The formation of mixed-species troops seems to be fairly
widespread among sympatric Saguinus species. For most areas where quantitative
information is available, between 50% and 100% of sightings of tamarins are
of mixed-species troops (Table I). The Santa Cecilia site is apparently an exception.
This site is characterized by large extensions of swamp forest ["bajial";
Glander et al., 1984], a forest type that is used only marginally by S. mystax
and S. fuscicollis. Population densities of these two species are low
at Santa Cecilia in comparison to other sites I Glander et al., 19841, which
probably affects patterns of interspecific association. The figure given for
the Rio Yarapa site has to be considered with care since it is based on trapping
results rather than direct observations. The possibility cannot be ruled out
that the two tamarin species were attracted to the traps independently.
Struhsaker [1981] found that census data reveal higher frequencies
of polyspecific associations than longitudinal studies on single groups in African
cercopithecids. Associations generally comprise more individuals than monospecific
groups and thus are more likely to be detected during censuses. During longitudinal
studies, differential habituation to an observer by the associated species may
result in much lower figures for the degree of association. In contrast to these
findings the frequency of sightings of mixed-species troops at the Rio Blanco
study site (72% for S. mystax) was in the same order of magnitude as
the average frequency of association of the study group (82%). Therefore, the
presented figures seem not to be biased by the sample method but rather give
a reliable estimate of the degree of polyspecific association of S. mystax
and S. fuscicollis at the study site. Time spent in association is comparable
to the figure given by Garber [1988a] for a mixed-species troop of moustached
and saddle-back tamarins on the left bank of the Rio Blanco (72%), but is essentially
higher than those for mixed-species troops of S. labiatus and S. fuscicollis,
where 43 to 57% of time was spent in association [Pook & Pook, 1982]. Mixed-species
troops of S. imperator and S. fuscicollis also spent much time apart
[Terborgh, 1983]. Whether different observational techniques, interspecific
variability, or the different habitats are responsible for these differences
cannot be determined with the limited data base available so far.
Time spent in association varied diurnally. Temporal variation
in the degree of association is also reported from a mixed-species troop of
S. mystax and S. fuscicollis from the left bank of the Rio Blanco
[Garber, 1988a] and for associations of S. labiatus and S. fuscicollis
[Pook & Pook, 1982]. While Pook and Pook observed that association was high
in the first half of the day and in the early afternoon but low around midday,
the pattern found by Garber is more similar to that described in this paper
with a relatively low degree of association during the first hours of the day.
Intergroup encounters, i.e., encounters of the study group with neighboring
groups, influence the time spent in association. Since these encounters generally
take place in the morning, the lower degree of association at this time of day
can be explained at least in part. It is possible that other factors also affect
the time spent in association, e.g., the activity patterns of the species. However,
this remains to be tested.
Interspecific interactions generally seem to be rare in mixed-species troops of tamarins. Pook and Pook [1982] observed only two "close encounters" between S. labiatus and S. fuscicollis. In both cases an S. labiatus threatened a S. fuscicollis. Yoneda [1981, 1984] occasionally observed threatening behavior between S. labiatus and S. fuscicollis in feeding trees and Terborgh [1983] reports aggressive exclusion of S. fuscicollis from feeding trees by S. imperator. In the present study, nearly all interspecific social interactions were agonistic and occurred at plant food resources. From the data a trend is apparent for agonistic interactions to occur at small food resources. However, it should be taken into consideration that small resources are generally located in the lower strata of the forest where visibility by the observer is good, while large resources are found at greater heights where observations are more difficult or even impeded by the dense vegetation. Nevertheless, aggressive interactions (which comprised the majority of agonistic interactions) are normally accompanied by specific vocalizations, and very few such vocalizations were heard while the troop was feeding in large resources. Agonistic interactions among moustached tamarins occurred at a rate of one in 12 h of observation and were thus much rarer than interspecific agonistic interactions. Furthermore, more than 50% of these Intraspecific agonistic interactions were observed in the context of intergroup encounters (but between individuals of the study group) and about 30% during feeding on plant food resources (Heymann, in prep. a).
Interspecific agonistic interactions at food resources incur some costs to both moustached and saddle-back tamarins. The exact amount of these costs cannot be determined in this study since the relative importance of the debated resources is not known. In his study on S. fuscicollis illigeri (which is not sympatric with another Saguinus species), Soini [1987] found that the quantitative contribution of small food resources to the diet was small. Assuming that this holds true also in the present case, this would mean that the costs of being deprived (at least partially) from these resources are probably rather low for the saddle-back tamarins. Since aggressive interactions only very rarely involve direct physical contact, there is little risk of being,hurt during such an interaction. Moustached tamarins are always dominant, even when one moustached tamarin is confronted with two or more saddle-backs. Therefore, the costs to the former of behaving aggressively should be low, too, and should be outweighed by the benefit of gaining first or exclusive access to small resources. Since interspecific aggressive interactions are rather rare (approx. one in 7 h) the overall costs are expected to be very low.
No interspecific social grooming was observed between moustached
and saddle-back tamarins, although both species frequently rested together at
the same resting site and occasionally within a few meters. The possibility
cannot be ruled out that some interspecific grooming has escaped the observer's
attention due to the often reduced observability when the tamarins were resting
high up in the canopy. Anyway, interspecific grooming would be extremely rare
in comparison to Intraspecific grooming in the moustached tamarins (approx.
one grooming interaction per h of observation; Heymann, in prep.a). Struhsaker
[1981] found a significant inverse correlation between dietary overlap and rates
of interspecific social grooming in mixed-species troops of cercopithecines.
Dietary overlap is very high in mixed-species troops of tamarins [Garber, 1986,
1988b; Terborgh, 1983] (personal observation), which probably might explain
why interspecific social grooming did not occur. In a captive mixed-species
troop of moustached and saddle-back tamarins, interspecific social grooming
was regularly observed [Heymann & Sicchar, 1988]. The provisioning of this
troop may have reduced struggling for food and thus allowed for interspecific
grooming to occur.
A mutual attraction between moustached and saddle-back tamarins
is evident from the observation of mutual long calling. Such mutual calling
has also been observed in mixed-species troops of S. labiatus and S.
fuscicollis [Pook & Pook,1982; Yoneda, 1984] and S. imperator and S.
fuscicollis [Terborgh, 1983] and has been interpreted as the means by which
association is established and maintained. In the present study it took longer
for the two species to join each other when a long call was given than when
no long call was given. In those cases where no long call was given the two
species generally had slept in adjacent sleeping sites. The tamarins thus were
already close together after leaving the sleeping sites and contact could be
established visually or by means of low-pitched contact calls. Long calls were
given when the sleeping sites were more distant from each other and therefore
it also took longer until the two species join each other.
Interspecific long calling incurs some costs to the caller.
These costs are made up of the energetic expenditure of calling and by the probably
higher risk of being spotted by a potential predator. However, the number of
long calls that are given in an interspecific context are very low in comparison
to those given in an Intraspecific context (intergroup encounters). Therefore,
the overall costs of interspecific calling are very low.
The benefits of forming mixed-species troops in tamarins
remain to be determined. A number of possible benefits have been proposed [Terborgh,
1983], amongst them enhanced detection and avoidance of predators. There are
very few observations of predation on tamarins, and these indicate that raptors,
snakes, and felids may be the most important predators [Emmons, 1987; Heymann,
1987] (Goldizen, 1987). Alarm calls given in response to potential avian predators
are mutually understood by moustached and saddle-back tamarins (Heymann, in
press). Since moustached tamarins occupy higher forest strata than the saddle-back
tamarins [Norconk, this symposium] (personal observation), they may detect raptors
earlier, which would benefit the saddle-back tamarins. On the other hand, moustached
tamarins might benefit from the saddle-back tamarins' earlier detection of terrestrial
predators or predators located at lower forest strata. Such an interspecific
division of roles has been found in polyspecific associations of Cercopithecus
cephus, Cercopithecus pogonias, and Cercopithecus nictitans
[Gautier-Hion et al., 1983]. It is noteworthy in this context that on an occasion
of snakemobbing by saddle-back tamarins, the moustached tamarins did not participate
[Bartecki & Heymann, 1987c].
Another benefit of mixed-species troops in tamarins that
has been proposed is the joint defense of the territory and food resources against
neighboring troops [Garber, 1988a; Terborgh, 1983]. It is not clear from the
descriptions given by Garber and Terborgh whether this means that intergroup
encounters may also involve interactions between heterospecific individuals
from different troops. In the present study,intergroup encounters were
always interspecific. Confrontations between a moustached tamarin of the study
group and a saddle-back tamarin of a neighboring group have never been observed.
During intergroup encounters moustached and saddle-back tamarins spent more
time apart from each other than at other times of the day. However, if joint
territorial defense means the defense of the same territory or the same resources
regardless of whether intergroup encounters are Intraspecific or interspecific,
then in the present study joint territorial defense was also evident. The home-range
of the moustached tamarin group and of the saddle-back tamarin group overlapped
nearly completely, with that of the saddle-back tamarins being slightly larger
(Heymann, in prep. b).
To elucidate the proximate and ultimate factors in the formation of mixed-species troops in tamarins requires further data. Particularly, it will be necessary to combine ecological information (e.g., diet, foraging patterns) with behavioural data on interspecific interactions to obtain more detailed estimates of costs and benefits of associations.
