To date some level of lateralisation in forelimb use is known for a substantial number of vertebrate species from fish to mammals [1–5]. This kind of lateralization is a result of brain interhemispheric asymmetry due to differential motor activity or information processing (reviewed in [3, 6, 7]. Among mammals only primates historically received considerable attention and have become the subjects of detailed investigation (reviewed in [3, 8, 9]). At the same time, there is a growing body of evidence showing manual laterality in other mammalian groups. For instance, limb preferences at individual and/or group levels in various species-typical motor activities have been described in rodents (Mongolian gerbils , mice and rats [11–14]), carnivores (black bears , domestic cats [16–20], and dogs [21–24]), ungulates (horses [25–28], plains zebras and impalas , donkeys ), bats (Schreiber’s long-fingered bat ) and even in marine mammals such as pinnipeds (walruses ) and cetaceans (humpback whales , bottlenose dolphins , and Commerson’s dolphins ). Thus, some kind of phylogenetic continuity in the evolution of motor lateralization in mammals could be traced . Meanwhile, the more species were studied, the more was becoming known about the specific factors shaping the expression of laterality in limb use. Animal’s posture was considered to be one of these factors.
In primates the upright posture was determined as a factor facilitating manual laterality irrespective of the bias direction: individual hand preferences significantly increased when experimental subjects had to reach a food item from a bipedal compared to a quadrupedal position (reviewed in [8, 37–39]). Notably, this shift was observed not only during the comparison of the same individuals in different body postures, but also at the interspecific level. In prosimians a more vertical body orientation in a species was shown to be associated with a stronger laterality in hand use [8, 40].
The species postural characteristics were proposed to influence not only the strength but also the very presence of the population level hand preference: in contrast to more upright and large-bodied species, small-bodied, quadrupedal prosimians showed no population level handedness irrespective to the subjects’ posture [39, 41]. The same could be applied to some ape species. The bipedal locomotion is more typical for gorillas and gibbons than for orangutans; and indeed gorillas and gibbons are more liable to group-level handedness . The latter author has hypothesised that readiness to exhibit a unilateral hand preference at the population level correlates with the degree of bipedality in a species. Furthermore, the best documented and most obvious case of behavioural laterality is pronounced handedness in humans ― the most bipedal primate [7, 43]. Some authors argued that bipedalism may have facilitated species-typical right-handedness in humans . This hypothesis was supported by findings in human infants, in which postural changes during early development are associated with establishing of stable handedness (reviewed in ). Before the age of three, infants display fluctuating patterns of manual preferences shifting together with the development of new forms of locomotion. Notably, on the stage of crawling on hands-and-knees (which in fact is a quadrupedal gait) infants exhibit no stable patterns of hand preferences, while the establishment of the latter follows closely the adoption of upright posture and bipedal locomotion.
Recently, we have explored manual laterality in a bipedal hopping marsupial: red-necked wallabies (Macropus rufogriseus, Diprotodontia), during their usual daily activity in zoo conditions . In this work wallabies were shown to display group-level preferences to use their left forelimb in feeding from the bipedal position and to lean on the right paw in the tripedal stance. Left-forelimb bias was also traced in unimanual autogrooming. The young wallabies displayed forelimb specializations resembling that in adults: they more often used their left forelimb for pulling down the mother’s pouch and simultaneously supported the body with the right forelimb during milk suckling. In contrast, in feeding from the quadrupedal position no group-level bias was found and only a few wallabies showed individual preferences. Thus, unimanual actions performed from upright posture were more suited to reveal individual and population forelimb preferences. These results led us to a conclusion that the bipedal stance favours the expression of lateralization in wallabies .
Besides the subject’s and species postural characteristics, manual laterality in mammals has been shown to be influenced by such factors as sex, age, and task complexity. Sex differences in motor preferences have been described in many primates (e.g., [46–51]). Generally, bias for use the left hand is more characteristic of males, whereas a greater right-hand use has been noted for females (e.g., [8, 48, 50, 52, 53]); although a number of primate studies failed to reveal any differences in motor laterality between the sexes [54–60]. The most pronounced sex differences in manual laterality have been reported for non-primate quadrupedal mammals. In horses, domestic cats, and dogs two sexes showed oppositely directed task specializations for the forelimb use resembling primates in the tendency for females to be more right-handed and males ― more left-handed [20, 22, 23, 26, 61].
The effect of age on limb preferences also appears to be a labile category in mammals. In many species researchers have failed to reveal significant age differences in manual laterality [56, 58, 62–64]. However, for some primates shifts toward stronger [65–69] or weaker [46, 70–72] hand preferences with age have been reported. The complexity of the manipulation task might be another factor affecting the degree of motor laterality in primates [53, 70, 73–77]. Fagot and Vauclair  suggested that in primates hand preferences at the population level most likely appear in relatively complex tasks requiring postural, perceptual or cognitive demands, such as bimanual manipulation or catching live prey. For instance, in squirrel monkeys group-level preference of the left hand was expressed in catching a fish, but was absent in reaching of stable food item, i.e., simple reaching .
Despite that great number of mammalian species studied to date in terms of behavioural lateralization in general and in particular in the aspect of manual laterality, on closer inspection, the set of studied taxa still has a number of white spaces. One of such an underrepresented group is the marsupial mammals. To the best of our knowledge, studies of marsupial laterality existing to date only report sensory lateralization in captive stripe-faced dunnarts  and hairy-nosed wombats , individual forelimb preferences in captive brush-tailed possums , and grey short-tailed opossums , as well as population-level manual laterality in captive red-necked wallabies . Nonetheless, data on asymmetrical limb use in marsupials would lead to gain a broader picture of the motor laterality evolution and its possible adaptive value. In addition, the comparison of lateralized limb use between the placentals and marsupials is important because these groups developed in parallel during evolution [83, 84] and share largely similar ecological adaptations and lifestyles.
Marsupials appear to be good candidates for investigating the effect of posture on the manifestation of forelimb preferences. Overall, it is considered that the species-typical posture interacts with the expression of manual preferences in primates. However, very little is known about whether this point is applicable to non-primates. Konerding et al.  showed that in domestic cats forelimb preferences are not affected by task’s postural demands. No differences in the direction or the strength of cats’ lateral biases were revealed between two variants of unimanual task ― food grasping from stable (sitting or standing) vs. unstable body posture (vertical clinging). To our knowledge, only this study together with the one on wallabies  aimed to investigate postural effect on laterality in non-primate mammals. It is, thus, clear that further investigation is needed to understand whether the body posture could shape manual laterality in non-primates or this phenomenon emerged only in the course of primate evolution. Because of the high diversity of postural habits and gaits varying from obligatory quadrupedal to entirely bipedal locomotion [86, 87], marsupials are a suitable model to gain insight into this issue. In red-necked wallabies the bipedal posture was found to increase the laterality , but whether the postural effect across species takes place in marsupials is not known. Since we have previously investigated lateral forelimb biases in a bipedal hopping marsupial, i.e., the species with bipedal locomotion as the preferred gait, we now aimed to study limb preferences in marsupial quadrupeds and compare them with the results in the bipedal species. Here we examine the forelimb preferences at the individual and population levels in two marsupial species, whose typical mode of locomotion is walking and climbing on all four limbs: grey short-tailed opossums, Monodelphis domestica (Didelphimorphia, Didelphidae) and sugar gliders, Petaurus breviceps (Diprotodontia, Petauridae). Basing on the primate data, demonstrating that quadrupedal locomotion tends to hinder the expression of handedness [8, 37, 42, 44], we hypothesized that quadrupedal marsupials should be less lateralized at individual and population level than the bipedal one.
The principal aim of the present study was to explore first the influence of the main factors affecting the motor preferences in placental mammals (such as species-typical posture, sex, age and task complexity) on manual laterality in marsupials. Consistency between the effects of these factors in marsupials and placentals as well as the implication of marsupial data for the current theories of manual laterality is then discussed.
All animals used in the present study were housed in the zoo. Forelimb preferences in both species were investigated in four types of unimanual behaviour: feeding on non-living food, feeding on live insects, supporting the body in the tripedal stance, and nest-material collecting. Unimanual behaviours were not artificially evoked. The animals were video recorded during their usual activity in the dark phase of day-night cycle using the cameras with infrared lighting. Video recording was conducted outside the cages to minimize possible disturbance.