Donald Stone Sade

Patterns of group splitting within matrilineal kinship groups

Summary1.When social groups of free-ranging rhesus monkeys on Cayo Santiago, Puerto Rico, undergo fission, they usually divide between genealogies.2.If a genealogy divides, it is usually between an eldest daughter with her family and the rest of the genealogy.3.The separation of the eldest daughter from her genealogy is the extreme case of peripheralization of low-ranking females among rhesus monkeys.4.The founders of new groups that disperse from the former home range are likely to be subordinate individuals in the parent group, as predicted by Christian (1970).5.The dispersal of families as units is likely to lead to ‘lineal effects’ (Neel and Salzano, 1967) in the genetical substructure of the population.

Genetic differentiation among matrilines in social groups of rhesus monkeys

SummaryIn rhesus monkey populations, animals related by descent to some female comprise a matriline or genealogy. Data on blood protein polymorphisms in the Cayo Santiago rhesus colony indicate that allele frequency variations among matrilines in social groups are large. These variations occur despite high levels of outbreeding. Computer simulation analyses indicate that pedigree or linear effect account for much of the observed genetic differentiation among genealogies. A sampling with correlation model in which genealogy sizes and average kinship levels are parameters predicts among matriline genetic differentiation. This study indicates that substantial genetic substructure is present within rhesus social groups. Our analyses also predict that large variances in allele frequencies should be common among social or trait groups based on kinship relationships.

Sociometrics of Macaca Mulatta III: n-path centrality in grooming networks

Abstract An individuals “status” is the position in a social network, which is usefuly modeled by the mathematical properties of directed graphs. Katz (1953) argued that an index of status based on “who chooses whom” should also include information on the status of the choosers. This recursive concept os status was adapted by Sade (1972), who used the sum of inward directed paths as a measure of centrality to distinguish the position of individual monkeys in a grooming network. Sade et al. (1988) used the same measure to describe the progression through time of the statuses of a set of male monkeys. These studies were limited by the perceived computational difficulties in finding paths longer than 3-paths. Although ad hoc formulae for up to 6-paths were available from Ross and Harary (1952) and a general algorithm was known from Parthasarathy (1964) the methods were cumbersome and impractical to apply to large networks. The present paper makes use of a simple counting algorithm to find the n-paths in a binary matrix. The characteristics of centrality statuses in the grooming network are then describing using paths of up to maximum length, thus avoiding the limitations of the earlier studies.

Rank-Related Fitness Differences and Their Demographic Pathways in Semi-Free-Ranging Rhesus Macaques ( Macaca mulatta )

Researchers have explored the fitness consequences of female dominance hierarchies in many primate populations, with most studies highlighting differences in age of maturation, fertility, and offspring survival. We use resampling techniques and van Tienderen’s (2000) elasticity path analysis to identify rank-related differences in finite rate of increase (λ) and their demographic correlates among segments of a semi-free-ranging rhesus macaque population. Higher-ranking population segments grew at greater rates for some portions of the 40-yr study period. The female members of these segments achieved these lifetime fitness differences through higher fertility and especially higher adult survival rates. This is the first clear evidence that social rank influences female primate adult survival, and is a crucial fitness component for any long-lived, slow-reproducing animal. Traditional methods of comparing lifespans, and other life history variables, among rank categories fail to identify most of the rank-related differences primarily because they require completed life histories that are available only on a small number of the females known in the population.

Sociometrics of Macaca Mulatta IV: Network analysis of social structure of a pre-fission group

Abstract Cluster analysis is applied to the grooming network of a group of free-ranging rhesus monkeys undergoing group fission to examine the social structure of the pre-fission group. A matrix of grooming interactions was compiled from detailed field notes collected over a 6-month period, during the mating season of 1972. The group underwent fission at the beginning of the 1973 mating season. The network analyses, based on an algorithm developed by Mizoguchi and Shimura (1980), and adapted by Karl Reitz for application to social structures (Reitz 1982, 1988) are designed to detect natural hierarchically arranged clusters of individuals within a group. The resulting sociograms provide measures of the cohesiveness of a group as a whole, and show how smaller clusters of close grooming partners are grouped into larger clusters within the group based on less frequent grooming interactions. The results of the network analyses are discussed in light of behavioral and demographic observations of the groups structure over the study period, and are found to compare well with the observers intuitive understanding of the social structure of the group as described in Chepko-Sade and Sade (1979).

Social behavior of the emperor tamarin in captivity: Components of agonistic display and the agonistic network

Agonistic behavior was studied longitudinally for 16 months in an intact family groups of captive emperor tamarins (Saguinus imperator subgrisescens) using methods from quantitative ethology and social network analysis. A motivational analysis of the components of agonistic display revealed the relative strength of each component along a continuum from strongly dominant to strongly subordinate. Tabulations of exchanges of strongly dominant and strongly subordinate components in interactions among the tamarins revealed an agonistic network (“dominance hierarchy”) that approached, but did not quite reach, the ideal state of a transitive order (“linear dominance hierarchy”). The frequency with which individual tamarins long called and scent marked was not closely correlated with their position (“dominance rank”) in the agonistic network. Instead, individuals undergoing change in status long called and scent marked frequently, irrespective of their rank.

Natural History of the Self

For a generation, a dominant theme in behavioral primatology has been the interpretation of social interactions as fitness strategies. This approach seeks the ultimate causes of sociality in adaptation through natural selection. Studies from Cayo Santiago have contributed to this approach in a number of studies of fitness consequences of agonistic status of females, most recently and extensively by Blomquist (2007, 2009) and Blomquist et al. (2011), and of males (Berard et al. 1993, 1994). Here a second, yet complementary, approach, which seeks the proximate causes of behavior in the organization of the neural mechanisms of perception and action, is taken. Each approach makes inferences about processes that are not directly observable, in the first case natural selection, in the second the activity of neural networks in the brain.