The common marmoset, Callithrix jacchus, belongs to the Callitrichidae family (New World Monkeys). Endemic in the Northeast forest of Brazil, they have conspicuous large white ear tufts and an annulled tail that changes between black and grey bars. The weight of an adult common marmoset vary between 300 gm to 450 gm, and the average body measurement of an adult common marmoset are neck to tail base 25 cm, tail 28 cm. Common marmosets normally live in extended family groups varying in size between 5-15 individuals. Typically, groups contain only one breeding pair, the dominant male and its mate. The breeding female suppresses reproduction in the other adult females. Breeding females show a high reproductive rate resulting from a biannual twinning and postpartum ovulation.

We have chosen the common marmoset (Callithrix jacchus) as model species, because it has been shown to be a promising model in which to study many aspects of social cognition. Marmosets, like all other callithrichids, show marked tolerance toward each other while feeding, and they frequently and closely attend to the foods others are eating. Callithrichids organize their behavior to a greater extent than do many other monkeys around the task of maintaining spatial and behavioral cohesion with their social partners. Due to their small size, these monkeys are more vulnerable to predation than are other monkeys, and, correspondingly, they are more cautious toward novel spaces and objects. As generalists, they consume fruits, snails, amphibians, lizards, spiders, small birds and nestlings in addition to insects and exudates. The predatory behavior of the genus Callithrix has been described as predominantly a stealthy, stalk and pounce, foliage-gleaning method. Nevertheless, extraction of hidden, embedded prey is a complex foraging technique and as such may not be readily mastered. Youngsters may therefore considerably benefit from skilled adults through observation.

A further positive impact on the role of social learning in common marmosets comes from their well-known cooperative breeding system. As the dominant female usually becomes pregnant again during the lactation period, all individuals of the group care for the young, not just the breeding female. In addition to carrying infants, care giving in marmosets includes provisioning of offspring with food. Large insects are the most frequently shared food items, but other animal matters and fruit are shared as well. The phase of socialization is considered particularly long, taking approximately 15 months. That extension of the period of intense social contact between infant and caring adults also prolongs the time during which socially mediated foraging skills may be developed.

We study marmosets both in captivity and in free-living conditions. Captive subjects are housed at the Biocenter (Althanstrasse) of the Faculty of Life Sciences (University of Vienna). Here we have two indoor cages (250 x 250 x 250 cm), attached to outdoor cages of the same size, in one observation room of the animal keeping facilities. All cages are equipped with branches, ropes and living plants. All animals are fed fruits, vegetables, monkey pellets and protein supplements. The temperature is 24-30°C during the day and 21-23°C at night. The humidity is ca. 50-70%. In summer daylight is the main source of lighting, while in winter additional UV-fluorescent tubes are used to maintain a 12:12 h light:dark cycle. We adhere to the guidelines for the Use of Animals in Research and the legal requirements in Austria. Due to necessary renovations we moved to a provisional keeping facility (in the faculty's greenhouse) in January 2005.

Our main Study site in the field is an area of 32 ha consisting of primary and secondary Atlantic Forest . The area is part of a condominium (a housing estate surrounded by a wall, which comprises about a hundred houses, greens, and several leisure facilities). It is situated 40km west of Recife in the state of Pernambuco, in the Northeast of Brazil (7°56'97''S, 35°1'23''W, 174m above sea level). The climate is seasonal. In the rainy season (June - August) temperatures range from 17 to 28°C and humidity from 90 to 100%. In the dry season temperatures can vary between 25 and 32°C with 70 to 80% humidity. Resources are more abundant in the dry season. Despite human impact, the area has preserved its richness and diversity regarding fauna and flora. The forest and its animal diversity are under law protection (IBAMA). The common marmosets are well-accustomed to the presence of humans.

We have selected marmosets to study several aspects of social cognition. Among them are imitation as an important mechanism of social learning, the function of social learning in the wild, the conditions for social learning during foraging (social relationships, dominance, food tolerance, cooperation), the development of social learning skill and the time course of its application. Furthermore, we want to adress in which form this primate species represent objects in its internal world, thereby studying object permanence and cross-modal representation.

Imitation in marmosets

The capacity to imitate the movements of conspecifics during foraging behaviour, while highly adaptive, is among the least common and most complex forms of learning yet identified. Although not uniquely human, it has been shown convincingly in few other species. It therefore provides a very challenging and potentially informative model which to study the evolution of learning. The primary purpose of this project is to examine whether monkeys, in addition to apes, are able to learn through imitation. Our attempt is based on challenging the assumption of cognitivists that, in proposing the emergence of meta-representation in higher primates, apes can, and monkeys cannot imitate. The results of this project, bublished in Bugnyar & Huber (1997 Anim Behav) and Völkl & Huber (2000 Anim Behav), are - to our knowledge - the first experimental evidence of marmoset (perhaps monkey) imitation in a functional task designed to simulate foraging behaviour in the wild.

As part of a European funded project (EDICI; www.univie.ac.at/edici) we aimed a carefully controlled investigation of the mechanisms mediating, and the factors contributing to, motor imitation in non-human species. Previous research has primarily been focused on imitative performance rather than imitation learning. The latter has been called the ‘transfer of skill problem’. It has received great attention due to its ostensible role in the development of culture and the cognitive demands it poses on the individual.

However, despite a century of research and the detection of mirror neurons, the empirical basis for this most advanced form of observational learning was somewhat weak. Few, if any, studies have shown that the animal observer has learned the response topography, i.e., the specific action by which the response is made. The purpose of our research was, first, to establish a methodology that could objectively determine the matching degree between the performance of a model and that of its observer, and also to look for any evidence of imitative learning. In an experimental set-up, we confronted our marmosets with a conspecific model that had been previously trained to open a plastic box in a peculiar way (i.e. with the mouth instead of the hand). Employing detailed motion analyses we showed that the observers precisely copied the movement patterns of the novel action demonstrated by the model (Voelkl & Huber 2007 PLoS ONE).

This finding challenges two dominant views of imitation. 1) The evidence of imitation in non-human primates questions the dominant opinion that imitation is a human-specific ability; that it is either an innate capacity of humans or is based on some human-specific cognitive abilities (eg intention reading). 2) The high matching degree suggests that marmosets possess the neuronal mechanism to code the actions of others and to map them onto their own motor repertoire, rather than priming existing motor-templates (Huber et al. 2009 Phil Trans).

The development of social learning on foraging in free-living common marmosets

It is widely agreed that foraging skills in young primates are acquired during social interactions with more experienced individuals. However, the developmental patterns of those social learning effects have been rarely explored in detail. We therefore investigated the development of social information acquisition in free-living common marmosets by observing 32 individuals during a period of eight months in the Atlantic forest of Camaragibe (Pernambuco-Brazil). The subject groups were divided into four age classes or developmental classes: class I (1 and 2 months old), class II (3 and 4 months old), class III (5 and 10 months old) and class IV (older than 15 months). We recorded different foraging patterns and classified them into three behavioural categories: ‘model independent foraging', ‘model observation' and ‘model dependent foraging'. The latter behavioural category was further divided into three sub-categories: ‘follow the model', ‘manipulate same object' and ‘forage together'. We found a significant increase of the frequency of all foraging patterns from class I to class II, and a significant decrease from class II to class III in the social mediated patterns. In addition to this general trend we found a significant difference in relative frequencies between behavioural patterns labeled ‘manipulate same object' and ‘model independent foraging'. Animals from class III observed and responded to models more often than individuals from the other classes. Our results thus provide evidence that mechanisms of social learning are age-dependent. While the increase in frequency of the observed foraging patterns in class II may be explained by maturation processes, the decrease in class III indicates learning effects. The results also suggest that in early cognitive development of common marmosets social learning is primarily based on attention processes (Schiel & Huber 2006 Am J Primatol). This developmental pattern of social learning has recently been confirmed in a laboratory study (Dell'mour, Range & Huber 2009 Am J Primatol).

Which factors determine cooperation?

Here we investigate the conditions under which marmosets work together and thereby benefit in a food manipulation task. Our focus of interest is towards the social conditions, particularly the level of inter-individual tolerance and the distribution of roles, in accounting for the appearance of collective achievements in groups, whatever the goals and tactics of, or benefits for, individuals may be . From what is known about the cooperative breeding system one might also expect an effect of age, sex and dominance status. In a first study (Werdenich & Huber 2002) we found very interesting effects. During the approach phase each of 8 individuals of a family group learned in isolation to pull a handle moving a bowl with attractive food towards its reach. In the following two experimental phases we used 16 dyads. In a dyadic training phase we assessed whether the partners were willing to manipulate the apparatus and to share food in reciprocity. In the subsequent cooperation test we examined whether they were willing to cooperate at a slightly modified apparatus whose solution required pulling of the handle by one individual (the producer) and grasping the bowl by the other animal (the scrounger). Although all individuals were willing to cooperate with at least one partner, only half of the dyads solved the task in the cooperation phase. Examination of the factors that correlated with success in the cooperation phase revealed that primarily those dyads cooperated in which the dominant subject took the role of the scrounger and the submissive one took the role of the producer. However, in these successful dyads the dominant animal didn't force the subordinate partner to pull the handle. Rather, the partners of cooperative dyads shared the reward and pulled equally often in both the dyadic training and the cooperation test. Thus, cooperation of marmosets in a food manipulation task seems to depend on a specific distribution of roles and the tolerance of higher-ranking individuals.

Object permanence in marmosets

On the basis of experiments with rhesus monkeys and squirrel monkeys it has been concluded that the ability to mentally represent unperceived events is shared by humans and great apes but not monkeys. However, interactions with moving objects are crucial for the survival of many organisms. In a context in which important objects like a mate, a predator, or a prey appears, disappears, and reappears, the interactions require functional invariants. In particular, animals that move efficiently through a structured environment in three dimensions of space, like primates, must be equipped with a capacity to recognize the important object as being the same through its repeated appearances over time. Therefore, we don't share the dichotomous view of the distribution of cognitive skills in primates. We instead recommend that attention be given to the lesser apes to gain a more accurate understanding of the distribution and the evolution of cognitive skills in nonhuman primates. Any general hypotheses about the object permanence skills of primates will remain premature until more research is done with marmosets and tamarins (Tomasello and Call 1997).

Therefore, the purpose of this project is to assess as accurately as possible the level of object permanence that is displayed in the behavior of marmosets and to analyze their search activities. Object permanence is essentially assessed through standardized visible and invisible displacement tests, in which subjects have to search for and find occluded objects. In these tests, the standard testing procedure involves the use of a number of screens—for instance three—behind which an object that the subject will have to search for can be hidden.

In a first study (Mendes & Huber 2004) a series of 9 search tasks corresponding to the Piagetian Stages 3–6 of object permanence were administered to 11 common marmosets. Success rates varied strongly among tasks and marmosets, but the performances of most subjects were above chance level on the majority of tasks of visible and invisible displacements. Although up to 24 trials were administered in the tests, subjects did not improve their performance across trials. Errors were due to preferences for specific locations or boxes, simple search strategies, and attentional deficits. The performances of at least 2 subjects that achieved very high scores up to the successive invisible displacement task suggest that this species is able to represent the existence and the movements of unperceived objects.