Reporting social behaviours of mixed-species troops formed by Callithrix jacchus and Callithrix penicillata ( Primate , Callitrichidae )

In New World primates, mixed-species troops have been reported. Here, we analysed the performance of affiliative and agonistic behaviours of Callithrix jacchus and Callithrix penicillata living in mixed groups. For this purpose, we recorded the interaction of the individuals from two groups located in Bauru city, in the state of São Paulo (Brazil). Our data show that in both groups, affiliative behaviours appeared more frequently than agonistic ones. We concluded that there is cohesion inside the mixed-species troops observed. We suggest that a deeper knowledge about the social behaviour of mixed-species troop species certainly may be useful in projects linked with the management of the impact caused by them.


Introduction
The common marmoset, Callithrix jacchus (Linnaeus, 1758), and the black tufted-ear marmosets, Callithrix penicillata (E.Geoffroy, 1812), are considered the two most successful species of the Callithrix genus, if we take into account their large geographic distribution, their high population density, and their efficient capacity for habitat exploration (Stevenson and Rylands, 1988;Miranda and Faria, 2001).
C. jacchus and C. penicillata are primates native to northeast and central Brazil, respectively, that have been introduced to regions outside their native geographical distribution (Ruiz-Miranda et al., 2000;Pinto et al., 2009).Both species are invading and possibly replacing other species in many regions of eastern Brazil, such as in the states of Minas Gerais, Espírito Santo, São Paulo, and Rio de Janeiro (Rylands et al., 2009), reaching the southern by a the state of Santa Catarina (Santos et al., 2005).In addition, inside the new habitats where they previously not occurred, these species may act as invaders, with potential damage to the preexisting biota, such as by predation of bird´s eggs (Alexandrino et al., 2012;Almeida et al., 2013).
One study conducted in the city of Bauru, central-western of the São Paulo state, Brazil, described the presence of both species (de Paula et al., 2005), and reported that in this region, C. jacchus and C. penicillata formed mixedspecies troops, with some individuals seeming to be hybrid derivates of the two species.The existence of mixed-species troops of the genus Callithrix has been reported in other studies (Alonso et al., 1987;Modesto and Bergallo, 2008;Sales et al., 2010).A review by Sodaro (1999) emphasises that the successful formation of a mixed-species troops does not guarantee long term advantages and, in comparison to the traditional housing of single species groups of primates, may require higher levels of monitoring to ensure that welfare is not compromised.On the other hand, the associations can be beneficial to both species, as well as to prey, as they help controlling their population (Peres, 1992;Lopes and Ferrari, 1994), predator defense (Peres, 1993), and exploration and defense of larger territories (Garber, 1988;Heymann, 1990).
Studies of interspecific behaviour are important for understanding the nature of associations (Struhsaker, 1981).In this sense, we evaluated the affiliative and agonistic behaviours of two mixed-species troops of C. jacchus and C. penicillata.These behaviours are used in a variety of contexts (Sussman and Garber, 2007), and help to understand the social group structure.

Study area and subjects
We conducted the study in the municipal district of Bauru (22°19'S, 49°04'W), in the central region of São Paulo state (Brazil).At this location, there are large mixed-species troops of both species, i.e. C. jacchus and C. penicillata, as well as their hybrids, broadly distributed throughout the area (de Paula et al., 2005).Some groups have home range near or inside private properties, and are habituated to human presence.The climate is Cwa (Koopen's classification), with a moderately rainy, hot, and humid season in the summer, and a dry and cold winter.
We evaluated two mixed groups: (a) Group 1 (n = 10): one adult male, two adult females, one sub-adult male and one sub-adult female of C. penicillata and two adult males, one adult female of C. jacchus and two infants; (b) Group 2 (n = 8): one adult male of C. penicillata, two adult females of C. jacchus, three juvenile individuals and two infants with intermediate tufts in colour (grey).Each animal within groups was identified by distinctive marks in its morphology, such as scars, size and colour of ear tufts and coat colour pattern.
We estimated individual ages via body size and state of development of sexual organs, in addition to fur characteristics that include the development of ear tufts and white frontal blaze.In this way, adults had fully developed the ear tufts and the sexual organs, whereas sub-adults had incomplete development of them (Decanini and Macedo, 2008).

Experimental design
The two groups are habituated to human presence and, occasionally, received fruits in improvised feeders.Nevertheless, previously to data recording, the two mixed groups of marmosets were monitored in order that they became habituated to our presence.To start the recording of the interactions between the individuals, we place the food in improved feeders.We determined the beginning of interactions when more than one individual arrived at the feeder.After that, we kept the data record during 1 h per day.We collected data of the all-occurrence (Altmann, 1974) of the affiliative and agonistic behaviours.We considered an individual as a participant in the event when it performed any of the behaviours described in Table 1, directed toward another individual of the group.
The data from both groups were recorded for 10 consecutive days, at morning (8h00 -10h00) or afternoon periods (14h00 -16h00), totalling 10 hours per group.

Data analysis
For data analysis, we divided the total number of each registered behaviour by the number of individuals in each group.Within each of them, we compared the total manifestation of affiliative and agonistic behaviours, as well Animal walks or runs after chasing another animal Food dispute When one individual tries to take food from another animal as the rate of each behavioural category via Goodman's proportion test, which considers the dependence between variables (Goodman, 1965).We used a = 0.05 for all the statistical analyses.

Results
In both groups, the affiliative behaviours appeared more frequently than agonistic ones (Goodman's test, p < 0.05; Figure 1).Figure 2 shows the rate of performance of each expressed behaviour.Within Group 1 (Figure 2a), the affiliative play and allogrooming behaviours appeared in the same rate (Goodman´s test, p > 0.05).The frequency of chase was similar to allogrooming (Goodman´s test, p > 0.05) and the fights and bites (agonistic behaviours) occurred in lower proportion than affiliative.Allogrooming was the kind of behaviour most expressed in Group 2 (Goodman's test, p < 0.05; Figure 2b).The play behaviour was expressed in the same rate of bites and fights.There was participation of the infants in all cases of the play behaviour.

Discussion
Our data show that, in both groups, the affiliative behaviours appeared more frequently than the agonistic ones.This indicates that the pattern of primate social structure is maintained with the coexistence of C. jacchus and C. penicillata.
The affiliative behaviours form the basis of associations between individuals in the group (Sussman et al., 2005;Sussman and Garber, 2007).The allogrooming and play were the kinds of affiliative behaviours more frequently observed among the marmosets groups.Allogrooming is a social behaviour common in primates.It is well known that it has a hygienic function by the removal of ectoparasites (Zamma, 2002) and reduction of social tension by the production of endorphins (Schino et al. 1988).It is also linked to group cohesion (Lehmann et al. 2007), and the strengthening of links between individuals (Sánchez- Villagra et al. 1998).Allogrooming can also ensure preferential access to resources (Ventura et al, 2006).In Group 1, the higher rate of play behaviour may be associated with group cohesion, but also could be explained by the presence of infants in it.This behaviour is more frequent among juveniles and infants and certainly helps to improve their social development, motor skills, and encourage learning (Lewis 2000).
The frequency of chase was similar to allogrooming in Group 1.The elevated frequency of chase behaviour in this group may be associated with the presence of more than one adult male.However the marmosets release these agonistic interactions without physical aggression.According to Hinde (1974), when the hierarchy inside the group is well established, individuals with high rank rarely engage in agonistic behaviour.Subordinates usually avoid confrontation, facing away from the simple approach of alpha individuals, thus exerting an active role in the maintenance of hierarchical relationships in social groups.
Our results demonstrate that there is cohesion between the two groups composed of C. jacchus and C. penicillata.It is well known that groups of the Callithrix genus are usually structured around a dominant breeding couple with low levels of aggression among members.Wrangham (1980) stated, ''Groups have evolved as a result of the benefits of cooperation, between allies competing against others of the same species.''This association can bring mutual benefits to the group members, thus keeping the group cohesion can be advantageous.C. jacchus and C. penicillata vary diet types and the number of species consumed (Rylands and Faria, 1993).The existence of abundant food items in the Brazilian savannah (Cerrado) and the diversity of items  that are part of the Callithrix diet can also contribute to the stability of the group.It certainly reduces disputes and consequentially increases the links between individuals.When resources are distributed heterogeneously in time and space, feeding competition and group-living might be less costly (Johnson et al., 2002).
On the other hand, such mixed groups may also bring serious concern, as they may increase the chances of success in foraging, and thus the likelihood of predation of eggs and nests, an important factor in the extinction of bird species in the tropics.Almeida et al. (2013) found high levels of predation on artificial nests by these marmosets and their hybrid derivates, suggesting that it may also occur in natural nests.
A deeper knowledge about the social ecology and interactions between these species may be useful in projects linked with the management of the impact caused by them.Furthermore, additional studies in behavioural, ecological, and genetic factors may supply better understanding of the social behaviours of mixed-species troops of C. jacchus and C. penicillata, as well as in other groups of primates.

Figure 1 .
Figure 1.Rates of performance of affiliative and agonistic behaviours in each mixed groups.The total data was divided by the number of individuals per group (Group 1, n = 10 and Group 2, n = 8).Different letters above the means indicate a significant difference (Goodman´s proportion test, p < 0.05).

Figure 2 .
Figure 2. Rate of performance of each behaviour category.(a) Group 1 and (b) Group 2. Different letters above the means indicate a significant difference (Goodman´s proportion test, p < 0.05).

Table 1 .
Behaviour during observed period for mixed groups of Callithrix jacchus and Callithrix penicillata.
AgonisticBiteThe animal uses teeth to wound parts of the body of another animal Fight Two marmosets grasp each other and bite each other´s head, thorax and abdomen.They tumble, roll, and often fall onto the floor Chase