The search for stability on narrow supports: an experimental study in cats and dogs

Abstract

Kinematic and coordination variables were studied in two carnivorans, one with known locomotor capabilities in arboreal substrates (cat), and the other a completely terrestrial species (dog). Two horizontal substrates were used: a flat trackway on the ground (overground locomotion) and an elevated and narrow runway (narrow-support locomotion). Despite their different degree of familiarity with the ‘arboreal’ situation, both species developed a strategy to adapt to narrow supports. The strategy of cats was based on using slower speeds, coupled with modifications to swing phase duration, to keep balance on narrow supports. The strategy of dogs relied on high speeds to gain in dynamic stability, and they increased cycle frequency by reducing swing phase duration. Furthermore, dogs showed a high variability in limb coordination, although a tendency to canter-like coordination was observed, and also avoided whole-body aerial phases. In different ways, both strategies suggested a reduction of peak vertical forces, and hence a reduction of the vertical oscillations of the centre of mass. Finally, lateral oscillation was reduced by the use of a crouched posture.

Introduction

The gaits employed by animals when walking or running overground, and their corresponding dynamics and kinematics, have been rigorously studied since the 19th century (e.g., Marey, 1873, Muybridge, 1899, Manter, 1938, Hildebrand, 1966, Hildebrand, 1980, Hildebrand, 1985, Demes et al., 1994, Lee et al., 1999, Larson et al., 2000, Cartmill et al., 2002, Fischer et al., 2002, Abourachid, 2003, Hutchinson et al., 2006, Maes et al., 2008). Nevertheless, the ground is not the only support on which animals move; they also move on branches high in the forest canopy or dig their way through the ground. Each substrate requires different anatomical, morphological, and mechanical adaptations, as well as modifications to the dynamics and kinematics of locomotion (Biewener, 2003).

Locomotion on arboreal substrates has not been as thoroughly studied as overground locomotion, but its main particularities have already been covered (Cartmill, 1974, Cartmill, 1985, Meldrum, 1991, Schmitt, 1999, Schmitt, 2003a, Schmitt and Lemelin, 2002, Lemelin et al., 2003, Lammers and Biknevicius, 2004, Schmidt and Fischer, 2010). The main problem affecting arboreal locomotion is the tendency of animals to roll (rotate around their sagittal axis) and topple from the support because all their support points are effectively collinear, which greatly reduces their support polygon. Several solutions to this problem, each involving different morphological adaptations, have been described (Cartmill, 1985): (i) relatively short limbs, as in arboreal viverrids (Taylor, 1970), or the use of a crouched posture (Schmidt and Fischer, 2010), keep the body's centre of mass close to the support and minimise lateral oscillation; (ii) prehensile hands and/or feet allow gripping the branch and thus exerting a torque that resists the toppling moment, as in primates (Rollinson and Martin, 1981, Vilensky and Larson, 1989, Schmitt, 1999), some opossums (Schmitt and Lemelin, 2002, Lemelin et al., 2003), and tupaiids (Sargis, 2001); (iii) the reduced body size of small animals, like squirrels, overcomes the toppling problem by spreading the support points relatively more widely on the surface of the branch; and (iv) a foolproof solution to totally avoid toppling is hanging underneath the branch, like sloths do. Another source of locomotor instability during arboreal locomotion is the round section of branches, which increases the potential of slipping off them. Animals with prehensile hands and/or feet avoid this problem by firmly grasping the support; while clawed animals, whose grasping abilities are reduced or absent, change limb placement during arboreal locomotion to reorient substrate reaction forces inwards to the branch, and thus prevent slipping off it (Schmitt, 2003a, Lammers and Biknevicius, 2004, Schmidt and Fischer, 2010). Finally, another problem affecting arboreal locomotion are vertical oscillations of the support. Branches, especially the fine ones, tend to deflect under an animal's weight, which not only hinders joint stabilisation, but might also toss the animal from the support due to elastic recovery. Schmitt (1999) proposed that animals use compliant gaits as a solution to this problem. Compliant gaits are characterised by substantial limb yield, which reduces vertical oscillations of the body (and thus of the support) and encourages long contact times, which in turn allows the reduction of stride frequency (and thus the potential of branch sway). Furthermore, compliant gaits reduce bone and joint stresses associated with flexed-limb gaits (Schmitt, 1999). The use of compliant gaits in primates, marsupials and other arboreal mammals was later confirmed by Larney and Larson (2004). In addition to compliant gaits, the use of a crouched posture has also been proposed as a mechanism to reduce vertical oscillations of the body both in compliant (Schmitt, 1999) and stiff gaits (i.e., when limb yield is low; Bishop et al., 2008). For the latter case, the authors proposed that, if limb protraction and angular excursion remained unaltered, the use of a crouched posture would reduce vertical displacement of the centre of mass by creating a smaller pendulum (and thus reducing potential energy fluctuations; Bishop et al., 2008). Finally, at higher speeds, ambling gaits have also been proposed as a solution to reduce vertical oscillations of the support, since they allow animals to maintain at least one foot in contact with the substrate during a stride, thus reducing peak vertical forces on the support (Schmitt et al., 2006).

Most studies on arboreal locomotion, though, focus on primates and, to a lesser extent, on some didelphids, since they consider these groups arboreal specialists, which present a set of adaptations to moving and foraging in an arboreal setting so marked that it makes their terrestrial locomotion distinct from that of other animals. These adaptations involve prehensile extremities, showing more protracted arm postures at touch-down, producing lower peak vertical substrate reaction forces with the forelimbs than with the hindlimbs, and using diagonal-sequence gaits almost exclusively when walking on narrow supports (Hildebrand, 1967, Vilensky and Larson, 1989, Demes et al., 1994, Larson et al., 2000, Schmitt and Lemelin, 2002). Nevertheless, arboreal specialists are not the only animals known to use arboreal substrates. As stated by Lammers and Biknevicius (2004), many small mammals use fallen logs and branches on the forest floor as arboreal runways. Furthermore, many terrestrial species often climb trees to escape predators or while hunting (MacDonald, 1984, Wilson and Mittermeier, 2009). Since stability in locomotion is directly linked to performance in escaping or hunting behaviours, and thus directly linked to fitness, it would be vital for these terrestrial mammals navigating arboreal substrates (non-arboreal specialists) to adapt their locomotion and increase their stability.

To date, locomotion on arboreal supports in non-arboreal specialists has only been studied in small species: the common marmoset (Callithrix jaccus) (Schmitt, 2003b), the grey short-tailed opossum (Monodelphis domestica) (Lammers and Biknevicius, 2004), and the rat (Rattus norvegicus) (Schmidt and Fischer, 2010). To increase their stability on arboreal supports, these animals reduced peak vertical forces to reduce the vertical oscillation of the centre of mass. Both the common marmoset and the rat used similar speeds and had similar contact times (i.e., duty factor, and thus stance phase duration) in overground and arboreal locomotion, while the grey short-tailed opossum used lower speeds and had longer contact times during arboreal locomotion. Schmidt and Fischer (2010) proposed that the reduction of speed could only be accomplished if some grasping ability is retained.

In the light of these results, we wonder how a larger non-arboreal specialist (for instance, a ground-dwelling carnivoran pursuing its prey up into the forest canopy) will adapt its kinematics and coordination to the arboreal substrate. Will larger mammals use the same strategy as the smaller ones? The first aim of this study was thus to determine how a medium-sized non-arboreal specialist adjusts its kinematics and coordination to adapt to an arboreal substrate. For our experiments, we chose the domestic cat (Felis silvestris catus), which is accustomed to moving comfortably along branches, rails, and similar narrow, elevated supports. Taking into account the possible solutions for increasing stability presented above, cats were expected to increase stance phase duration, and thus decrease stride frequency. Slower speeds on narrow supports than on flat ground, as was found for the grey short-tailed opossum (Lammers and Biknevicius, 2004), were also expected, since cats can use their claws to grip the support. We also expected that they would display a more crouched posture in the ‘arboreal’ situation to bring the centre of mass closer to the support.

Secondly, we wondered whether the strategy employed by non-arboreal specialists to adapt to the arboreal situation, if there was any, would be a universal solution for all terrestrial species. That is, if we encouraged a completely terrestrial species into an arboreal-like situation, would it arrive at the same solution to keep balance and advance on the narrow support? To answer this question, we used a protocol similar to the one used for cats to study the kinematics and coordination of the domestic dog (Canis lupus familiaris) when moving along a narrow, elevated support, before comparing both strategies. We chose the dog because it is a completely terrestrial species whose kinematics and coordination overground have already been thoroughly studied (Hildebrand, 1968, Lee et al., 1999, Maes et al., 2008).