We evaluated variation in group size and composition of Udzungwa red colobus (Procolobus gordonorum) in relation to gross-habitat and sociological parameters. The endangered species is endemic to the Udzungwa Mountains and nearby forests in the Kilombero Valley of south-central Tanzania. We counted 63 groups in 10 forests, ranging in altitude from 250 to 1,800 m. Group size ranged from 7 to 83 (x = 36.3) individuals and adult sex ratios (females/male) ranged from 1.5 to 7.3 (x = 3.5), excluding solitary individuals. Group size was influenced by several habitat parameters, including tree density, degree of deciduousness, and forest size. Groups were largest in large blocks of mature, moist, mixed evergreen and semideciduous forests, but group size is not correlated with altitude. Groups in a highly degraded forest appeared to have fission-fusion societies. The effect of habitat quality on age-sex composition of groups was most apparent in natality and less so in survivorship of adult females and juveniles. The number of adult males in groups accounted for 50% of the variance in group size and 34-39% of the variance in numbers of adult females in groups. Habitat quality affects natality more than demographic parameters do. Groups with a low proportion of

A distinct species of mangabey was independently found at two sites 370 kilometers apart in southern Tanzania (Mount Rungwe and Livingstone in the Southern Highlands and Ndundulu in the Udzungwa Mountains). This new species is described here and given the name "highland mangabey" Lophocebus kipunji sp. nov. We place this monkey in Lophocebus, because it possesses noncontrasting black eyelids and is arboreal. L. kipunji is distinguished from other mangabeys by the color of its pelage; long, upright crest; off-white tail and ventrum; and loud call. This find has implications for primate evolution, African biogeography, and forest conservation.

An analysis of 3,774 episodes of agonistic aiding collected during a two-year study of a rhesus monkey group (Macaca mulatta) indicated the differential influence of kinship and rank relationships on the participation of different age-sex classes in both aid to victims and aid to aggressors. Most aiding favored victims rather than aggressors and was much more likely to occur when matrilineal kin were involved. Females were more likely to aid than were males, and the frequency of their participation increased with age. Females were much more influenced by kinship than were males and defended or aggressively supported kin against any third party regardless of dominance relationships. Adult males seldom aided against animals that were dominant to themselves; the rare exceptions occurred when adult males defended kin. Aiding was far more likely to occur if the victim was squealing, and noisy agonistic episodes often involved multiple aiders on both sides. Aiding patterns had some potential to insure dominance rank inheritance within families, in accordance with the Kawamura hypothesis. In aiding animals outside of their own matrilines, however, group members aided randomly with respect to this model. There was little evidence that aiding functioned to support individuals when they targeted animals to which they should be dominant as adults based on matrilineal dominance relationships. Most defensive aiding seemed to function primarily to defend victims (primarily kin) of aggression. Aggressive support of the attacker, on the other hand, seemed to function primarily to reinforce coalitions with the attacker. The identity of the victim was unimportant as long as it was neither kin to nor dominant to the aider. Aggressive support of attackers did not overturn existing dominance relationships.

Sanje mangabeys (Cercocebus sanjei), first described in 1981, are among the most endangered primates in the world. They are endemic to the Udzungwa Mountains of Tanzania, in a biogeographic region designated one of the world's biodiversity hotspots. Conservation research since 1997 has documented the presence of the mangabey in only 3 of the relict montane forest blocks of the Udzungwas. The total population, possibly <1,500 animals, is fragmented and not adequately protected. A substantial proportion (perhaps 40%) live in forest reserves outside the protective confines of the Udzungwa Mountains National Park, and they are affected by habitat loss and hunting. Efforts to improve their conservation status include assessment of distribution, relative abundance, and habitat quality, and initiation of observational research with habituated individuals to acquire critically important data on their habitat requirements, diet, movement patterns, socioecology, and community ecology. These interrelated research activities should contribute to effective management for conservation, provide baseline information to support current efforts to expand the boundaries of the national park, and guide potential future establishment of corridors between the major forests known to support mangabey groups.

Both age and sex influence rhesus monkey agonistic behavior. In intrapoup episodes, submission was most frequent in juveniles, but aggression increased steadily with age, albeit much more sharply in females. As infants, males were more often involved in agonistic behavior than were females, but this sex difference reversed with age. A notable change in the frequency and forms of agonistic expression occurred in adolescent males. By the time they became adults, their participation in agonistic episodes was silent and brief and rarely involved biting. Adolescent males received high frequencies of aggressive responses, and this is hypothesized to account for the marked shift in adult male patterns of participation in intragroup agonistic interactions, as relative to females whose basic pattern of agonistic expression does not change with age. Scott (1958) defined agonistic behavior as aggression and responses to aggression (including submission). These responses have attracted the attention of investigators of primate behavior because they are often noisy, result in severe injuries, which are noticeable for long periods of time, and seem, subjectively, to be socially disruptive. Agonistic behavior, however, is relatively rare in primate societies, constituting less than 5% and often less than 2% of social behavior (

No abstract

Many individual researchers have used line transect counts to estimate forest primate abundance. They have devoted less attention to the interpretation of line transect data obtained by several observers, as is often the case in long-term monitoring programs. We present primate relative abundance data that 5 observers collected over 6 yr (not continuous) along 4 different transects each 4 km long in the Mwanihana Forest, Udzungwa Mountains National Park, Tanzania. Total distance walked during transect repetitions is ca. 700 km. The species we saw most frequently was the endemic Udzungwa red colobus Procolobus gordonorum (mean 0.59 groups/km walked), followed by the Angolan black-and-white colobus Colobus angolensis (0.43 groups/km) and Sykes's monkey Cercopithecus mitis (0.35 groups/km). We sighted the endemic Sanje mangabey Cercocebus galeritus sanjei and the yellow baboon Papio cynocephalus infrequently, the latter being confined to the deciduous forest parts of the transects.We analyzed sighting frequency by gross habitat type, transect, season, and observer. Interobserver differences in the relative abundance of each species were moderate and the few cases of significant variations were due to discordance of only 1 observer from the others. Estimated distances of primate group sightings differ significantly among observers, thus preventing us from deriving estimates of absolute density. Frequency distributions of distance-class intervals are not significantly different among observers, which may indicate gross interobserver consistency in the width of the area sampled. We conclude that unless consistency in data collection is checked, as we did for 2 observers who collected data simultaneously, potential interobserver differences remain an underlying source of variance in the results that cannot be separated from other sources of variance.

No abstract