Ecological Factors and Sociality in Rodents

Large (>1 kg) arboreal rodent genera that include an important amount of fruits in their diet (i.e., frugi-omnivorous and frugi-herbivorous forms), are more likely to be social than other rodents (Müller and Soligo, 2002, 2005). In contrast, activity cycle did not appear to influence the presence or absence of sociality (Müller and Soligo, 2005), which is in agreement with the fact that all primates exhibit a social lifestyle, whether they are nocturnal or diurnal. A diurnal lifestyle, however, seems to be an important attribute that promotes living in cohesive groups (van Schaik and van Hooff, 1983). The nocturnal group-living species Avahi spp. and Aotus spp. may not represent valid counterarguments as it has been suggested that the nocturnal lifestyle of these species is a secondary attribute (Eisenberg et al., 1972; Martin, 1990; Müller and Thalmann, 2000; Sussman, 1999, 2000; Thalmann, 2001) and that their gregarious pattern of social organization arose from the gregarious pattern of diurnal primates and not from the dispersed nocturnal type (Müller and Thalmann, 2000).

The fact that large rodents (>1 kg) exhibit a social lifestyle more often than medium-sized and small forms would support the suggestion that small body size is an important determinant of a "solitary" lifestyle as found among nocturnal prosimians (e.g., Charles-Dominique, 1978; Clutton-Brock and Harvey, 1977). However, most "solitary" nocturnal prosimians are social, but weigh less than 1 kg. Large body size in primates might therefore at best be a determinant for foraging in cohesive groups. Alternatively, sociality may have evolved in an ancestral primate with a body weight above 1 kg and was subsequently retained throughout the evolution of the order. As body size influences many life-history traits and affects social behavior, sociality may in fact not be directly linked to body size but to size-related factors instead (Armitage, 1981).

It is puzzling that arboreal rodents are more likely to be social than their terrestrial counterparts, as this seems to contradict the hypothesis that predation pressure encourages sociality and because terrestrial animals are assumed to be more susceptible to predation than arboreal species (Dunbar, 1988). However, since the presence of social networks is not equivalent to living in groups, it does not in itself carry benefits concerning predation avoidance. Furthermore, at least among primates, terrestrial species do not necessarily suffer higher predation risks than arboreal forms (Isbell, 1994). The study on rodents, however, revealed that arboreal and terrestrial genera differ in their dietary preferences (Figure 2). Whereas the amount of frugi-omnivorous genera is the same both in the tree and on the ground, more terrestrial rodents are faunivorous or herbivorous and more arboreal forms have a granivorous or frugi-herbivorous diet (Müller and Soligo, 2005) (Figure 2). This is not surprising since food availability is not the same

Arboreal

Terrestrial

Arboreal

Terrestrial

■ Frugi-omnivore mi Frugi-herbivore

□ Herbivore s Faunivore

Figure 2. Percentage of dietary categories in arboreal and terrestrial rodent genera (after Müller and Soligo, 2005).

for arboreal and terrestrial animals: insects occur mainly in the leaf-litter on the ground or in the shrub and underground layer (Cartmill, 1972) and grasses, forbs, roots, and stalks are not found in trees. By contrast, fruits and berries grow in trees and shrubs and their availability to terrestrial species is therefore very limited.

Insects generally occur in small and highly dispersed patches (Oates, 1987). Green vegetation is either uniformly scattered and more-or-less stable (leaves, roots, stems, and stalks) or clumped and unpredictable (grasses, herbs, and forbs) (Clutton-Brock, 1974). By contrast, fruits, berries, seeds, young leaves, and flowers are patchily distributed in space and time and therefore, difficult to find. Hence, the limiting factor for feeding on these items is to find a good site (Krebs and Davies, 1987; Milton, 1981; Oates, 1987). Once a good location is known, the resource can be depended on because fruits, seeds, flowers, and young leaves show a high degree of predictability. Animals can move directly between resources if they know where to find them and when they are fruiting or flowering. This is obviously preferable to wasting time and energy by searching in a random fashion. Because the selective factor enabling animals to locate such resources seems to be increased mental complexity with an emphasis on learning and retention rather than genetic coding, membership in a close social unit can increase the efficiency of food exploitation if information on feeding sites is shared (Cords, 2000; Gautier-Hion, 1988; Milton, 1981). This has also been suggested for nesting colonies and communal roots of birds that may serve as "information centers" where individuals can find out where good feeding sites are (Ward and Zahavi, 1973). The same might be true of nocturnal prosimians and other animals that feed on dispersed but predictable food resources. In a social network, adult animals can teach their offspring or siblings about food locations (Clark, 1985; Müller, 1999a; Müller and Soligo, 2002).

This hypothesis is supported by the fact that the presence of a social lifestyle in rodents varies according to dietary preferences (Figure 3). The largest number of social rodents are found within frugi-omnivorous and frugi-herbivorous genera and the proportion of social rodents within these categories lies well above the rodent average (Müller and Soligo, 2005). Although seeds are believed to be patchily distributed, granivorous rodents are less likely to be social than other rodents. Many granivorous rodents hoard food and limiting pilferage could be more important to those animals than maximizing the efficiency of localizing food (Müller and Soligo, 2005).

Frugi- Frugi- Herbivore Faunivore Granivore herbivore omnivore

Dietary category

Figure 3. Percentage of dietary categories in social and nonsocial rodent genera. Rodent average (i.e., overall percentage of social genera in rodents) is 66.2% (after Müller and Soligo, 2005).

Frugi- Frugi- Herbivore Faunivore Granivore herbivore omnivore

Dietary category

Figure 3. Percentage of dietary categories in social and nonsocial rodent genera. Rodent average (i.e., overall percentage of social genera in rodents) is 66.2% (after Müller and Soligo, 2005).

The fact that large (>1 kg) arboreal genera that feed on fruits and vegetation or on fruits and animal matter are more likely to be social than other rodents supports the notion of a link between a diet that includes an important amount of fruits and the evolution of social networks. The relevant factor between frugivory and the development of social networks lies in the importance of sharing information on food resources. This can be extended to argue that a frugivorous diet lay behind the evolution of social networks in early primates (Müller and Soligo, 2002, 2005). Indeed, social networks are thought to have already been present in the last common ancestor of living primates (Müller and Thalmann, 2000). Based on the ecological correlates of sociality found in rodents, Müller and Soligo (2005) have predicted that social stem-primates were relatively large (>1 kg) and arboreal and had a diet that included a significant amount of fruit. This hypothesis is also supported by data on marsupials where those species that include fruits in their diets exhibit social networks whereas others do not (Müller, 1999b; Rasmussen and Sussman, this volume). It is, however, partially at odds with the current majority view concerning the nature of the first primates, as these are often believed to have been small, insectivorous, nocturnal, and arboreal.

0 0

Post a comment