The social group of the brown howler monkey, Alouatta guariba clamitans Cabrera, 1940 is typically small (2–12 individuals), with one or two adult males, and more females than males (Silva Jr., 1981; Neville et al., 1988; Mendes, 1989; Steinmetz, 2000, 2001; Treves, 2001; Miranda and Passos, 2005). In primates where males outnumber females, males compete for both females and food, which normally results in groups with dominant males (Clutton-Brock and Harvey, 1977; Dunbar, 1988; Mittermeier et al., 1999). As in all howler monkeys, A. g. clamitans is sexually dimorphic; adult males are larger than females and are dominant to them (von Ihering, 1914; Altmann, 1959; Neville et al., 1988; Bonvicino, 1989; Mendes, 1989; Chiarello, 1995; Mittermeier et al., 1999; Treves, 2001). Although howler monkeys are the most studied of the Neotropical primates, most of the research has been concentrated on a few species, in particular A. palliata and A. seniculus, and only more recently A. caraya, A. pigra and A. guariba (Neville et al., 1988; Bicca-Marques, 2003). Field studies on the composition, social structure and hierarchy of A. g. clamitans are rare (Miranda, 2004; Jardim, 2005; Miranda and Passos, 2005), and long-term investigation will be necessary to obtain more detailed information on their social patterns. Here we describe a particular case of an A. g. clamitans group in which the dominant animal was an adult female — in effect, a matriarchal group.
The study area is a forest fragment of approximately 700 ha in the Chácara Payquerê, a component of the Área de Proteção Ambiental da Escarpa Devoniana, in the municipality of Balsa Nova, Paraná, Brazil (25°29′52″S, 49°39′24″W). This remnant forest is within the Floresta Ombrófila Mista (Araucaria Pine Forest biome) and consists mainly of disturbed primary forest, broken by patches of secondary forest.
Observations of the behavior and ecology of brown howler monkeys in the study area have been continuous since February 2002 (Miranda, 2004; Miranda and Passos, 2004, 2005; Miranda et al., 2005; Miranda et al., in press). The study group, the Forninho Group, was the best known in the area while its members remained together. Our observations here were recorded ad libitum (Altmann, 1974). We used the system of age-sex classification proposed by Mendes (1989) and used by Miranda and Passos (2005).
In February 2002, the Forninho Group included two adult males (AM), three adult females (AF), and a juvenile I (JI). After four births through two years of study (two births in 2003 and two in 2004), the group consisted of two AM, a sub-adult male (SAM), three AF, two juveniles II (JII), one male and one female, and two JI (both males). In January 2004 the dominant adult male (AM1) disappeared, which apparently disrupted the social structure of the group. This resulted in the Forninho Group splitting into two: the Carrano Group with one AM, one AF and one JII (female), and the Vavá Group with one SAM, two AF (AF1 and AF2), one JII (male), two JI (males) and one infant (INF) born one month after the disappearance and presumed death of AM1 (Table 1). Another birth occurred in July of 2004, this time to an AF2 in the Vavá Group. Miranda and Passos (2005) give the details of the dynamics of this group.
Composition of rhe Forninho social group and of the resulting (Vavá and Carrano) groups which were formed after it split up, coincident with the disappearance of the adult male AM1.
AF1 of the Vavá Group was the mother of the SAM, as well as the JII, one of the JI and the older INF. AF1 may also have been the mother of AF2, but we could not confirm this because AF2 was already an adult at the beginning of our study. If AF1 was indeed the mother of AF2, then every member of the Vavá Group would have been the progeny of this single adult female.
In confrontations with neighboring groups, AF1 vocalized the most (bark and howl sensu Oliveira, 2002) and challenged the dominant AM of Carrano Group, forcing him away from the confrontation. This female behaved in the same way toward humans when one of the investigators accidentally startled the animals, alarming the group and provoking AF1 to come their defense. On two occasions, AF1 challenged the subadult male over food. The male retreated both times.
The breakup of the original Forninho Group may have been prompted by AM2, who was possibly the son of AF1, and, as such, leaving the group to avoid endogamy. Following his departure, AM2 became the nucleus of the new Carrano Group. However, given the occurrence of infanticide in howler monkeys (Clarke, 1983; Zunino et al., 1986; Rumiz, 1990; Clarke et al., 1994; Galetti et al., 1994; Calegaro-Marques and Bicca-Marques, 1996; Palacios, 2000; Aguiar et al., in press), it is also possible that AF1 may have forced AM2 out of the group, since AF1 was pregnant by AM1, who had been the dominant male until his disappearance. AF1, therefore, may have been trying to protect her infant from possible infanticide by AM2. Mendes (1989) observed a similar case in which an AM drove out the dominant male of his study group. Although successful in his ouster of the dominant male, the AM in Mendes' study group was forced out by the two adult females of the group, primarily by the adult female who was pregnant. This behavior on the part of the female ended only when her infant disappeared, and she finally accepted the presence of the new male. Mendes (1989) interpreted these events as an attempt by the pregnant female to avoid infanticide.
The AF1 was followed by her offspring with AM1 who had disappeared — INF, JI, JII and SAM, and possibly AF2 — which gave rise to a group dominated by a female rather than a male. AF1 remained dominant despite the presence of the subadult male; she was the one who defended the group against neighboring howlers and the researchers who accidentally startled her group. Hirano (2003), studying A. guariba, and Calegaro-Marques and Bicca-Marques (1997), studying A. caraya, have reported that adult females are dominant to subadult males.
In howler monkeys, individuals of both sexes emigrate, forming new groups or simply integrating with another group (Neville et al., 1988; Mendes, 1989; Bonvicino, 1989; Calegaro-Marques and Bicca-Marques, 1996; Giudice, 1997; Brockett et al., 2000; Ostro et al., 2001; Jardim, 2005). Bonvicino (1989) and Mendes (1989) recorded only subadult males as solitary individuals. We saw solitary adults of both sexes in our study area, showing that females emigrate as well as males (Miranda, 2004; Miranda and Passos, 2005). The subadult male of the Vavá Group could have remained in the group to become the dominant adult, or else moved out to avoid endogamy. Groups without adult males may be more susceptible to predation, as well as to infanticide by strange adult males or to an upset in group hierarchy (Zunino et al., 1986; Dunbar, 1988; Clarke et al., 1994; Galetti et al., 1994; Palacios, 2000). We would expect that sooner or later an adult male, either the resident subadult or an immigrant, would take over as the dominant individual in the group.
Groups of A. g. clamitans are usually dominated by males, but they may be adaptable enough to change this pattern in certain cases. We believe that this is evident in the case of our original focal group, the Forninho Group, in which an adult female left in order to avoid imminent infanticide, establishing, at least temporarily, a female-dominated group.
We thank Édner L. Rosa for permission to work on his property and Kauê Cachuba Abreu for his help with the fieldwork. We also appreciate the assistance of Dr. Albert Leyva in preparing the manuscript.
 João M. D. Miranda, Postgraduate in Zoology – UFPR, Rua Rio Guaporé 1275, Bairro Alto, Curitiba, 82840-320 Paraná, Brazil, e-mail: <email@example.com>
 Itiberê P. Bernardi, Graduate in Biology – URI do Alto Uruguai e das Missões, Rua Guerino Cerutti 161, ap. 702, Centro, 98400-000 Frederico Westphalen, Rio Grande do Sul
 Rodrigo F. Moro-Rios, Graduate in Biology – UFPR, Rua Wilson Valdívia Domingues 203, casa 6, Bairro Jardim das Américas, Curitiba 81540-170, Paraná
 Lucas M. Aguiar and Gabriela Ludwig, Postgraduate in Zoology – UFPR, Rua Bronislau Ostoja Roguski 649, Casa 3, Bairro Jardim das Américas, Curitiba 81540-080, Paraná