Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE (www.oxfordhandbooks.com). (c) Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy).

date: 19 February 2018

Control of Locomotion in Hexapods

Abstract and Keywords

This article discusses legged locomotion in insects. It describes the basic patterns of coordinated movement both within each leg and among the various legs. The nervous system controls these actions through groups of joint pattern generators coupled through interneurons and interjoint reflexes in a range of insect species. These local control systems within the thoracic ganglia rely on leg proprioceptors that monitor joint movement and cuticular strain interacting with central pattern generation interneurons. The local control systems can change quantitatively and qualitatively as needed to generate turns or more forceful movements. In dealing with substantial obstacles or changes in navigational movements, more profound changes are required. These rely on sensory information processed in the brain that projects to the multimodal sensorimotor neuropils collectively referred to as the central complex. The central complex affects descending commands that alter local control circuits to accomplish appropriate redirected movements.

Keywords: tripod gait, proprioception, campaniform sensill, central pattern generator, reflex reversal, central complex

Insects are, by most criteria, the most successful creatures in the animal kingdom. Certainly, their agility contributes greatly to this status. Few if any forms of terrain present an insurmountable barrier to all insects. Although many insects fly over complex terrain, most typically move by walking, running, and jumping. Foraging insects walk slowly over floors or branches, scurry under rocks, climb up walls and over ceilings, or jump over barriers that if scaled to human dimensions would represent achievements beyond comprehension by the most accomplished athletes. Predatory insects stalk prey or lay in ambush before rapidly striking. Not to be undone, prey individuals have effective escape strategies that include running, jumping, or flying to avoid attack. These abilities have not been lost on engineers, who have turned to observing insect locomotion as inspiration for legged robotic devices. Here, we will describe our current understanding of how insect nervous systems generate basic movements and then modify them as needed to move efficiently through complex natural terrain.

Although these creatures are often described as “simple systems,” a close examination of their abilities reveals mechanisms that are elegant but not really simplistic. Insect locomotion represents a remarkable combination of mechanical principles, neural control, and sensory input leading to efficient muscular movement of leg joints. The problem encompasses a wide range of issues from biomechanics to both central and peripheral neurobiological factors as well as force development in muscle. No review could possibly cover all of these topics in any depth, and we will not attempt to do so in the space allotted to us. Rather, we will concentrate on the local neurobiological circuits of the thoracic ganglia that generate basic leg movements while allowing for some modification in the timing and level of motor activity. These circuits allow insects to adjust their movements as they encounter inclines, holes, or even walls. However, some barriers require the insect to actively redirect leg movements to alter posture or turn its body. Moreover, insect locomotion is not a simple reactive process. Rather, context-dependent factors including multisensory cues in the insect’s immediate surroundings, physiological state, and learning all impact locomotory decisions. These actions require further processing in brain regions leading to interactions between brain and thoracic circuits. Absent this higher control, an insect could walk with a normal gait and deal with small barriers, but the exceptional behaviors that were mentioned earlier would not occur. Thus, we will also describe more recent findings regarding interactions between the brain and thoracic ganglia.

Basic Leg Movements Associated With Walking

Although people have undoubtedly studied the leg movements of insects as long as humans have been aware of these remarkable creatures, formal reports on leg movement generally start with the observations of Hughes (1952) and Wilson (1966). They described the reproducible patterns of leg movements that most insects use. All insects have six legs. At slow speeds, the legs follow a metachronal pattern moving from the hind legs to middle and then to front legs on either side. Each leg alternates between a stance phase when the tarsi are on the ground and the animal is pushed forward and a swing phase when the tarsus is moved forward through the air. At higher speeds, this pattern moves into a modification of the metachronal pattern called the tripod gait (Fig. 1). Here the front and rear legs on one side of the animal move as a unit with the middle leg on the opposite side. This tripod alternates between swing and stance with a second tripod made up of the remaining legs. The tripod gait is very stable, because at most speeds the animal’s center of mass remains within the base of support, and it will not tip over. In cockroaches, the tripod gait can be further subdivided into a slower “amble,” a faster “trot” that has much less variability in leg coordination (Bender et al., 2011). An even faster escape run is used on rare occasions when the insect is threatened (only about 1% of the time).

Similar patterns of leg coordination occur in other insects (stick insects [Cruse, 1990], locusts [Duch & Pluger, 1995], Drosophila [Wosnitza et al., 2013]). Small insects such as Drosophila, which weigh less than 1 milligram, can move their legs very rapidly when walking (Fig. 1C, D). The importance of Drosophila as a model organism requires some special attention. Their small size, minute weight, and specializations for flight might suggest the use of unique mechanisms in the neural control of walking. However, Drosophila also show “gliding coordination” in walking and patterns of leg movements that change systematically and gradually with the rate of stepping speed. As with larger insects, their legs form tripods of support in fast running (15 steps/s), while more variable and metachronal gaits are evident in slow walking (8 steps/s).

Control of Locomotion in HexapodsClick to view larger

Figure 1 Leg coordination in walking. (A) Tripod gait pattern for a cockroach on a tether. Video frames showing the extreme rear position during stance of the two tripods of legs. (B) Stance and swing timing of each leg separated by color for the two tripods. Boxes indicate times when each leg is in swing. Note the Left front and hind legs move in time with the Right middle leg as do the remaining three legs. (C) Walking in freely moving fruit flies (Drosophila) was studied by digital imaging. (D) Gaits of flies vary with walking speed. Drosophila show tripod coordination at very rapid rates of stepping but less precise coordination and metachronal gaits at slower rates

(panels C and D from Wosnitza et al., 2013).

In all insects, each leg is made up of segments that are similar from leg to leg but differ in dimensions. From the most proximal to distal location, the leg segments are the coxa, trochanter, femur, tibia, and a series of tarsal segments ending in a retractable claw (Fig. 2). In the cockroach, the most important joints for walking are the coxa-trochanter (CTr) joint and the femur-tibia (FTi) joint. The CTr joint actually moves the femur relative to the coxa because the trochanter-femur (TrF) joint makes only small movements. Flexion of the TrF joint effectively rotates the tarsus and, in the middle and hind legs, this action is critical to initiating the swing phase of walking (Bender et al., 2010). Nevertheless, during many movements it acts mechanically as a fused joint. In stick insects and locusts, the coxa is much smaller and no movement occurs at the TrF joint (Büschges & Gruhn, 2008). The relative proportions of the other leg segments are also very different. These factors may reflect the locus of connection on the side of the body rather than the ventral surface, as in the cockroach. Drosophila legs also attach on the ventral side of the thorax (Wosnitza et al., 2013).

There are many specializations in leg design found in various insects. Thus, for example, the hind leg of a locust has a very long muscular femur, making a powerful jumping leg (Heitler, 1977; Heitler & Burrows, 1977). For most insects, the joints of the front legs are much more flexible, leading to greater range of motion (Ritzmann et al., 2004) that can be used in searching movements as, for example, stick insects that forage in natural habitats (Dürr, 2001).

Control of Locomotion in HexapodsClick to view larger

Figure 2 Diagrams describing the leg segments of the front and hind legs of the cockroach. (A) The positions of the legs on the body are shown at various different parts of the leg cycle. Note that the movements through swing and stance are different for front and hind legs (see text for details.) (B) Photographs of the front and hind legs. (C) The various segments are color coded and names are designated. Note that the same segments are present in both front and hind legs, but their relative sizes differ.

Although the legs within a tripod move their feet as a unit, the joint movements and resulting forces are unique for each pair of legs (Full et al., 1991; Watson & Ritzmann, 1998b; Tryba & Ritzmann, 2000). In cockroaches, the hind legs make powerful movements that drive the animal forward. To accomplish this action, the CTr joint and the FTi joint move in near synchrony. This action allows these rotary joints to direct the movements of the tarsi (feet) in a line nearly parallel to the long axis of the animal’s body. The middle legs make similar movements, but with smaller CTr extensions. The resulting actions of the middle legs generate forces that first brake the forward movement of the animal and then push it forward (Full et al., 1991).

In contrast, the front legs of cockroaches make much greater use of the thoraco-coxa (ThC) joint, which attaches the leg to the thorax (Ritzmann et al., 2004; Bender et al., 2010). The ThC joint has three degrees of freedom, similar to a ball-and-socket joint (Laurent & Richard, 1986), which might complicate mechanisms of control. However, measurements of ground reactions forces and kinematics in freely moving stick insects, which have highly mobile ThC joints in all legs, have shown that the major forces for support of body weight and propulsion in all legs are generated by the CTr joint (Dallmann et al., 2016). In contrast, the torques produced by protraction and retraction at the ThC joint mainly serve to stabilize the leg in walking on horizontal substrates. Thus, the functions of different joints in stick insect legs are similar to that seen in the hind and middle legs of cockroaches and other insects.

Despite these differences, the sum of the ground reaction forces of all three pairs of legs is similar to that seen in the bipedal leg movements of a human (Full et al., 1991). The middle and hind legs are also important for rotating the animal during turns (Dürr & Ebeling, 2005; Mu & Ritzmann, 2005).

Local Control of Basic Leg Movements: Reflexes and Central Pattern Generators

Even though differences are found in the detailed morphology of various insect legs, common neural control systems are found even in insects as different as cockroaches and stick insects. The muscles that extend leg joints are typically innervated by very few motor neurons. Most of the muscles that extend the CTr joints of cockroaches are innervated by two excitatory motor neurons: one fast and one slow. In slow movements, the slow motor neuron fires in bursts that lead to a series of facilitating muscle potentials that generate joint extension (Delcomyn & Usherwood, 1973; Delcomyn, 1985a, 1987; Watson & Ritzmann, 1988b). At faster speeds, the fast motor neurons are added (Delcomyn, 1985b; Watson & Ritzmann, 1988a). These neurons generate few spikes, but each causes a rapid extension of the joint, greatly increasing the joint velocity. Flexion of the joint is generated by antagonistic muscles that are typically controlled by a larger number of motor neurons.

How are these motor neurons activated in patterns that will move the legs efficiently? The local control systems found in the thoracic ganglia must be able to generate basic swing-stance patterns but still be open to adjustments in both timing and force. First, the individual joints must be made to alternate between extension and flexion. In each leg, these joints must be coordinated to generate effective leg movements, and then individual legs must be coordinated to generate effective gaits. Both intra- and interleg coordination is susceptible to change as the insect moves around its environment negotiating barriers and changing walking speed. On top of altered timing patterns, the forces generated by stance phase muscles must also adjust as the insect walks up and down inclines or walls, carries items, or even climbs over substantial barriers.

To accomplish these control functions, insects have evolved local control circuits that incorporate an elegant interplay between central pattern generators (CPGs) for each leg joint and sensory reflexes. The existence of joint CPGs was demonstrated in stick insects by cutting the peripheral nerves and recording from the proximal leg stumps. In this deafferented condition, application of the muscarinic agonist pilocarpine causes the motor neurons to burst. However, the timing of motor bursts serving various joints of deafferented preparations is not coordinated (Büschges et al., 1995). Thus, although there is no evidence suggesting an overarching walking or running pattern generator, there are joint CPGs that set timing for individual joints.

In addition to joint pattern generators, local circuit nonspiking interneurons have been identified that can directly generate activities at multiple joints of single legs. One interneuron, I4, is present on the right and left sides of each thoracic ganglion and plays a key role when stick insects search for a foothold (Berg et al., 2015). I4 is both necessary and sufficient to initiate and maintain searching movements in single legs. Furthermore, depolarization of I4 does not elicit searching after a leg makes contact with the substrate, showing that it can both integrate sensory inputs and be gated by signals of substrate contact. Other interneurons have also been shown to affect muscle activity at a number of leg joints (Büshges, 1995; von Uckermann & Büschges, 2009) and may be important in establishing activation of leg muscles as synergists (Zill et al., 2015).

What then is the role of the numerous sensors that are found on each leg? Insect legs contain a remarkable wealth of sensors (Zill, 1990), including hair plates that detect joint flexion, internal stretch receptors called chordotonal organs that monitor joint angle and velocity of motion, and campaniform sensilla that detect both the level and direction of strain. The hair plates are placed strategically on the leg so that extreme flexion on a given joint deflects the hairs and serve as limit sensors. The chordotonal organs are made of elastic strands that span each joint. As the joint moves, the strands are made to stretch and relax, and sensory neurons embedded in them report on the joint position and motion. Campaniform sensilla are made of flexible domes that are embedded in the insect’s cuticle at strategic locations to detect strain on the leg. Many have elliptical domes that expand as strain occurs in one axis of the leg and squeeze in the orthogonal axis. As such, they are positioned to detect loading of the leg during stance as well as decreases in loading that occur as the leg enters swing (Ridgel et al., 1999; Noah et al., 2004b; Zill et al., 2004).

Activation of these sensors reflexively affects both the level of motor activation and the timing of phase changes of various leg joints. For example, campaniform sensilla embedded in specific areas of the cuticle respond to changes in strain that allow the insect’s nervous system to adjust to changing load (Zill & Seyfarth, 1996; Zill et al., 2004) (Fig. 3). The signals of loads and muscle forces detected by campaniform sensilla can be used as positive feedback to rapidly develop muscle contractions needed in support and propulsion (Zill et al., 2012, 2015). They can also influence groups of muscles at different joints (e.g., leg extensors) to insure that they are tuned to work together as synergists (Akay et al., 2001; Zill et al., 2012]. Moreover, interjoint reflexes influence the mechanisms that change [switch] the phase of walking from flexion to extension and back again (Akay et al., 2004). Thus, the central timing of individual joint CPGs may be considered a switch of last resort that assures that a change in direction of joint movement will occur eventually. However, normally interjoint reflexes force that change before the limit is reached. This timing control is similar to that which has been described at the cellular level for leech heart (Arbas & Calabrese, 1987a, 1987b). There, pacemaker cells on the right and left side of the head ganglia have intrinsic properties that will cause their membrane potential to depolarize activating a Ca2+ dependent plateau potential which persists for a period of time before returning to baseline. In isolation, these intracellular properties will cause these cells to burst on their own. However, each neuron also receives reciprocal inhibition from its contralateral homolog. As that cell depolarizes, it shuts off the plateau potential on the first neuron. Thus, the intracellular properties act as a switch of last resort, which, in the intact system, may rarely occur.

The interplay between joint CPGs and sensory reflexes produces coordinated leg movements that are appropriate to the actions required at any time (Hess & Büschges, 1997; Akay et al., 2004). An advantage in determining joint coordination by this kind of sensory feedback is that it can be changed as needed by manipulating reflex properties. For example, if the insect reverses direction to move backward, the hind leg thorax-coxa joint must alter its pattern of movement, activating retractor motor neurons during swing and protractors during stance. To accomplish this change, the inter-joint reflexes associated with trochanteral campaniform sensilla (TrCS) switch from initiating coxal retractor motor neurons for forward walking to protractor motor neurons for backward movement (Akay et al., 2007). However, this mechanism is probably complemented by other elements in the local leg circuitry, as recent experiments have shown that reversals of the phase of activation of protractor and retractor muscles can occur in “fictive” preparations after elimination of sensory feedback from those muscles (Rosenbaum et al., 2015).

Likewise, in some turning movements, the legs on the inside of the turn switch from pushing the insect forward during stance to a pattern in which the joints extend during swing, then depress to the substrate and pull the body laterally during flexion. Again, these alterations are associated with reversals in specific interjoint reflexes (Hellekes et al., 2012).

Control of Locomotion in HexapodsClick to view larger

Figure 3 Campaniform sensilla of insect legs. (A) Drawing of a stick insect indicating the position of the middle leg trochanter. (B) Scanning electron micrograph of the trochanter cuticle. Many groups of sensory receptors are found in the trochanter including a hair plate and groups of campaniform sensilla. (C) Micrographs of whole mount preparations showing cuticular caps of two homologous groups of trochanteral campaniform sensilla in stick insects (left) and Drosophila (right). Strains in the exoskeleton are detected by processes of sensory neurons that attach to the caps. The two groups have similar mutually perpendicular cap orientations in both species.

(Panel B from Josef Schmitz and Annelie Exter, University of Bielefeld)

The flexibility in use of sensory information is also evident in studies documenting the effects of ablating sensory receptors. Discrete deficits in walking occur after selective ablation of some sense organs such as the trochanteral hairplate (Wong & Pearson, 1976; Dean & Schmitz, 1992). Removal of the hair plate, which is excited by joint flexion, produces exaggerated swing movements and overstepping in walking. However, ablation of most sense organs by themselves has little or no effect. The minimal impact of removing individual sensors even includes the femoral chordotonal organ, which has strong effects on motor output (Bässler, 1977). Similar results have been seen in recent experiments in Drosophila using genetic tools. Inactivation of sensory neurons in the fly’s legs, to block proprioceptive feedback, led to deficient step precision, but interleg coordination and the ability to execute a tripod gait were unaffected (Mendes et al., 2013).

The lack of effects found in sensory ablation studies could lie, at least in part, in the convergence of sensory modalities in motor control. Many studies have shown that inputs from different types of receptors converge in the central nervous system. Indeed, it is important to note that signals from many sense organs vary simultaneously during leg movement. For example, experiments in cockroaches showed that increasing body weight (through small magnets attached to the thorax) produced activation of leg campaniform sensilla concurrent with changes in leg joint angles (Keller et al., 2007). Thus, both these signals indicate the load perturbation and could be used by the nervous system to generate appropriate compensatory adjustments. Removal of one such sensory source would still leave the other.

The functions of sensory inputs are often better revealed by techniques that challenge the system through perturbations or reversals of sensory signals rather than ablation. Although removal of the femoral chordotonal organ had little effect, reversing the ligament of the receptor produced dramatic changes in posture and walking (Bässler, 1979; Graham & Bässler, 1981). Increases in body load (by applying small weights to the thorax) have also been shown to modify the rate of stepping and patterns of gait in walking in Drosophila (Mendes et al., 2014) in a manner similar to effects seen in cockroaches. The increasingly extensive array of genetic tools available in Drosophila could be applied to precisely analyze the effects of perturbing sensory inputs in walking.

Leg sensors also impact coordination between legs. This can occur as a result of the mechanics between the legs or through interganglionic neural connections. The mechanical effects on gait can be demonstrated by so-called peg leg experiments (von Buddenbrock, 1921). Prior to amputation of a middle leg, the insect walks in a tripod gait (Fig. 1). However, immediately after amputation, the insect switches to a diagonal gait where each front leg steps with the contralateral hind leg. This change could occur as a result of loss of loading information from the amputated leg; this hypothesis was confirmed by Wendler (1966), who returned an amputated stick insect leg to normal use by gluing a stick to the amputation stump. The same effect was accomplished with more accurate mechanics in cockroaches by creating a peg leg out of the insect’s own leg through denervation (Noah et al., 2004a). All of the nerves in a cockroach leg were severed in the femur, making the distal part of the leg into an anatomically perfectly matched peg. If the nerves were severed below the trochanteral segment, the leg stump was used normally in walking. However, if the nerves were cut above (proximal to) the trochanter, the leg showed repetitive flexion/extensions similar to searching movements. These findings demonstrated that sensory inputs from the trochanter that detect leg contact are essential for normal use of the leg in walking.

Similar effects of leg ablation have recently been demonstrated in fruit flies. The very rapid rates of stepping in flies might suggest that walking was entirely driven by a CPG with little modulation from sensory feedback. However, ablation of a single hind leg produces immediate, compensatory changes: The rate of walking is decreased and the stance trajectory of the adjacent middle leg is shifted to provide support (Wosnitza et al., 2013). During slow walking, the leg’s stumps make repetitive movements within single cycles of the remaining legs at a rate that is relatively constant (about 10 Hz) and largely independent of the movements of other legs (Berendes et al., 2015). However, in rapid running the stump movements tend to occur at the appropriate interleg phase. These findings suggest that walking in flies is similar to other insects in being driven by the interaction of central oscillatory circuits that are extensively modified by sensory feedback.

These experiments indicate that loading and unloading that occurs during the leg cycle not only influences joints within a leg but also affects the timing activation in adjacent legs. Some of this activation appears to occur through interganglionic neural connections. Experiments in stick insects demonstrated that when the front legs start walking the muscles of the middle legs will show bursting activities (Bergmann et al., 2009). Moreover, the influence of one leg on another can be demonstrated on an oiled-plate tether that mechanically isolates each leg from the others (Gruhn et al., 2009). Finally, stimulation of sensors located on the trochanter leg segment was necessary to establish the appropriate phase of bursting in which the middle legs alternated with the front legs (Noah et al., 2004a).

The data supporting neural properties that influence interleg coordination should not be taken as eliminating a role for mechanics. Mechanical linkages among the legs also affect gait. In freely walking animals, coordination of leg movements may be aided by mechanical forces transmitted through the substrate (ground), without relying on connections within the nervous system (Cruse, 1990; Cruse et al., 1998, 2007). In cockroaches, discharges of campaniform sensilla that detect unloading occur after an adjacent leg is placed in stance and pushes down on the substrate (Zill et al., 2009). These signals insure that the leg is not lifted until another leg is providing adequate support. Thus, once again, intact coordination is accomplished by a complex interplay among neural and mechanical properties.

Many of these effects have also been reproduced in dynamic simulations (Ekeberg et al., 2004; Daun-Grühn & Büschges, 2011). In these models, sensory reflexes are used to adjust the strength of motor activity, thereby creating effective movements for any situation in which the insect finds itself (e.g., walking on a flat horizontal surface, going up or down an incline, or climbing over an object). Another simulation accounted for all joint movements of all legs in a tethered cockroach as it transitioned between forward walking and turning in response to optomotor stimulation (Szczecinski et al., 2014). This last model relied heavily on known and predicted reflex reversals and with the addition of descending inputs that adjusted the strength of joint changes could produce a family of curves in a freely walking artificial agent.

More detail on the organization of the local control system found in thoracic ganglia of insects can be found in several excellent reviews (Zill et al., 2004; Ritzmann & Buschges, 2007; Büschges & Gruhn, 2008; Büschges et al., 2008, Büschges, 2012).

Negotiating Obstacles

The local control that we described earlier will allow insects to walk on horizontal surfaces, up and down inclines, or even on walls and ceilings (Duch & Plugerm, 1995; Larsen et al., 1995). However, some barriers require more profound adjustments. Faced with a large block, an intact cockroach traveling at normal speed rears up before it places its front legs on the top of the barrier. It can do that without actually contacting the front of the object with its legs (Fig. 4), by using information gained from sensors on its head (Watson et al., 2002). In particular, the cockroach appears to use its antennae to judge the height of the barrier (Harley et al., 2009) and then climbs accordingly.

Antennae are remarkable sensory organs that provide critical mechanical cues to insects as they move through their environments (Staudacher et al., 2005). The antennae are used in gap crossing to estimate distance and guide the legs across the gap (Blaesig & Cruse, 2004a, 2004b). In stick insects, ablation of the antennae demonstrates that the gap is estimated by mechanical cues from the antennae. However, in Drosophila, which have much shorter antennae, a similar gap estimate occurs through visual mechanisms (Pick & Strauss, 2005). In more complex environments, stick insects actively probe objects with their antennae to guide the insect in climbing maneuvers in a context-dependent manner (Schutz & Durr, 2011).

Detection of a substantial object in an insect’s path leads to precise climbing or turning movements. In stick insects the height of the last antennal contact prior to leg movement predicts the point where the front leg will contact the object (Schutz & Durr, 2011). Intact cockroaches moving at normal foraging speeds palpate a block or shelf with their antennae prior to rearing up as they climb over the object (Harley et al., 2009). The higher the object, the higher the rearing movement. As with stick insects, ablation of the antennae compromises this detection mechanism. Antennae can also be used to seek out objects of interest such as bars placed in the periphery (Okada & Toh, 2000). In this case, hair plates at the bases of the antennae report where the antennae were upon contact and this information leads to activation of neurons in the brain that direct turns toward the object (Guo & Ritzmann, 2013).

For climbing, intact insects rear up to the appropriate height as dictated by antennal or visual information (Watson et al., 2002; Harley et al., 2009; Schutz & Durr, 2011). In cockroaches, the change in leg movement that occurs during rearing requires a rotation of the ThC joint of the middle legs (Fig. 4) (Watson et al., 2002). That rotation points the leg downward and with this change in leg orientation, extension of the more distal leg joints now raises the front of the body upward. In this new posture, the cockroach can now easily swing the front legs to the top of the block and then, by extending the hind legs, push its body up to the top surface. Although the rearing and climbing movements use the same distal leg joints moving through the same joint angles as seen in horizontal walking, the motor activity is typically enhanced as the cockroach moves upward against gravity (Watson et al., 2002). This observation suggests that the same reflex circuits that were described previously come into play here. Having altered its posture, the cockroach moves upward against gravity, which may increase the strain on its cuticle as well as on the actual muscles. These changes should be detected by campaniform sensilla along the leg which would then activate reflexes that increase motor activity. Thus, the entire behavior to surmount the block would involve detection of the large object with sensors such as antennae or eyes, leading to descending commands that alter posture through changes in orientation of the middle leg’s ThC joints. After the cockroach rears up, control is returned to the local circuits of the thoracic ganglia that generate normal leg joint extensions, but now with enhanced force to push the insect upward.

Control of Locomotion in HexapodsClick to view larger

Figure 4 Sequence of climbing movements made by cockroach negotiating a block. Elapses of 50 msec between each frame. Note that the front legs never contact the front of the block. Also note, at 150 msec, the middle leg (outlined in red) has rotated at the ThC joint to bring that legs tibia nearly perpendicular to the substrate. Thus, when the leg extends in the next frame the body is pushed upward rather than forward.

If the block is replaced with a shelf, there are two possible outcomes. The cockroach can either climb over the shelf or tunnel under it. Again the antennae are critical to this decision. If they tap the shelf from above, the cockroaches typically climb, but if they tap the under surface, they tunnel under the shelf (Harley et al., 2009). Interestingly, this decision is also influenced by ambient light levels. In the light, there is a much greater tendency for the cockroach to tunnel. However, in the dark, there is no significant difference between climbing and tunneling. This observation suggests that in the light, these nocturnal insects are predisposed to seek out shelter, whereas in the dark they perform more natural foraging behaviors. The detection of light that is important to this process is performed by the ocelli (two simple eyes on the head). If they are covered, there is no longer a preference for tunneling, even in the light, whereas if the compound eyes are covered, there is no change in the behavior (Harley et al., 2009).

A Role for Brain Circuits in Motor Control

The observation that multiple sensory inputs detected by sensors on the head affect climbing behaviors suggests that association areas of the brain may be very much involved in related decisions. This conclusion is supported by observations of cockroach behavior after ablation of connectives that link higher centers to the thoracic ganglia (Ridgel & Ritzmann, 2005). With neck connectives cut, cockroaches make very small movements, if any. However, if the circumoesophageal connectives are cut removing the brain, but retaining connections between the suboesophageal and thoracic ganglia, normal tripod gait movements are observed but with little ability to alter walking patterns in response to objects in the insect’s path. Indeed, they willingly walk off the ends of tables. In effect, removal of the brain appears to have reduced movement to a static pattern that loses many of the complex behaviors that attract robotic engineers to insects as model systems while retaining the basic tripod gaits.

Within the brain, the central complex (CX) has been specifically suggested to be involved in supervising such motor activity (Strausfeld, 1999, 2012). It is highly conserved in arthropods and can even be detected in the fossils of their ancestors (Ma et al., 2015). The CX is made up of several prominent neuropils on the midline of the brain of all insects (Pfeiffer & Homberg, 2014). These neuropils are highly columnar with fibers projecting between neuropils in a very regular array that suggests that they may be comparing left and right inputs. In locusts and monarch butterflies, the CX contains numerous neurons that are sensitive to polarized light and that could be used to guide them as they migrate (Heinze & Homberg, 2009; Heinze & Reppert, 2011). In cockroaches, neurons sensitive to both mechanical stimulation of antennae and changes in ambient light levels have been recorded using extracellular methods (Ritzmann et al., 2008).

Extracellular recordings during visual responses to moving stripes also identified a range of response properties in CX neurons, including phasic responses when the stripe patterns were turned on or off and tonic responses during the stripe movements (Kathman et al., 2014). The tonic responses were often directional in that movement of stripes in one direction generated strong activation, while stripe movement in the opposite direction resulted in weak responses, no activation at all, or suppression of ongoing activity. These observations suggest that CX neurons could play a critical role in optomotor responses. In support of this hypothesis, injection of the local anesthetic, procaine, into the CX of tethered cockroaches silenced CX neurons and temporarily blocked optomotor turning (Kathman et al., 2014). Approximately 20 minutes after the injection, activity returned to CX neurons as did the optomotor response. Control experiments with saline injection had no such effect.

Thus, the CX appears to process various types of sensory cues. Does it affect movement? Drosophila with genetic mutations that generate breaks in one of the CX neuropils (the protocerebral bridge) have difficulty walking (Strauss et al., 1992; Strauss, 2002). Moreover, electrolytic lesions within various regions of the CX in cockroach result in deficits in specific behaviors (Harley & Ritzmann, 2010). For example, lesion in lateral regions of one of the CX neuropils called the fan-shaped body increases the number of wrong turns when the cockroach is asked to walk in a U-shaped track. Lesions on the midline of this neuropil do not affect turning but do alter climbing over blocks or shelves. Finally, extracellular recordings in the CXs of tethered cockroaches reveal neurons that alter their firing rate in tandem with and, for some neurons, in advance of changes in step frequency. Moreover, stimulation through the same electrodes evokes changes in step frequency (Bender et al., 2010). Similarly, neurons in the lateral regions of the fan-shaped body are tuned to turning movements toward the side on which they are recorded, and stimulation in that region reliably evokes turning in the appropriate direction (Guo & Ritzmann, 2013). This and other bits of evidence are consistent with the notion that the CX takes in various different forms of sensory information and uses them to influence descending commands that alter activity in the thoracic ganglia, leading to turning, climbing, or tunneling movements (Ritzmann et al., 2012).

Although these recordings in tethered cockroaches clearly show a motor effect in CX neurons, the complex feedback mechanisms that must be present in freely behaving insects were eliminated by the open loop tether. A more recent study recorded CX neurons of freely walking cockroaches (Martin et al., 2015). Here extracellular activity recorded from individual CX neurons was related to forward translational movements and left-right rotations in an open arena. At any given point in time these two factors generate a vector, and a color-coded square at the tip of the vector shows the neuron’s activity level during, before, or after the change in movement. By filling in the various directional vectors, a plot was generated of the neuron’s activity pattern as the insect explored the arena (Fig. 5). This plot was then smoothed and contours were generated representing 0% to 100% activity levels. The 50% contours were then extracted and compared with other CX neurons recorded in the same and other insects. Importantly, most of the changes in activity that were recorded relative to both turning and translational movement changes occurred before the movement was altered, suggesting that they were part of a descending command system. However, some did occur after the movement, suggesting feedback mechanisms also occurred, and a few trials included both prior and subsequent activation.

Control of Locomotion in HexapodsClick to view larger

Figure 5 Locomotor-related activity in the central complex of the freely moving cockroach. (A) The path of a cockroach exploring the open arena. Color indicates firing rate of an example central-complex neuron during each segment. Black dots indicate the position of the cockroach every five frames. (B) Raw spike times (rasters), the smoothed firing rate of a central-complex neuron (orange), and translational velocity (blue) and rotational velocity (cyan) of the animal during the bouts indicated at (i) and (ii) in (A). Gray shading indicates the delay between peaks in the firing rate and peak rotational velocity of the resulting movement bout. (C) A point-process generalized linear model (GLM) was used to determine the contribution of each movement parameter on the firing probability of each recorded neuron. The result is a linear filter or “kernel” quantifying the relative weight of each movement parameter in predicting the relative weight of each movement parameter in predicting spikes preceding or following each video frame. Kernels generated for this example neuron show the relative influence of the translation (blue) right rotation (red) and left rotation (green) components, with lags up to ten frames (0.5 s). Kernels are interpolated and smoothed for display. (D) Mean rotational velocity versus the peak firing rate of this example neuron during spontaneous walking in the open arena. (E) A firing-rate map for this example neuron for all rotational and translational velocity combinations. The contour for the 50% firing-rate level is overlaid (green line, from the smoothed and gap-filled firing-rate map).

(Reprinted from Martin et al., 2015, with permission from Elsevier)

The 50% contour plots of all neurons recorded in that study covered the entire range of movements seen in the arena (Fig. 6). A closer examination shows that selection of different arrays of neurons occurs in association with either right or left turning and rapid or slow translation. This finding suggests that CX neurons could represent a population code that is altered by the cockroach according it its desired direction and speed of movement.

Control of Locomotion in HexapodsClick to view larger

Figure 6 Contour lines in the translational and rotational velocity axes at the 50% of maximum firing-rate levels from the firing-rate map at the best delay (as in Fig. 5E). Note that these contours span the entire range of translational and rotational movements.

(Reprinted from Martin et al., 2015, with permission from Elsevier)

How might these descending commands redirect leg movements? Again we must return to the local reflex circuits of the thoracic ganglia. Remember that the CPGs for each joint are coordinated by local reflexes. However, there appear to be multiple reflex circuits present. When a stick insect walks forward, stimulation of the trCS causes a switch in the ThC motor neurons from protraction to retraction (Akay et al., 2007). However, if the insect walks backward, reflex effects on the ThC joint reverse. Now trCS activation causes the opposite switch from retraction to protraction. The local reflex effects are, therefore, not constant but change according to the direction of walking.

A similar effect could account for changes in the cockroach’s turning movements. When the cockroach turns on a tether, the middle leg on the inside of the turn changes it movements so that joint extension now occurs during swing rather than stance (Mu & Ritzmann, 2005). This alteration causes the leg to reach out laterally and then after setting down pull medially in an attempt to move the body toward the tarsus. If opposing reflexes exist in the cockroach, as in the stick insect, this switch could be caused by descending commands enhancing one reflex while reducing the strength of the opposite one.

Such a descending effect on reflex reversals can be demonstrated by removing all descending activity through bilateral ablation of neck connectives. Stretching the femoral chordotonal organ (FCo) that monitors femur-tibia joint angle normally enhances activity in the trochanteral slow depressor (Ds) motor neuron. Relaxation of the FCo inhibits ongoing Ds activity. However, after cutting both neck connectives, the activation associated with FCo stretch is greatly reduced while the inhibition associated with relaxation actually reverses to now produce excitation. Similar reflex reversals are seen in appropriate legs during turning movements of stick insects (Hellekes et al., 2012). These studies suggest that modulation or switching of reflexes may be an important function of higher centers.

A test of this hypothesis requires that electrodes implanted into the CX for recording purposes could also generate turns in a consistent direction via electrical stimulation. If this could be demonstrated, one could expose the FCo and ask whether the appropriate reflex reversal for that turning direction could be generated by the same CX stimulation.

Indeed, in the study described earlier, several cockroaches with tetrode wires in place for recording were subsequently stimulated through the same wires (Martin et al., 2015). Most of these trials generated turns to the side ipsilateral to the side that the wires were on. A few generated contralateral turns, but in all cases the direction of turning was consistent from trial to trial and was also consistent with the directional biases that had been noted in the recorded CX neuron activity. These subjects also had electromyogram (EMG) electrodes implanted in the coxa to monitor the slow depressor (Ds) motor neuron. When the turn was to the same direction as the EMG implants, this inside leg showed characteristic changes in Ds activity (Fig. 7A). During walking, Ds turned on just prior to the femur-tibia (FTi) extension at the end of swing, and it was maintained throughout stance (most of the FTi extension). It then rapidly turned off at the end of FTi extension. During this leg cycle the FCo is stretched during FTi flexion and relaxed during FTi extension. Thus, the rapid decline in Ds activity with FCo relaxation is consistent with the inhibition seen during the FCo reflex. When the insect turns, the inside leg extends during swing and flexes during stance. Under this condition, Ds activity tends to have a strong activation during FTi extension (FCo relaxation) and weaker activation during FTi flexion (FCo stretch).

Control of Locomotion in HexapodsClick to view larger

Figure 7 Responses of the slow depressor (Ds) motor neuron during (A) free walking and (B) reflex activation of femoral chordotonal organ. (A) Blue plot was taken from the cockroach as it walked forward. Red line was taken from the inside leg during turning induced by stimulation through tetrode wires implanted in the CX. (B) Same insect moved to a restraint preparation for dissection of the femoral chordotonal organ. Here the cyan plot is the baseline reflex, while the red plot is the same response to the chordotonal organ movement while the CX is stimulated using the same tetrodes. In both plots, the black curve indicates the real (A) and predicted (B) femur-tibia joint movement while the green curve the predicted (A) and real (B) movement of the femoral chordotonal organ.

(Modified from Martin et al., 2015, with permission from Elsevier)

After CX and EMG responses were noted in the free walking condition, the insect was moved with the CX tetrodes in place to a preparation dish where the FCo could be exposed and attached to a transducer (Fig. 7B). Under these conditions, stretching FCo activated Ds (consistent with the activation seen at the end of swing in the free walking condition) and then turned off as FCo was relaxed (consistent with the rapid termination at the end of stance). In the free walking condition, Ds activity was maintained for a longer period during stance, presumably due to the loading effects on c.s. receptors during that period. When the FCo reflex was repeated along with CX stimulation (that caused turning in the free walking state), the Ds inhibition seen during FCo relaxation reversed to generate strong activation reminiscent of the strong activation that occurred in the inside leg during turning as FTi extended.

This observation strongly suggests that CX activity affects descending commands that in turn act by altering local reflexes in the thoracic ganglia, leading to redirected leg movements. In this way, CX circuits can use global sensory and physiological state information to orchestrate the insect’s leg movements in a context-dependent manner.

In addition to the CX, descending interneurons originating in the suboesophageal (gnathal) ganglion appear to provide a quicker shortcut from antennal activation to the local circuits of the thoracic ganglia, bypassing CX involvement (Ache et al., 2015). At least one such neuron has an ascending branch that projects to the brain, possibly informing the CX circuitry.

The alterations that are associated with CX activity are not restricted to horizontal turning. Activation changes also occur in CX neurons just prior to climbing behaviors over blocks (Martin et al., 2015). Plots of the relationship between walking speed and CX neuron activity show that some CX neurons have the same pattern in walking and climbing, but others are either elevated or show changes in slope when the cockroach is engaged in climbing behavior. Thus, the pattern of the CX population code seen in Figure 6 is transformed to a new pattern when the cockroach enters into a climbing state. It is reasonable to suggest that other state changes also occur as the insect, for example, flies, feeds, or seeks mates.

Summary of Motor Control in Complex Environments

The image that arises from the observations that are described herein is one that includes complex interactions of neural circuits that reside in different regions of the insect CNS. Local reflexes and CPGs interact in thoracic ganglia to generate basic leg movements. Reflexes also account for changes in strength of motor activity at appropriate times, such as when walking up or down inclines or after postural adjustments to rear up in preparation for climbing over blocks. When the insect approaches an object that is too large to be negotiated with simple reflex adjustments or something that it wishes to further investigate, sensors on the head evaluate it, and the resulting information is integrated within brain circuits such as those that reside in the CX. Here appropriate commands are formulated that descend to the thoracic ganglia where they can redirect leg movement by altering the strength of competing interjoint reflexes.

With this type of hierarchical motor control, the insect has the best of both worlds. It has local reflex circuits that can quickly adjust movement as needed, but it also has a sophisticated brain that can take in large amounts of data from the wealth of sensors located on its head and then use that information to temporarily redirect leg movements as it initiates transient turning, climbing, or tunneling movements while maintaining the stability that is inherent in the local feedback circuits. These systems allow insects to move through very diverse environments at will even as environmental conditions and its own internal state change dramatically. As such, they serve as excellent models for human-engineered vehicles that are designed to move through a range of dangerous terrains, but only when all parts of the hierarchical controls that we describe are intact.

Acknowledgments

This material is based in part upon work supported by the National Science Foundation under grant no. IOS-0845417 and the AFOSR under grant FA9550-10-1-0054 to RER and grant nos. IBN-0235997 (to SZ) and MRI 0959012 to the Marshall University Molecular and Biological Imaging Center.

References

Ache, J. M., Haupt, S., & Dürr, V. (2015). A direct descending pathway informing locomotor networks about tactile sensor movement. Journal of Neuroscience, 35(9), 4081–4091. doi:10.1523/jneurosci.3350-14.2015Find this resource:

Akay, T., Bässler, U., Gerharz, P., & Büschges, A. (2001). The role of sensory signals from the insect coxa-trochanteral joint in controlling motor activity of the femur-tibia joint. Journal of Neurophysiology, 85(2), 594–604.Find this resource:

Akay, T., Haehn, S., Schmitz, J., & Büschges, A. (2004). Signals from load sensors underlie interjoint coordination during stepping movements of the stick insect leg. Journal of Neurophysiology, 92(1), 42–51.Find this resource:

Akay, T., Ludwar, B. C., Goritz, M. L., Schmitz, J., & Büschges, A. (2007). Segment specificity of load signal processing depends on walking direction in the stick insect leg muscle control system. Journal of Neuroscience, 27(12), 3285–3294. doi:10.1523/jneurosci.5202-06.2007.Find this resource:

Arbas, E. A., & Calabrese, R. L. (1987a). Ionic conductances underlying the activity of interneurons that control heartbeat in the medicinal leech. Journal of Neuroscience, 7(12), 3945–3952.Find this resource:

Arbas, E. A., & Calabrese, R. L. (1987b). Slow oscillation of membrane potential in interneurons that control heartbeat in the medicinal leech. Journal of Neuroscience, 7(12), 3953–3960.Find this resource:

Bässler, U. (1977). Sensory control of leg movement in the stick insect, Carausius morosus. Biological Cybernetics, 25, 61–71.Find this resource:

Bässler, U. (1979). Effects of crossing the receptor apodeme of the femoral chordotonal organ on walking, jumping and singing in locusts and grasshoppers. Journal of Comparative Physiology A, 134, 173–176.Find this resource:

Bender, J. A., Pollack, A. J., & Ritzmann, R. E. (2010). Neural activity in the central complex of the insect brain is linked to locomotor changes. Current Biology, 20(10), 921–926.Find this resource:

Bender, J. A., Simpson, E. M., & Ritzmann, R. E. (2010). Computer-assisted 3D kinematic analysis of all leg joints in walking insects. PLoS One, 5(10), e13617. doi:10.1371/journal.pone.0013617.Find this resource:

Bender, J. A., Simpson, E. M., Tietz, B. R., Daltorio, K. A., Quinn, R. D., & Ritzmann, R. E. (2011). Kinematic and behavioral evidence for a distinction between trotting and ambling gaits in the cockroach Blaberus discoidalis. Journal of Experimental Biology, 214(12), 2057–2064. doi:10.1242/jeb.056481.Find this resource:

Berendes, V., Büschges, A., & Bockemühl, T. (2015). Speed-dependent interplay between CPG activity and sensory feedback during walking in Drosophila. 11th Goettingen Meeting of the German Neuroscience Society, March 2015. Goettingen, Germany.Find this resource:

Berg, E. M., Hooper, S. L., Schmidt, J., & Büschges, A. (2015). A leg-local neural mechanism mediates the decision to search in stick insects. Current Biology, 25(15), 2012–2017.Find this resource:

Blaesing, B., & Cruse, H. (2004a). Stick insect locomotion in a complex environment: Climbing over large gaps. Journal of Experimental Biology, 207(Pt 8), 1273–1286.Find this resource:

Blaesing, B., & Cruse, H. (2004b). Mechanisms of stick insect locomotion in a gap-crossing paradigm. Journal of Comparative Physiology A, 190(3), 173–183.Find this resource:

Borgmann, A., Hooper, S. L., & Büschges, A. (2009). Sensory feedback induced by front-leg stepping entrains the activity of central pattern generators in caudal segments of the stick insect walking system. Journal of Neuroscience, 29(9), 2972–2983.Find this resource:

Büschges, A. (1995). Role of local nonspiking interneurons in the generation of rhythmic motor activity in the stick insect. Journal of Neurobiology, 27(4), 488–512.Find this resource:

Büschges, A., Schmitz, J., & Bässler, U. (1995). Rhythmic patterns in the thoracic nerve cord of the stick insect induced by pilocarpine. Journal of Experimental Biology, 198, 435–456.Find this resource:

Büschges, A., Akay, T. Gabriel, J. P., & Schmidt, J. (2008). Organizing network action for locomotion: Insights from studying insect walking. Brain Research Review, 57(1), 162–171.Find this resource:

Büschges, A., & Gruhn, M. (2008). Mechanosensory feedback in walking: From joint control to locomotory patterns. Advances in Insect Physiology, 34, 194–234.Find this resource:

Büschges, A. (2012). Lessons for circuit function from large insects: Towards understanding the neural basis of motor flexibility. Current Opinion in Neurobiology, 22, 602–608.Find this resource:

Cruse, H. (1990). What mechanisms coordinate leg movement in walking arthropods? Trends in Neurosciences, 13, 15–21.Find this resource:

Cruse, H., Kindermann, T., Schumm, M., Dean, J., & Schmitz, J. (1998). Walknet—a biologically inspired network to control six-legged walking. Neural Network Office of the Journal of International Neural Network Society, 11(7–8), 1435–1447.Find this resource:

Cruse, H., Dürr, V., & Schmitz, J. (2007). Insect walking is based on a decentralized architecture revealing a simple and robust controller. Philosophical Transactions Series A. Mathematics, Physical, and Engineering Sciences, 365(1850), 221–250.Find this resource:

Dallmann, C., Dürr, V., & Schmitz, J. (2016). Joint torques in a freely walking insect reveal distinct functions of leg joints in propulsion and posture control. Proceedings of the Biological Sciences, 27(1823), pii: 20151708. doi:10.1098/rspb.2015.1708Find this resource:

Daun-Gruhn, S., & Büschges, A. (2011). From neuron to behavior: Dynamic equation-based prediction of biological processes in motor control. Biological Cybernetics, 105, 71–88.Find this resource:

Dean, J., & Schmitz, J. (1992). The two groups of sensilla in the ventral coxal hairplate of Carausius morosus have different roles during walking. Physiology and Entomology, 17, 331–341.Find this resource:

Delcomyn, F., & Usherwood, P. N. R. (1973). Motor activity during walking in the cockroach Periplaneta americana. I. Free walking. Journal of Experimental Biology, 148, 629–642.Find this resource:

Delcomyn, F. (1985a). Walking and running. In G. A. Kerkut & L. I. Gilbert (Eds.), Comprehensive insect physiology, biochemistry and pharmacology: Nervous system: Structure and motor function (pp. 439–466). Oxford, England: Pergamon Press.Find this resource:

Delcomyn, F. (1985b). Factors regulating insect walking. Annual Review of Entomology, 30, 239–256.Find this resource:

Delcomyn, F. (1987). Motor activity during searching and walking movements of cockroach legs. Journal of Experimental Biology, 133, 111–120.Find this resource:

Duch, C., & Pfluger, H.-J. (1995). Motor patterns for horizontal and upside down walking and vertical climbing in the locust. Journal of Experimental Biology, 198, 1963–1976.Find this resource:

Dürr, V. (2001). Stereotypic leg searching movements in the stick insect: Kinematic analysis, behavioural context and simulation. Journal of Experimental Biology, 204, 1589–1604.Find this resource:

Dürr, V., & Ebeling, W. (2005). The behavioural transition from straight to curve walking: Kinetics of leg movement parameters and the initiation of turning. Journal of Experimental Biology, 208, 2237–2252.Find this resource:

Ekeberg, Ö., Blümel, M., & Büschges, A. (2004). Dynamic simulation of insect walking. Arthropod Structure and Development, 33, 287–300.Find this resource:

Full, R. J., Blickhan, R., & Ting, L. H. (1991). Leg design in hexapedal runners. Journal of Experimental Biology, 158, 369–390.Find this resource:

Graham, D., & Bässler, U. (1981). Effects of afference sign reversal on motor activity in walking stick insects (Carausius morosus). Journal of Experimental Biology, 91, 179–193.Find this resource:

Gruhn, M., Zehl, L., & Büschges, A. (2009). Straight walking and turning on a slippery surface. Journal of Experimental Biology, 212(2), 194–209. doi:10.1242/jeb.018317Find this resource:

Guo, P., & Ritzmann, R. E. (2013). Neural activity in the central complex of the cockroach brain is linked to turning behaviors. Journal of Experimental Biology, 216(6), 992–1002. doi:10.1242/jeb.080473Find this resource:

Harley, C. M., English, B. A., & Ritzmann, R. E. (2009). Characterization of obstacle negotiation behaviors in the cockroach, Blaberus discoidalis. Journal of Experimental Biology, 212(10), 1463–1476. doi:10.1242/jeb.028381Find this resource:

Harley, C. M., & Ritzmann, R. E. (2010). Electrolytic lesions within central complex neuropils of the cockroach brain affect negotiation of barriers. Journal of Experimental Biology, 213(16), 2851–2864. doi:10.1242/jeb.042499Find this resource:

Heinze, S., & Homberg, U. (2009). Linking the input to the output: New sets of neurons complement the polarization vision network in the locust central complex. Journal of Neuroscience, 29(15), 4911–4921.Find this resource:

Heinze, S., & Reppert, S. M. (2011). Sun compass integration of skylight cues in migratory monarch butterflies. Neuron, 69(2), 345–358.Find this resource:

Heitler, W. J. (1977). The locust jump. III. Structural specializations of the metathoracic tibiae. Journal of Experimental Biology, 67, 29–36.Find this resource:

Heitler, W. J., & Burrows, M. (1977). The locust jump. I. The motor programme. Journal of Experimental Biology, 66, 203–219.Find this resource:

Hellekes, K., Blincow, E., Hoffmann, J., & Büschges, A. (2012). Control of reflex reversal in stick insect walking: Effects of intersegmental signals, changes in direction, and optomotor-induced turning. Journal of Neurophysiology, 107(1), 239–249. doi:10.1152/jn.00718.2011Find this resource:

Hess, D., & Büschges, A. (1997). Sensorimotor pathways involved in interjoint reflex action of an insect leg. Journal of Neurobiology, 33(7), 891–913.Find this resource:

Hughes, G. M. (1952). The co-ordination of insect movements. I. The walking movements of insects. Journal of Experimental Biology, 29, 267–284.Find this resource:

Kathman, N. D., M. Kesavan, & R. E. Ritzmann (2014). Encoding wide-field motion and direction in the central complex of the cockroach Blaberus discoidalis. Journal of Experimental Biology, 217(22), 4079–4090. doi:10.1242/jeb.112391Find this resource:

Keller, B. R., Duke, E. F., Amer, A. S., & Zill, S. N. (2007). Tuning posture to body load: Decreases in load produce discrete sensory signals in the legs of freely standing cockroaches. Journal of Comparative Physiology A, 193, 881–891.Find this resource:

Larsen, G. S., Frazier, S. F., Fish, S. E., & Zill, S. N. (1995). Effects of load inversion in cockroach walking. Journal of Comparative Physiology A, 176(2), 229–238.Find this resource:

Laurent, G., & Richard, D. (1986). The organization and role during locomotion of the proximal musculature of the cricket foreleg. I. Anatomy and innervation. Journal of Experimental Biology, 123, 255–283.Find this resource:

Ma, X., Edgecombe, G. D., Hou, X., Goral, T., & Strausfeld, N. J. (2015). Preservational pathways of corresponding brains of a Cambrian euarthropod. Current Biology, 25, 1–7.Find this resource:

Martin, J. P., Guo, P., Mu, L., Harley, C. M., & Ritzmann, R. E. (2015). Central-complex control of movement in the freely walking cockroach. Current Biology, 25, 1–9.Find this resource:

Mendes, C. S., Bartos, I., Akay, T., Marka, S., & Mann, R. S. (2013). Quantification of gait parameters in freely walking wild type and sensory deprived Drosophila melanogaster. eLife, 2, e00231.Find this resource:

Mendes, C. S., Rajendren, S. V., Bartos, I., Marka, S., & Mann, R. S. (2014). Kinematic responses to changes in walking orientation and gravitational load in Drosophila melanogaster. PloS One, 9, e109204.Find this resource:

Mu, L., & Ritzmann, R. E. (2005). Kinematics and motor activity during tethered walking and turning in the cockroach, Blaberus discoidalis. Journal of Comparative Physiology A, 191, 1037–1054.Find this resource:

Noah, J. A., Quimby, L., Frazier, S. F., & Zill, S. N. (2004a). Walking on a “peg leg”: Extensor muscle activities and sensory feedback after distal leg denervation in cockroaches. Journal of Comparative Physiology A, 190(3), 217–231.Find this resource:

Noah, J. A., Quimby, L., Frazier, S. F., & Zill, S. N. (2004b). Sensing the effect of body load in legs: Responses of tibial campaniform sensilla to forces applied to the thorax in freely standing cockroaches. Journal of Comparative Physiology A, 190(3), 201–215.Find this resource:

Okada, J., & Toh, Y. (2000). The role of antennal hair plates in object-guided tactile orientation of the cockroach (Periplaneta americana). Journal of Comparative Physiology A, 186, 849–857.Find this resource:

Pfeiffer, K., & Homberg, U. (2014). Organization and functional roles of the central complex in the insect brain. Annual Review of Entomology, 59(1), 165–184. doi:10.1146/annurev-ento-011613-162031.Find this resource:

Pick, S., & Strauss, R. (2005). Goal-driven behavioral adaptations in gap-climbing Drosophila. Current Biology, 15, 1473–1478.Find this resource:

Ridgel, A. L., Frazier, S. F., DiCaprio, R. A., & Zill, S. N. (1999). Active signalling of leg Loading and unloading in the cockroach. Journal of Neurophysiology, 81, 1432–1437.Find this resource:

Ridgel, A. L., & Ritzmann, R. E. (2005). Effects of neck and circumoesophageal connective lesions on posture and locomotion in the cockroach. Journal of Comparative Physiology A, 191(6), 559–573.Find this resource:

Ritzmann, R. E., Quinn, R. D., & Fischer, M. S. (2004). Convergent evolution and locomotion through complex terrain by insects, vertebrates and robots. Arthropod Structure and Development, 33, 361–379.Find this resource:

Ritzmann, R. E., & Büschges, A. (2007). Insect walking: From reduced preparations to natural terrain. In G. North & R. J. Greenspan (Eds.), Invertebrate neurobiology (pp. 229–250). Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.Find this resource:

Ritzmann, R. E., Ridgel, A. L., & Pollack, A. J. (2008). Multi-unit recording of antennal mechanosensitive units in the central complex of the cockroach, Blaberus discoidalis. Journal of Comparative Physiology A, 194, 341–360.Find this resource:

Ritzmann, R. E., Harley, C. M., Daltorio, K. A., Tietz, B. R., Pollack, A. J., Bender, J. A. & Quinn, R. D. (2012). Deciding which way to go: How do insects alter movements to negotiate barriers? Frontiers in Neuroscience, 6, 97. doi:org/10.3389/fnins.2012.00097.Find this resource:

Rosenbaum, P., J. Schmitz, J. Schmidt, & A. Büschges (2015). Task-dependent modification of leg motor neuron synaptic input underlying changes in walking direction and walking speed. Journal of Neuorophysiology, 114(2), 1090–1101. doi:10.1152/jn.00006.2015Find this resource:

Schutz, C., & Durr, V. (2011). Active tactile exploration for adaptive locomotion in the stick insect. Proceedings of the National Academy of Sciences B, 366(1581), 2996–3005. doi:10.1098/rstb.2011.0126Find this resource:

Staudacher, E. M., Gebhardt, M., & Durr, V. (2005). Antennal movements and mechanoreception: Neurobiology of active tactile sensors. Advances in Insect Physiology, 32, 49–205.Find this resource:

Strausfeld, N. J. (1999). A brain region in insects that supervises walking. Progress in Brain Research, 123, 273–284.Find this resource:

Strausfeld, N. J. (2012). Arthropod brains: Evolution, functional elegance and historical significance. Cambridge, MA: Belknap Press of Harvard University Press.Find this resource:

Strauss, R., Hanesch, U., Kinkelin, M., Wolf, R., & Heisenberg, M. (1992). No-bridge of Drosophila melanogaster: Portrait of a structural brain mutant of the central complex. Journal of Neurogenetics, 8, 125–155.Find this resource:

Strauss, R. (2002). The central complex and the genetic dissection of locomotor behaviour. Current Opinion in Neurobiology, 12, 633–638.Find this resource:

Szczecinski, N. S., Brown, A. E., Bender, J. A., Quinn, R. D., & Ritzmann, R. E. (2014). A neuromechanical simulation of insect walking and transition to turning of the cockroach Blaberus discoidalis. Biological Cybernetics, 108, 1–21.Find this resource:

Tryba, A. K., & Ritzmann, R. E. (2000). Multi-joint coordination during walking and foothold searching in the Blaberus cockroach. I. Kinematics and electromyograms. Journal of Neurophysiology, 83, 3323–3336.Find this resource:

von Buddenbrock, W. (1921). Der Rhythmus der Schreitbewegungen der Stabheuschrecke Dyxippus. Biol. Zentralbl, 41, 41–48.Find this resource:

von Uckermann, G., & Büschges, A. (2009). Premotor interneurons in the local control of stepping motor output for the stick insect single middle leg. Journal of Neurophysiology, 102(3), 1956–1975.Find this resource:

Watson, J. T., & Ritzmann, R. E. (1998a). Leg kinematics and muscle activity during treadmill running in the cockroach, Blaberus discoidalis: II. Fast running. Journal of Comparative Physiology A, 182, 23–33.Find this resource:

Watson, J. T., & Ritzmann, R. E. (1998b). Leg kinematics and muscle activity during treadmill running in the cockroach, Blaberus discoidalis: I. Slow running. Journal of Comparative Physiology A, 182, 11–22.Find this resource:

Watson, J. T., Ritzmann, R. E., Zill, S. N., & Pollack, A. J. (2002). Control of obstacle climbing in the cockroach, Blaberus discoidalis I. Kinematics. Journal of Comparative Physiology A, 188, 39–53.Find this resource:

Wendler, G. (1966). The co-ordination of walking movements in arthropods. In Nervous and hormonal mechanisms of integration. Symposium for the Society of Experimental Biology, 20, 229–249.Find this resource:

Wilson, D. M. (1966). Insect walking. Annual Review of Entomology, 11, 103–123.Find this resource:

Wong, R. K., & Pearson, K. G. (1976). Properties of the trochanteral hair plate and its function in the control of walking in the cockroach. Journal of Experimental Biology, 64(1), 233–249.Find this resource:

Wosnitza, A., Bockemühl, T., Dübert, M., Scholz, H., & Büschges, A. (2013). Inter-leg coordination in the control of walking speed in Drosophila. Journal of Experimental Biology, 216(3), 480–491. doi:10.1242/jeb.078139Find this resource:

Zill, S., Chaudhry, S., Schmitz, J., & Büschges, A. (2012). Force encoding in stick insect legs delineates a reference frame for motor control. Journal of Neurophysiology, 108, 1453–1472.Find this resource:

Zill, S., Keller, B., & Duke, E. (2009). Sensory signals of unloading in one leg follow stance onset in another leg: Transfer of load and emergent coordination in cockroach walking. Journal of Neurophysiology, 101, 2297–2304.Find this resource:

Zill, S. N. (1990). Mechanoreceptors: Exteroceptors and proprioceptors. In I. Huber, E. P. Masler, & B. R. Rao (Eds.), Cockroaches as models for neurobiology: Applications in biomedical research (pp. 247–267). Boca Raton, FL: CRC Press.Find this resource:

Zill, S. N., Chaudhry, S., Büschges, A., & Schmitz, J. (2015). Force feedback reinforces muscle synergies in insect legs. Arthropod Structure and Development, 44(6), 541–553.Find this resource:

Zill, S. N., Schmitz, J., & Büschges, A. (2004). Load sensing and control of posture and locomotion. Arthropod Structure and Development, 33(3), 273–286.Find this resource:

Zill, S. N., & Seyfarth, E. A. (1996). Exoskeletal sensors for walking. Scientific American, 275(1), 86–90.Find this resource: