On Heels and Toes: How Ants Climb with Adhesive Pads and Tarsal Friction Hair Arrays

Ants are able to climb effortlessly on vertical and inverted smooth surfaces. When climbing, their feet touch the substrate not only with their pretarsal adhesive pads but also with dense arrays of fine hairs on the ventral side of the 3^rd and 4^th tarsal segments. To understand what role these different attachment structures play during locomotion, we analysed leg kinematics and recorded single-leg ground reaction forces in Weaver ants (Oecophylla smaragdina) climbing vertically on a smooth glass substrate. We found that the ants engaged different attachment structures depending on whether their feet were above or below their Centre of Mass (CoM). Legs above the CoM pulled and engaged the arolia (‘toes’), whereas legs below the CoM pushed with the 3^rd and 4^th tarsomeres (‘heels’) in surface contact. Legs above the CoM carried a significantly larger proportion of the body weight than legs below the CoM. Force measurements on individual ant tarsi showed that friction increased with normal load as a result of the bending and increasing side contact of the tarsal hairs. On a rough sandpaper substrate, the tarsal hairs generated higher friction forces in the pushing than in the pulling direction, whereas the reverse effect was found on the smooth substrate. When the tarsal hairs were pushed, buckling was observed for forces exceeding the shear forces found in climbing ants. Adhesion forces were small but not negligible, and higher on the smooth substrate. Our results indicate that the dense tarsal hair arrays produce friction forces when pressed against the substrate, and help the ants to push outwards during horizontal and vertical walking.     

Functionally Different Pads on the Same Foot Allow Control of Attachment: Stick Insects Have Load-Sensitive “Heel” Pads for Friction and Shear-Sensitive “Toe” Pads for Adhesion

Stick insects (Carausius morosus) have two distinct types of attachment pad per leg, tarsal “heel” pads (euplantulae) and a pre-tarsal “toe” pad (arolium). Here we show that these two pad types are specialised for fundamentally different functions. When standing upright, stick insects rested on their proximal euplantulae, while arolia were the only pads in surface contact when hanging upside down. Single-pad force measurements showed that the adhesion of euplantulae was extremely small, but friction forces strongly increased with normal load and coefficients of friction were 1. The pre-tarsal arolium, in contrast, generated adhesion that strongly increased with pulling forces, allowing adhesion to be activated and deactivated by shear forces, which can be produced actively, or passively as a result of the insects'' sprawled posture. The shear-sensitivity of the arolium was present even when corrected for contact area, and was independent of normal preloads covering nearly an order of magnitude. Attachment of both heel and toe pads is thus activated partly by the forces that arise passively in the situations in which they are used by the insects, ensuring safe attachment. Our results suggest that stick insect euplantulae are specialised “friction pads” that produce traction when pressed against the substrate, while arolia are “true” adhesive pads that stick to the substrate when activated by pulling forces.     

Friction ridges in cockroach climbing pads: anisotropy of shear stress measured on transparent,microstructured substrates

The contact of adhesive structures to rough surfaces has been difficult to investigate as rough surfaces are usually irregular and opaque. Here we use transparent, microstructured surfaces to investigate the performance of tarsal euplantulae in cockroaches (Nauphoeta cinerea). These pads are mainly used for generating pushing forces away from the body. Despite this biological function, shear stress (force per unit area) measurements in immobilized pads showed no significant difference between pushing and pulling on smooth surfaces and on 1-μm high microstructured substrates, where pads made full contact. In contrast, on 4-μm high microstructured substrates, where pads made contact only to the top of the microstructures, shear stress was maximal during a push. This specific direction dependence is explained by the interlocking of the microstructures with nanometre-sized “friction ridges” on the euplantulae. Scanning electron microscopy and atomic force microscopy revealed that these ridges are anisotropic, with steep slopes facing distally and shallow slopes proximally. The absence of a significant direction dependence on smooth and 1-μm high microstructured surfaces suggests the effect of interlocking is masked by the stronger influence of adhesion on friction, which acts equally in both directions. Our findings show that cockroach euplantulae generate friction using both interlocking and adhesion.     

Evolution of locomotory attachment pads in the Dermaptera (Insecta)

This contribution is the first comparative SEM study of tarsal and pretarsal structures of 18 dermapteran species, including epizoic Hemimeridae, rare Apachyidae, as well as basal Pygidicranidae. Our data reject the apparent uniformity of this taxon and show that representatives of Dermaptera have independently evolved both types of attachment mechanisms: hairy and smooth. Dermaptera possess a wide spectrum of attachment devices: arolia, euplantulae, tarsal surfaces covered with specialised tenent setae and other types of cuticular outgrowths. The groundpattern of the pretarsal and tarsal attachment structures was reconstructed by mapping their characters onto a cladogram, generated without tarsal characters. In the groundpattern of recent Dermaptera, the tarsus consists of three tarsomeres. Presumably, the last common ancestor of the Dermaptera possessed an arolium, since this structure occurs in the most basal taxa: Diplatyidae, Karschiellidae (partim, adults), Pygidicranidae partim, and Apachyidae. The absence of arolium in two of the pygidicranid taxa is probably due to a secondary loss. The arolium seems to be reduced in the 'higher Dermaptera' and amongst them, only the Geracinae, which belong to the Spongiphoridae and, hence, to the well supported Eudermaptera [European Journal of Entomology, 98 (2001), 445], evolved this structure convergently. The character state distribution for euplantulae suggests their evolution being similar to that of the arolium. All species of Tagalina possess a specialised tarsus with a strongly dilated second tarsomere. The same applies to the Forficulidae. However, their relatively remote phylogenetic position to Tagalina burri is a convincing reason to assume convergent evolution of this character. The Chelisochidae, with a slender, elongated second tarsomere, possess a unique structure, which supports their monophyly. The special, heart shaped structure of the second tarsal segments in the Forficulidae suggests their monophyly. The attachment structures of Hemimerus vosseleri are highly derived and probably autapomorphic for this taxon.     

Detecting substrate engagement: responses of tarsal campaniform sensilla in cockroaches

Sensory signals of contact and engagement with the substrate are important in the control and adaptation of posture and locomotion. We characterized responses of campaniform sensilla, receptors that encode forces as cuticular strains, in the tarsi (feet) of cockroaches using neurophysiological techniques and digital imaging. A campaniform sensillum on the fourth tarsal segment was readily identified by its large action potential in nerve recordings. The receptor discharged to contractions of the retractor unguis muscle, which engages the pretarsus (claws and arolium) with the substrate. We mimicked the effects of muscle contractions by applying displacements to the retractor apodeme (tendon). Sensillum firing did not occur to unopposed movements, but followed engagement of the claws with an object. Vector analysis of forces suggested that resisted muscle contractions produce counterforces that axially compress the tarsal segments. Close joint packing of tarsal segments was clearly observed following claw engagement. Physiological experiments showed that the sensillum responded vigorously to axial forces applied directly to the distal tarsus. Discharges of tarsal campaniform sensilla could effectively signal active substrate engagement when the pretarsal claws and arolium are used to grip the substrate in climbing, traversing irregular terrains or walking on inverted surfaces.     

Friction and adhesion in the tarsal and metatarsal scopulae of spiders

Friction and adhesion forces of the ventral surface of tarsi and metatarsi were measured in the bird spider Aphonopelma seemanni (Theraphosidae) and the hunting spider Cupiennius salei (Ctenidae). Adhesion measurements revealed no detectable attractive forces when the ventral surfaces of the leg segments were loaded and unloaded against the flat smooth glass surface. Strong friction anisotropy was observed: friction was considerably higher during sliding in the distal direction. Such anisotropy is explained by an anisotropic arrangement of microtrichia on setae: only the setal surface facing in the distal direction of the leg is covered by the microtrichia with spatula-like tips. When the leg is pushed, the spatula-shaped tips of microtrichia contact the substrate, whereas, when the leg is pulled over a surface, setae bend in the opposite direction and contact the substrate with their spatulae-lacking sides. In an additional series of experiments, it was shown that desiccation has an effect on the friction force. Presumably, drying of the legs results in reduction of the flexibility of the setae, microtrichia, spatulae, and underlying cuticle; this diminishes the ability to establish proper contact with the substrate and thus reduces the contact forces.     

Adhesive properties of the arolium of a lantern-fly, Lycorma delicatula (Auchenorrhyncha, Fulgoridae)

The arolium in Lycorma delicatula is shaped as a truncated pyramid, tapering proximally. The base or the terminal area is corrugated, forming parasagittal wrinkles (period 1.5-5.0 microm), which are supported from inside by cuticular dendrites. Side faces of the arolium are made up of sclerotized dorsolateral plates. When claws slip on a smooth substrate and pronate, the dorsolateral plates diverge and expand the sticky terminal area. The real contact area with the glass plate was recognized by light reflection on its periphery. This area was measured and shown to be smaller when the leg was pressed perpendicularly to the substrate (0.02 mm(2)) than when it was sheared in a direction parallel to the substrate (0.05 mm(2)). Attachment forces were measured with the aid of dynamometric platforms during pulling of active insects from horizontal or vertical glass surfaces. Normal adhesive force (about 9-12 mN) was much less than friction force during sliding with velocity of 6-17 mm/s (50-100 mN); however, when expressed in tenacity per unit contact area the difference was less pronounced: 170 and 375-625 mN/mm(2), respectively. Sliding of the arolium during shear displacement was shown to be oscillatory in frame-by-frame video analysis. Relaxative oscillations consisted of periodical sticks-slips of the arolium along the glass surface.     

The earliest Timematids in Burmese amber reveal diverse tarsal pads of stick insects in the mid‐Cretaceous

Many extant insects have developed pad structures, euplantulae or arolia on their tarsi to increase friction or enhance adhesion for better mobility. Many polyneopteran insects with euplantulae, for example, Grylloblattodea, Mantophasmatodea and Orthoptera, have been described from the Mesozoic. However, the origin and evolution of stick insects' euplantulae are poorly understood due to rare fossil records. Here, we report the earliest fossil records of Timematodea hitherto, Tumefactipes prolongates gen. et sp. nov. and Granosicorpes Urates gen. et sp. nov., based on three specimens from mid-Cretaceous Burmese amber. Specimens of Tumefactipes prolongates gen. et sp. nov. have extremely specialized and expanded euplantulae on their tarsomere II. These new findings are the first known and the earliest fossil records about euplantula structure within Phasmatodea, demonstrating the diversity of euplantulae in Polyneoptera during the Mesozoic. Such tarsal pads might have increased friction and helped these mid-Cretaceous stick insects to climb more firmly on various surfaces, such as broad leaves, wetted tree branches or ground. These specimens provide more morphological data for us to understand the relationships of Timematodea, Euphasmatodea, Orthoptera and Embioptera, suggesting that Timematodea might be monophyletic with Euphasmatodea rather than Embioptera and Phasmatodea should have a closer relationship with Orthoptera rather than Embioptera.     

Giant stick insects reveal unique ontogenetic changes in biological attachment devices

A strong modification of tarsal and pretarsal attachment pads during the postembryonic development is described for the first time. In the exceptionally large thorny devil stick insect Eurycantha calcarata a functional arolium is only present in the immature instars, enabling them to climb on smooth surfaces, especially leaves. Nymphs are also characterized by greyish and hairy euplantulae on tarsomeres 1–4. The gradual modifications of the arolium and the euplantula of tarsomere 5 in the nymphal development are probably mainly related to increased weight. The distinct switch in the life style between the leaf-dwelling nymphal stages and the ground-dwelling adults results in the final abrupt change of the adhesive devices, resulting in a far-reaching reduction of the arolium, the presence of a fully-developed, elongated euplantula on tarsomere 5, and white and smooth euplantulae on tarsomeres 1–4. The developmental remodelling of attachment pads also reflects a phylogenetic pattern. The attachment devices of the earlier instars are similar to those found in the basalmost lineage of extant stick insects, Timema, which is characterized by a very large pan-shaped arolium and a hairy surface of the tarsal and pretarsal attachment pads.     

Distal leg structures of the Aculeata (Hymenoptera): A comparative evolutionary study of Sceliphron (Sphecidae) and Formica (Formicidae)

The distal parts of the legs of Sceliphron caementarium (Sphecidae) and Formica rufa (Formicidae) are documented and discussed with respect to phylogenetic and functional aspects. The prolegs of Hymenoptera offer an array of evolutionary novelties, mainly linked with two functional syndromes, walking efficiently on different substrates and cleaning the body surface. The protibial-probasitarsomeral cleaning device is almost always well-developed. A complex evolutionary innovation is a triple set of tarsal and pretarsal attachment devices, including tarsal plantulae, probasitarsomeral spatulate setae, and an arolium with an internal spring-like arcus, a dorsal manubrium, and a ventral planta. The probasitarsal adhesive sole and a complex arolium are almost always preserved, whereas the plantulae are often missing. Sceliphron has retained most hymenopteran ground plan features of the legs, and also Formica, even though the adhesive apparatus of Formicidae shows some modifications, likely linked to ground-oriented habits of most ants. Plantulae are always absent in extant ants, and the arolium is often reduced in size, and sometimes vestigial. The arolium contains resilin in both examined species. Additionally, resilin enriched regions are also present in the antenna cleaners of both species, although they differ in which of the involved structures is more flexible, the calcar in Sceliphron and the basitarsal comb in Formica. Functionally, the hymenopteran distal leg combines (a) interlocking mechanisms (claws, spine-like setae) and (b) adhesion mechanisms (plantulae, arolium). On rough substrate, claws and spine-like setae interlock with asperities and secure a firm grip, whereas the unfolding arolium generates adhesive contact on smooth surfaces. Differences of the folded arolium of Sceliphron and Formica probably correlate with differences in the mechanism of folding/unfolding.     

Walking on smooth or rough ground: passive control of pretarsal attachment in ants

The hymenopteran tarsus is equipped with claws and a movable adhesive pad (arolium). Even though both organs are specialised for substrates of different roughness, they are moved by the same muscle, the claw flexor. Here we show that despite this seemingly unfavourable design, the use of arolium and claws can be adjusted according to surface roughness by mechanical control. Tendon pull experiments in ants (Oecophylla smaragdina) revealed that the claw flexor elicits rotary movements around several (pre-) tarsal joints. However, maximum angular change of claws, arolium and fifth tarsomere occurred at different pulling amplitudes, with arolium extension always being the last movement. This effect indicates that arolium use is regulated non-neuronally. Arolium unfolding can be suppressed on rough surfaces, when claw tips interlock and inhibit further contraction of the claw flexor or prevent legs from sliding towards the body. To test whether this hypothesised passive control operates in walking ants, we manipulated ants by clipping claw tips. Consistent with the proposed control mechanism, claw pruning resulted in stronger arolium extension on rough but not on smooth substrates. The control of attachment by the insect claw flexor system demonstrates how mechanical systems in the body periphery can simplify centralised, neuro-muscular feedback control.     

Control of climbing behavior in the cockroach, Blaberus discoidalis. II. Motor activities associated with joint movement

Deathhead cockroaches employ characteristic postural strategies for surmounting barriers. These include rotation of middle legs to re-direct leg extension and drive the animal upward. However, during climbing the excursions of the joints that play major roles in leg extension are not significantly altered from those seen during running movements. To determine if the motor activity associated with these actions is also unchanged, we examined the electromyogram activity produced by the slow trochanteral extensor and slow tibial extensor motor neurons as deathhead cockroaches climbed over obstacles of two different heights. As they climbed, activity in the slow trochanteral extensor produced a lower extension velocity of the coxal-trochanteral joint than the same frequency of slow trochanteral extensor activity produces during horizontal running. Moreover, the pattern of activity within specific leg cycles was altered. During running, the slow trochanteral extensor generates a high-frequency burst prior to foot set-down. This activity declines through the remainder of the stance phase. During climbing, motor neuron frequency no longer decreased after foot set-down, suggesting that reflex adjustments were made. This conclusion was further supported by the observation that front leg amputees generated even stronger slow trochanteral extensor activity in the middle leg during climbing movements.     

Adhesive and frictional properties of tarsal attachment pads in two species of stick insects (Phasmatodea) with smooth and nubby euplantulae

In the present study, the tarsal attachment pads (euplantulae) of two stick insect species (Phasmatodea) were compared. While the euplantulae of Cuniculina impigra (syn. Medauroidea extradentata) are smooth, those of Carausius morosus bear small nubs on their surfaces. In order to characterize the adhesive and frictional properties of both types of euplantulae, adhesion and friction measurements on smooth (Ra=0.054 μm) and rough (Ra=1.399 μm) substrates were carried out. The smooth pads of C. impigra generated stronger adhesion on the smooth substrate than on the rough one. The adhesive forces of the structured pads of C. morosus did not differ between the two substrates. Friction experiments showed anisotropy for both species with higher values for proximal pulls than for distal pushes. In C. impigra, friction was stronger on the smooth than on the rough surface for both directions, whereas in C. morosus friction was stronger on the smooth surface only for pushes. This shows that smooth attachment pads are able to generate relatively stronger adhesion and friction on a flat smooth surface than on a rough one. In contrast, nubby pads have similar adhesion on both substrates, and also show no difference in friction in the pulling direction. This leads to the conclusion that smooth pads are specialized for rather smooth substrates, whereas nubby pads are better adapted to generate stronger forces on a broader range of surfaces.     

Elasticity and movements of the cockroach tarsus in walking

Anatomical, kinematic and ablation studies were performed to evaluate the contribution of elasticity in use of the cockroach tarsus (foot) in walking. The distal tarsus (claws and arolium) engages the substrate during the stance phase of walking by the action of a single muscle, the retractor unguis. Kinematic and ablation studies demonstrated that tarsal disengagement occurs at the end of stance, in part via the action of elastic elements at the penultimate tarsal joint. In isolated legs, this joint exhibits very rapid (less than 20 ms duration) recoil to extension when released from the engaged position, and recoil is even more rapid (less than 10 ms) after removal of the retractor tendon (apodeme). The joint also possesses an enlarged cuticular condyle which is the attachment for ligaments and articular membranes, some of which fulfill morphological criteria consistent with the presence of the elastic protein resilin. Measurements of restoring forces generated by joint displacement indicate that they are graded but could readily lift the mass of the distal tarsus. This biomechanical design can facilitate efficient use of the tarsus in walking while under active control by only a single muscle and may also be highly advantageous when cockroaches very rapidly traverse irregular terrain. Accepted: 16 September 1998     

Locomotion of the Eurasian nuthatch on vertical and horizontal substrates

Legged locomotion of the Eurasian nuthatch Sitta europaea on horizontal and vertical substrates was examined using field observations and experiments. Although previous studies have reported that nuthatches use 'walking' on vertical substrates, we found that they usually used 'hopping' on both vertical and horizontal substrates. When climbing up a vertical substrate, the feet were staggered in position and small phase differences were observed between the left and right leg movements in the gait. In upward climbing, the body was inclined towards the substrate during the first stance phase similar to other tree-trunk climbers, but the tail was not used for helping body rotation unlike most tree-trunk climbers. The staggered position of the feet may allow the legs to play different roles in pulling towards and pushing away from the substrate. In downward climbing, the feet were staggered in position, but the phase difference was quite small. In field observations, the Eurasian nuthatch preferred to move vertically, rather than in an inclined direction.     

Structure and function of the arolium of Mantophasmatodea (Insecta)

All species of the insect order Mantophasmatodea characteristically keep the 5th tarsomere and pretarsus (arolium plus two claws) turned upwards and off the substrate. The unusually large arolium was studied in two species of Mantophasmatodea using bright field light microscopy, reflection microscopy, fluorescence microscopy, TEM, SEM, and Cryo‐SEM. It contains an epithelial gland, numerous tracheoles, and nerves. The gland consists of enlarged epithelial cells with large nuclei, mitochondria, RER, golgi complexes, microtubules, and numerous secretion vesicles. Evidence for exocytosis of the vesicles into the gland reservoir between the epithelial gland and the thick cuticle could be observed. Cryo‐SEM revealed that the ventral side of the arolium and distal part of its dorsal side are covered with a liquid film. Fluid footprints of arolia of individuals walking on a glass plate also indicate the presence of secretory fluid on the arolium surface. Behavioral experiments using animals with ablated arolia showed that representatives of Mantophasmatodea do not need their arolia to detect and respond to vibratory communication signals nor to catch small to medium‐sized prey. Individuals with ablated arolia were not able to move upside down on a smooth glass plate. We conclude that Mantophasmatodea use their arolia for attachment when additional adhesion force is required (e.g. windy conditions, handling large prey, mating). They can bring their arolia in contact with the surface in a very fast reflex (18.0 ± 9.9 ms). The secretory fluid found on the surface is produced by the gland and transported to the outside, presumably through small pore channels, to enhance adhesion to the substrate. J. Morphol., 2009. © 2009 Wiley‐Liss, Inc.     

Structure and mechanics of the tarsal chain in the hornet, Vespa crabro (Hymenoptera: Vespidae): implications on the attachment mechanism

Two combined mechanisms on the hornet tarsus are adapted to attachment to the substrate: a friction-based (claws and spines) and an adhesion-based one (arolium). There are two ranges of substrate roughness optimal for attachment, either very smooth or very rough. There is an intermediate range of substrate grains of small but non-zero size, where both of these mechanisms fail. The optimal size of substrate grains for hornet grasping was 50-100 microm. Maximal hold to the substrate was achieved when surface irregularities were clamped between the claws of opposite legs. In such a position, the insect could withstand an external force which was almost 25 times larger than its own weight. The tarsal chain is an important part of the entire attachment mechanism. The articulations in the kinematic chain of tibia-tarsus-pretarsus are monocondylar. Three tarsal muscles and one head of the claw retractor muscle originate in the tibia. On pull to the retractor tendon, the tarsus bends in a plane. All elements of the tarsal kinematic chain have one active degree of freedom. The distance between the intertarsomeric articulation point and the tendon of the claw retractor (75-194 microm) corresponds to an efficiency of 1 degrees per 1-3 mircom of pulling distance travelled by the tendon. The claw turns about 1 degrees per 4.3-5.0 microm of pulling distance travelled by the unguitractor. The arolium turns forward and downward simultaneously with flexion of the claws. The kinematic chain of the arolium lacks real condylar joints except the joint at the base of the manubrium. Other components are tied by flexible transmissions of the membranous cuticle. The walking hornet rests on distal tarsomeres of extended tarsi. If the retractor tendon inside the tarsus is fixed, passive extension of the tarsomeres might be replaced by claw flexion. Tarsal chain rigidity, measured with the force tester, increased when the retractor tendon was tightened. Probably, pull to the tendon compresses the tarsomeres, increasing friction within contacting areas of rippled surfaces surrounding condyles within articulations.     

Ground reaction forces in vertically ascending beetles and corresponding activity of the claw retractor muscle on smooth and rough substrates

We measured ground reaction forces in fore–aft and normal directions of single hind and front legs in vertically ascending Sagra femorata beetles (Coleoptera, Chrysomelidae) on a smooth and a rough substrate. Simultaneously, we performed electromyographic recordings (EMGs) of the hind leg claw retractor muscle that partly controls the attachment structures. On both substrates, hind legs produced upward- as well as downward-directed forces during one stance phase. Forces were equivalent in both directions. Front legs generated only upward-directed forces. The main function of hind legs in ascending beetles in the second half of the stance thus probably prevented the animals from tilting away from the substrate. The EMGs of hind legs showed an early spike during stance with large amplitude. It was mostly followed by few additional spikes with large amplitude and in some cases of spikes with smaller amplitude distributed throughout the stance phase. We found significantly more spikes on the rough substrate than on the smooth one. This is probably due to the more important role of pretarsal claws than tarsal hairy attachment pads on the rough substrate or to the reduced adhesive forces on the rough substrate that have to be compensated by additional muscle activity.     

Avoidance reflexes mediated by contact chemoreceptors on the legs of locusts

Taste receptors, or basiconic sensilla, are distributed over the legs of the locust and respond to direct contact with chemical stimulants. The same chemosensory neurones that responded to contact with salt solutions also responded to particular acidic odours. Odours of food and other chemicals had no effect on the chemosensory neurones. In locusts free to move, an acid odour presented to the tarsus of a hind leg evoked a rapid avoidance movement in which the tarsus was levated, the tibia flexed and the femur levated. Intracellular recordings from motor neurones that innervate muscles of the hind leg showed that when an acid odour was directed towards basiconic sensilla on the leg there was a reciprocal activation of antagonistic motor pools that move the leg segments about each joint. Thus an extensor tibiae motor neurone was inhibited while a flexor tibiae motor neurone was excited, and the tarsal depressor and retractor unguis motor neurones were inhibited while the tarsal levator motor neurone was excited. This method of odour stimulation of taste receptors generates less adaptation than direct contact with chemicals, and therefore represents an ideal method for stimulating taste receptors for further studies on the central pathways processing taste signals. Accepted: 2 June 1998     

Control of obstacle climbing in the cockroach, Blaberus discoidalis. I. Kinematics

An advantage of legged locomotion is the ability to climb over obstacles. We studied deathhead cockroaches as they climbed over plastic blocks in order to characterize the leg movements associated with climbing. Movements were recorded as animals surmounted 5.5-mm or 11-mm obstacles. The smaller obstacles were scaled with little change in running movements. The higher obstacles required altered gaits, leg positions and body posture. The most frequent sequence used was to first tilt the front of the body upward in a rearing stage, and then elevate the center of mass to the level of the top of the block. A horizontal running posture was re-assumed in a leveling-off stage. The action of the middle legs was redirected by rotations of the leg at the thoracal-coxal and the trochanteral-femoral joints. The subsequent extension movements of the coxal-trochanteral and femoral-tibial joints were within the range seen during horizontal running. The structure of proximal leg joints allows for flexibility in leg use by generating subtle, but effective changes in the direction of leg movement. This architecture, along with the resulting re-direction of movements, provides a range of strategies for both animals and walking machines.