to the identified descending interneuron cONv in stick insects. cONv may be involved in tactually induced reach-to-grasp movements. We show that cONv response delay, transient and steady state are velocity-dependent and that antennal proprioceptive hair fields contribute to the velocity encoding of cONv.

Evaluation of linear and non-linear activation dynamics models for insect muscle.

In computational modelling of sensory-motor control, the dynamics of muscle contraction is an important determinant of movement timing and joint stiffness. This is particularly so in animals with many slow muscles, as is the case in insects-many of which are important models for sensory-motor control. A muscle model is generally used to transform motoneuronal input into muscle force. Although standard models exist for vertebrate muscle innervated by many motoneurons, there is no agreement on a parametric model for single motoneuron stimulation of invertebrate muscle. Although several different models have been proposed, they have never been evaluated using a common experimental data set. We evaluate five models for isometric force production of a well-studied model system: the locust hind leg tibial extensor muscle. The response of this muscle to motoneuron spikes is best modelled as a non-linear low-pass system. Linear first-order models can approximate isometric force time courses well at high spike rates, but they cannot account for appropriate force time courses at low spike rates. A linear third-order model performs better, but only non-linear models can account for frequency-dependent change of decay time and force potentiation at intermediate stimulus frequencies. Some of the differences among published models are due to differences among experimental data sets. We developed a comprehensive toolbox for modelling muscle activation dynamics, and optimised model parameters using one data set. The "Hatze-Zakotnik model" that emphasizes an accurate single-twitch time course and uses frequency-dependent modulation of the twitch for force potentiation performs best for the slow motoneuron. Frequency-dependent modulation of a single twitch works less well for the fast motoneuron. The non-linear "Wilson" model that optimises parameters to all data set parts simultaneously performs better here. Our open-access toolbox provides powerful tools for researchers to fit appropriate models to a range of insect muscles.

Bimodal modulation of background activity in an identified descending interneuron.

In the absence of any obvious input, sensory neurons and interneurons can display resting or spontaneous activity. This is often regarded as noise and removed through trial averaging, although it may reflect history-dependent modulation of tuning or fidelity and, thus, be of functional relevance to downstream interneurons. We investigated the history dependence of spontaneous activity in a pair of identified, bimodal descending interneurons of the stick insect, called contralateral ON-type velocity-sensitive interneurons (cONv). The bilateral pair of cONv conveys antennal mechanosensory information to the thoracic ganglia, where it arborizes in regions containing locomotor networks. Each cONv encodes the movement velocity of the contralateral antenna, but also substrate vibration as induced by discrete tapping events. Moreover, cONv display highly fluctuating spontaneous activity that can reach rates similar to those during antennal movement at moderate velocities. Hence, cONv offer a unique opportunity to study history-dependent effects on spontaneous activity and, thus, encoding fidelity in two modalities. In this work, we studied unimodal and cross-modal effects as well as unilateral and bilateral effects, using bilateral recordings of both cONv neurons, while moving one antenna and/or delivering taps to induce substrate vibration. Tapping could reduce spontaneous activity of both neurons, whereas antennal movement reduced spontaneous activity of the contralateral cONv neuron only. Combination of both modalities showed a cooperative effect for some parameter constellations, suggesting bimodal enhancement. Since both stimulus modalities could cause a reduction of spontaneous activity at stimulus intensities occurring during natural locomotion, we conclude that this should enhance neuronal response fidelity during locomotion.NEW & NOTEWORTHY The spontaneous activity in a pair of identified, descending insect interneurons is reduced depending on stimulus history. At rest, spontaneous activity levels are correlated in both interneurons, indicating a common drive from background activity. Whereas taps on the substrate affect both interneurons, antennal movement affects the contralateral interneuron only. Cross-modal interaction occurs, too. Since spontaneous activity is reduced at stimulus intensities encountered during natural locomotion, the mechanism could enhance neuronal response fidelity during locomotion.

Motor control of an insect leg during level and incline walking.

During walking, the leg motor system must continually adjust to changes in mechanical conditions, such as the inclination of the ground. To understand the underlying control, it is important to know how changes in leg muscle activity relate to leg kinematics (movements) and leg dynamics (forces, torques). Here, we studied these parameters in hindlegs of stick insects (Carausius morosus) during level and uphill/downhill (±45 deg) walking, using a combination of electromyography, 3D motion capture and ground reaction force measurements. We find that some kinematic parameters including leg joint angles and body height vary across walking conditions. However, kinematics vary little compared with dynamics: horizontal leg forces and torques at the thorax-coxa joint (leg protraction/retraction) and femur-tibia joint (leg flexion/extension) tend to be stronger during uphill walking and are reversed in sign during downhill walking. At the thorax-coxa joint, the different mechanical demands are met by adjustments in the timing and magnitude of antagonistic muscle activity. Adjustments occur primarily in the first half of stance after the touch-down of the leg. When insects transition from level to incline walking, the characteristic adjustments in muscle activity occur with the first step of the leg on the incline, but not in anticipation. Together, these findings indicate that stick insects adjust leg muscle activity on a step-by-step basis so as to maintain a similar kinematic pattern under different mechanical demands. The underlying control might rely primarily on feedback from leg proprioceptors signaling leg position and movement.

A load-based mechanism for inter-leg coordination in insects.

Animals rely on an adaptive coordination of legs during walking. However, which specific mechanisms underlie coordination during natural locomotion remains largely unknown. One hypothesis is that legs can be coordinated mechanically based on a transfer of body load from one leg to another. To test this hypothesis, we simultaneously recorded leg kinematics, ground reaction forces and muscle activity in freely walking stick insects (Carausius morosus). Based on torque calculations, we show that load sensors (campaniform sensilla) at the proximal leg joints are well suited to encode the unloading of the leg in individual steps. The unloading coincides with a switch from stance to swing muscle activity, consistent with a load reflex promoting the stance-to-swing transition. Moreover, a mechanical simulation reveals that the unloading can be ascribed to the loading of a specific neighbouring leg, making it exploitable for inter-leg coordination. We propose that mechanically mediated load-based coordination is used across insects analogously to mammals.

A direct descending pathway informing locomotor networks about tactile sensor movement.

Much like visually impaired humans use a white-cane, nocturnal insects and mammals use antennae or whiskers for near-range orientation. Stick insects, for example, rely heavily on antennal tactile cues to find footholds and detect obstacles. Antennal contacts can even induce aimed reaching movements. Because tactile sensors are essentially one-dimensional, they must be moved to probe the surrounding space. Sensor movement is thus an essential cue for tactile sensing, which needs to be integrated by thoracic networks for generating appropriate adaptive leg movements. Based on single and double recordings, we describe a descending neural pathway comprising three identified ON- and OFF-type neurons that convey complementary, unambiguous, and short-latency information about antennal movement to thoracic networks in the stick insect. The neurons are sensitive to the velocity of antennal movements across the entire range covered by natural movements, regardless of movement direction and joint angle. Intriguingly, none of them originates from the brain. Instead, they descend from the gnathal ganglion and receive input from antennal mechanoreceptors in this lower region of the CNS. From there, they convey information about antennal movement to the thorax. One of the descending neurons, which is additionally sensitive to substrate vibration, feeds this information back to the brain via an ascending branch. We conclude that descending interneurons with complementary tuning characteristics, gains, input and output regions convey detailed information about antennal movement to thoracic networks. This pathway bypasses higher processing centers in the brain and thus constitutes a shortcut between tactile sensors on the head and the thorax.

Joint torques in a freely walking insect reveal distinct functions of leg joints in propulsion and posture control.

Determining the mechanical output of limb joints is critical for understanding the control of complex motor behaviours such as walking. In the case of insect walking, the neural infrastructure for single-joint control is well described. However, a detailed description of the motor output in form of time-varying joint torques is lacking. Here, we determine joint torques in the stick insect to identify leg joint function in the control of body height and propulsion. Torques were determined by measuring whole-body kinematics and ground reaction forces in freely walking animals. We demonstrate that despite strong differences in morphology and posture, stick insects show a functional division of joints similar to other insect model systems. Propulsion was generated by strong depression torques about the coxa-trochanter joint, not by retraction or flexion/extension torques. Torques about the respective thorax-coxa and femur-tibia joints were often directed opposite to fore-aft forces and joint movements. This suggests a posture-dependent mechanism that counteracts collapse of the leg under body load and directs the resultant force vector such that strong depression torques can control both body height and propulsion. Our findings parallel propulsive mechanisms described in other walking, jumping and flying insects, and challenge current control models of insect walking.

A Computational Model of a Descending Mechanosensory Pathway Involved in Active Tactile Sensing.

Many animals, including humans, rely on active tactile sensing to explore the environment and negotiate obstacles, especially in the dark. Here, we model a descending neural pathway that mediates short-latency proprioceptive information from a tactile sensor on the head to thoracic neural networks. We studied the nocturnal stick insect Carausius morosus, a model organism for the study of adaptive locomotion, including tactually mediated reaching movements. Like mammals, insects need to move their tactile sensors for probing the environment. Cues about sensor position and motion are therefore crucial for the spatial localization of tactile contacts and the coordination of fast, adaptive motor responses. Our model explains how proprioceptive information about motion and position of the antennae, the main tactile sensors in insects, can be encoded by a single type of mechanosensory afferents. Moreover, it explains how this information is integrated and mediated to thoracic neural networks by a diverse population of descending interneurons (DINs). First, we quantified responses of a DIN population to changes in antennal position, motion and direction of movement. Using principal component (PC) analysis, we find that only two PCs account for a large fraction of the variance in the DIN response properties. We call the two-dimensional space spanned by these PCs 'coding-space' because it captures essential features of the entire DIN population. Second, we model the mechanoreceptive input elements of this descending pathway, a population of proprioceptive mechanosensory hairs monitoring deflection of the antennal joints. Finally, we propose a computational framework that can model the response properties of all important DIN types, using the hair field model as its only input. This DIN model is validated by comparison of tuning characteristics, and by mapping the modelled neurons into the two-dimensional coding-space of the real DIN population. This reveals the versatility of the framework for modelling a complete descending neural pathway.

Comparative whole-body kinematics of closely related insect species with different body morphology.

Legged locomotion through natural environments is very complex and variable. For example, leg kinematics may differ strongly between species, but even within the same species it is adaptive and context-dependent. Inter-species differences in locomotion are often difficult to interpret, because both morphological and ecological differences among species may be strong and, as a consequence, confound each other's effects. In order to understand better how body morphology affects legged locomotion, we compare unrestrained whole-body kinematics of three stick insect species with different body proportions, but similar feeding ecology: Carausius morosus, Aretaon asperrimus and Medauroidea extradentata (=Cuniculina impigra). In order to co-vary locomotory context, we introduced a gradually increasing demand for climbing by varying the height of stairs in the setup. The species were similar in many aspects, for example in using distinct classes of steps, with minor differences concerning the spread of corrective short steps. Major differences were related to antenna length, segment lengths of thorax and head, and the ratio of leg length to body length. Whereas all species continuously moved their antennae, only Medauroidea executed high swing movements with its front legs to search for obstacles in the near-range environment. Although all species adjusted their body inclination, the range in which body segments moved differed considerably, with longer thorax segments tending to be moved more. Finally, leg posture, time courses of leg joint angles and intra-leg coordination differed most strongly in long-legged Medauroidea.

Spatial co-ordination of foot contacts in unrestrained climbing insects.

Animals that live in a spatially complex environment such as the canopy of a tree, constantly need to find reliable foothold in three-dimensional (3D) space. In multi-legged animals, spatial co-ordination among legs is thought to improve efficiency of finding foothold by avoiding searching-movements in trailing legs. In stick insects, a 'targeting mechanism' has been described that guides foot-placement of hind- and middle legs according to the position of their leading ipsilateral leg. So far, this mechanism has been shown for standing and tethered walking animals on horizontal surfaces. Here, we investigate the efficiency of this mechanism in spatial limb co-ordination of unrestrained climbing animals. For this, we recorded whole-body kinematics of freely climbing stick insects and analysed foot placement in 3D space. We found that touch-down positions of adjacent legs were highly correlated in all three spatial dimensions, revealing 3D co-ordinate transfer among legs. Furthermore, targeting precision depended on the position of the leading leg. A second objective was to test the importance of sensory information transfer between legs. For this, we ablated a proprioceptive hair field signaling the levation of the leg. After ablation, the operated leg swung higher and performed unexpected searching movements. Furthermore, targeting of the ipsilateral trailing leg was less precise in anteroposterior and dorsoventral directions. Our results reveal that the targeting mechanism is used by unrestrained climbing stick insects in 3D space and that information from the trochanteral hair field is used in ipsilateral spatial co-ordination among legs.

Encoding of near-range spatial information by descending interneurons in the stick insect antennal mechanosensory pathway.

Much like mammals use their whiskers, insects use their antennae for tactile near-range orientation during locomotion. Stick insects rapidly transfer spatial information about antennal touch location to the front legs, allowing for aimed reach-to-grasp movements. This adaptive behavior requires a spatial coordinate transformation from "antennal contact space" to "leg posture space." Therefore, a neural pathway must convey proprioceptive and tactile information about antennal posture and contact site to thoracic motor networks. Here we analyze proprioceptive encoding properties of descending interneurons (DINs) that convey information about antennal posture and movement to the thoracic ganglia. On the basis of response properties of 110 DINs to imposed movement of the distal antennal joint, we distinguish five functional DIN groups according to their sensitivity to three parameters: movement direction, movement velocity, and antennal joint angle. These groups are simple position-sensitive DINs, which signal the antennal joint angle; dynamic position-sensitive DINs, which signal the joint angle with strong dependence on movement; unspecific movement-sensitive DINs, which signal movement but not the velocity, position, or direction of movement; and ON- and OFF-type velocity-sensitive DINs. The activity of the latter two groups is increased/attenuated during antennal movement, with the spike rate increasing/decreasing linearly with antennal joint angle velocity. Some movement-sensitive DINs convey spikes to the thorax within 11 ms, suggesting a rapid, direct pathway from antennal mechanosensory to thoracic motor networks. We discuss how the population of DINs could provide the neural basis for the intersegmental spatial coordinate transfer between a touch sensor of the head and thoracic motor networks.

A hierarchical model for external electrical control of an insect, accounting for inter-individual variation of muscle force properties.

Cyborg control of insect movement is promising for developing miniature, high-mobility, and efficient biohybrid robots. However, considering the inter-individual variation of the insect neuromuscular apparatus and its neural control is challenging. We propose a hierarchical model including inter-individual variation of muscle properties of three leg muscles involved in propulsion (retractor coxae), joint stiffness (pro- and retractor coxae), and stance-swing transition (protractor coxae and levator trochanteris) in the stick insect Carausius morosus. To estimate mechanical effects induced by external muscle stimulation, the model is based on the systematic evaluation of joint torques as functions of electrical stimulation parameters. A nearly linear relationship between the stimulus burst duration and generated torque was observed. This stimulus-torque characteristic holds for burst durations of up to 500ms, corresponding to the stance and swing phase durations of medium to fast walking stick insects. Hierarchical Bayesian modeling revealed that linearity of the stimulus-torque characteristic was invariant, with individually varying slopes. Individual prediction of joint torques provides significant benefits for precise cyborg control.

Hybrid insect-computer robots ­ an exciting fusion of biology and technology ­ herald a future of small, highly mobile and efficient devices. However, these robots require a way to control the movements of the insects, a task made complex due to the differences between different insects' nervous and muscle systems. To bridge this gap, Owaki, Dürr and Schmitz studied the relationship between electrical stimulation of three leg muscles in the legs of stick insects and the resultant torque. To do these experiments, the scientists kept the body of the stick insects fixed and electrically stimulated one out of three leg muscles to produce walking-like movements. The results of these electrical stimulations allowed Owaki, Dürr and Schmitz to propose a model that predicts the torque created in the insect's joints when different patterns of electrical stimulation are applied to a leg muscle. The researchers identified a near-linear relationship between the duration of the electrical stimulus and the resultant torque. Moreover, the slope of this linear relationship can be estimated for individual animals with a few measurements only. This finding refines the precision of the motor control required to build individually tuned biohybrid robots. Investigating the precise control of insect biohybrid robots, particularly using stick insects, can lead to advancements in biohybrid robotics and enrich our understanding of insect locomotion. Owaki, Dürr and Schmitz's insights could lead to the creation of adaptable and highly mobile devices with many applications, but key challenges need to be addressed. First, model testing must be implemented in free-walking insects, and the electrical stimuli must be refined to mimic natural neuromuscular signals more closely.

Descending interneurons of the stick insect connecting brain neuropiles with the prothoracic ganglion.

Stick insects respond to visual or tactile stimuli with whole-body turning or directed reach-to-grasp movements. Such sensory-induced turning and reaching behaviour requires interneurons to convey information from sensory neuropils of the head ganglia to motor neuropils of the thoracic ganglia. To date, descending interneurons are largely unknown in stick insects. In particular, it is unclear whether the special role of the front legs in sensory-induced turning and reaching has a neuroanatomical correlate in terms of descending interneuron numbers. Here, we describe the population of descending interneurons with somata in the brain or gnathal ganglion in the stick insect Carausius morosus, providing a first map of soma cluster counts and locations. By comparison of interneuron populations with projections to the pro- and mesothoracic ganglia, we then estimate the fraction of descending interneurons that terminate in the prothoracic ganglion. With regard to short-latency, touch-mediated reach-to-grasp movements, we also locate likely sites of synaptic interactions between antennal proprioceptive afferents to the deutocerebrum and gnathal ganglion with descending or ascending interneuron fibres. To this end, we combine fluorescent dye stainings of thoracic connectives with stainings of antennal hair field sensilla. Backfills of neck connectives revealed up to 410 descending interneuron somata (brain: 205 in 19 clusters; gnathal ganglion: 205). In comparison, backfills of the prothorax-mesothorax connectives stained only up to 173 somata (brain: 83 in 16 clusters; gnathal ganglion: 90), suggesting that up to 60% of all descending interneurons may terminate in the prothoracic ganglion (estimated upper bound). Double stainings of connectives and antennal hair field sensilla revealed that ascending or descending fibres arborise in close proximity of afferent terminals in the deutocerebrum and in the middle part of the gnathal ganglia. We conclude that two cephalothoracic pathways may convey cues about antennal movement and pointing direction to thoracic motor centres via two synapses only.

Effect of Thoracic Connective Lesion on Inter-Leg Coordination in Freely Walking Stick Insects.

Multi-legged locomotion requires appropriate coordination of all legs with coincident ground contact. Whereas behaviourally derived coordination rules can adequately describe many aspects of inter-leg coordination, the neural mechanisms underlying these rules are still not entirely clear. The fact that inter-leg coordination is strongly affected by cut thoracic connectives in tethered walking insects, shows that neural information exchange among legs is important. As yet, recent studies have shown that load transfer among legs can contribute to inter-leg coordination through mechanical coupling alone, i.e., without neural information exchange among legs. Since naturalistic load transfer among legs works only in freely walking animals but not in tethered animals, we tested the hypothesis that connective lesions have less strong effects if mechanical coupling through load transfer among legs is possible. To do so, we recorded protraction/retraction angles of all legs in unrestrained walking stick insects that either had one thoracic connective cut or had undergone a corresponding sham operation. In lesioned animals, either a pro-to-mesothorax or a meso-to-metathorax connective was cut. Overall, our results on temporal coordination were similar to published reports on tethered walking animals, in that the phase relationship of the legs immediately adjacent to the lesion was much less precise, although the effect on mean phase was relatively weak or absent. Lesioned animals could walk at the same speed as the control group, though with a significant sideward bias toward the intact side. Detailed comparison of lesion effects in free-walking and supported animals reveal that the strongest differences concern the spatial coordination among legs. In free walking, lesioned animals, touch-down and lift-off positions shifted significantly in almost all legs, including legs of the intact body side. We conclude that insects with disrupted neural information transfer through one connective adjust to this disruption differently if they experience naturalistic load distribution. While mechanical load transfer cannot compensate for lesion-induced effects on temporal inter-leg coordination, several compensatory changes in spatial coordination occur only if animals carry their own weight.

Marker-Less Motion Capture of Insect Locomotion With Deep Neural Networks Pre-trained on Synthetic Videos.

Motion capture of unrestrained moving animals is a major analytic tool in neuroethology and behavioral physiology. At present, several motion capture methodologies have been developed, all of which have particular limitations regarding experimental application. Whereas marker-based motion capture systems are very robust and easily adjusted to suit different setups, tracked species, or body parts, they cannot be applied in experimental situations where markers obstruct the natural behavior (e.g., when tracking delicate, elastic, and/or sensitive body structures). On the other hand, marker-less motion capture systems typically require setup- and animal-specific adjustments, for example by means of tailored image processing, decision heuristics, and/or machine learning of specific sample data. Among the latter, deep-learning approaches have become very popular because of their applicability to virtually any sample of video data. Nevertheless, concise evaluation of their training requirements has rarely been done, particularly with regard to the transfer of trained networks from one application to another. To address this issue, the present study uses insect locomotion as a showcase example for systematic evaluation of variation and augmentation of the training data. For that, we use artificially generated video sequences with known combinations of observed, real animal postures and randomized body position, orientation, and size. Moreover, we evaluate the generalization ability of networks that have been pre-trained on synthetic videos to video recordings of real walking insects, and estimate the benefit in terms of reduced requirement for manual annotation. We show that tracking performance is affected only little by scaling factors ranging from 0.5 to 1.5. As expected from convolutional networks, the translation of the animal has no effect. On the other hand, we show that sufficient variation of rotation in the training data is essential for performance, and make concise suggestions about how much variation is required. Our results on transfer from synthetic to real videos show that pre-training reduces the amount of necessary manual annotation by about 50%.

Integrative Biomimetics of Autonomous Hexapedal Locomotion.

Despite substantial advances in many different fields of neurorobotics in general, and biomimetic robots in particular, a key challenge is the integration of concepts: to collate and combine research on disparate and conceptually disjunct research areas in the neurosciences and engineering sciences. We claim that the development of suitable robotic integration platforms is of particular relevance to make such integration of concepts work in practice. Here, we provide an example for a hexapod robotic integration platform for autonomous locomotion. In a sequence of six focus sections dealing with aspects of intelligent, embodied motor control in insects and multipedal robots-ranging from compliant actuation, distributed proprioception and control of multiple legs, the formation of internal representations to the use of an internal body model-we introduce the walking robot HECTOR as a research platform for integrative biomimetics of hexapedal locomotion. Owing to its 18 highly sensorized, compliant actuators, light-weight exoskeleton, distributed and expandable hardware architecture, and an appropriate dynamic simulation framework, HECTOR offers many opportunities to integrate research effort across biomimetics research on actuation, sensory-motor feedback, inter-leg coordination, and cognitive abilities such as motion planning and learning of its own body size.