Two years of painstaking observation on the social interactions of a troop of free-ranging monkeys and an analysis of their family trees has found signs of natural selection affecting the behavior of the descendants.

Rhesus macaques who had large, strong networks tended to be descendants of similarly social macaques, according to a Duke University team of researchers. And their ability to recognize relationships and play nice with others also won them more reproductive success.

"If you are a more social monkey, then you're going to have greater reproductive success, meaning your babies are more likely to survive their first year," said post-doctoral research fellow Lauren Brent, who led the study. "Natural selection appears to be favoring pro-social behavior."

The analysis, which appears Wednesday in Nature Scientific Reports, combined sophisticated social network maps with 75 years of pedigree data and some genetic analysis.

The monkeys are a free-ranging population of macaques descended from a 1938 release of monkeys from India on undeveloped 38-acre Cayo Santiago Island, off the eastern coast of Puerto Rico. They live in a natural setting with little human intervention other than food provisioning, but they do have university students watching them a lot of the time.

Field researchers who had learned to identify each of the nearly 90 monkeys on sight carefully logged interactions between individuals in 10-minute episodes over a two-year span. They compiled four or five hours of data per individual, logging grooming, proximity and aggression.

From that, the team built web-like network maps to analyze pro-social and anti-social interactions. They also looked at the maps for a measure they called "betweenness" -- the shortest paths between individuals -- and "eigenvector," a friends-of-friends measure that shows how many friends each friend of an individual has.

"The really 'popular' monkeys would have a high eigenvector, or a really big friends-of-friends network," Brent said. There were also less-popular outliers who had fewer social interactions and a lower eigenvector. "They're sort of the dorks," Brent said.

When these measures were then compared to family trees, "a lot of these network measures popped out as having significant heritability," Brent said. That is, the behaviors seemed to run in families.

"This is really a landmark paper," said James Fowler, a professor of medical genetics and political science at the University of California-San Diego who studies human social networks, including Facebook, but who was not part of the study. "They're showing that the positive behaviors which build social networks might be heritable, and that's consistent with what we've been seeing in human studies."

The analysis of aggression didn't reveal much heritability, but it did influence reproductive success. At either end of the aggression scale, monkeys who were the most aggressive and those who were the most passive had better reproductive success than the monkeys in the middle.

The team also collected blood samples and did some genetic analysis on two genes in the serotonin system of the monkeys. Variability in the two genes -- one that makes serotonin and one that carries it around -- was most closely associated with differences in grooming connections between the monkeys.

They chose to focus their genetic analysis on two genes in the serotonin system because there is a lot of literature on that area in humans. Serotonin, a molecule that carries signals between nerve cells, is part of the system acted on by antidepressant drugs, so it has been widely studied.

"The way that genes can affect behavior is by their influence on neural circuits," said Michael Platt, director of the Duke Institute for a Brain Sciences and the Center for Cognitive Neuroscience. "We know that neural circuits for a variety of things like social behavior, food and mood are under the influence of serotonin signaling, in both humans and monkeys."

Genes by themselves don't determine your social standing, Platt added. But social success comes from some combination of social skills and temperament, which appear to have a genetic basis.

"We can see that some of these behaviors have a genetic basis, from what we know of the pedigrees and the network map," Brent said. "But we've only scratched the surface of figuring out which specific genes are associated with each behavior."

Fowler said the article is especially interesting coming on the heels of a study in Nature last year that showed hunter-gatherer networks are not very different from those in modernized human societies. "So now the conversation is about where to draw the line -- how far back did our networks evolve?" Fowler asked. "This paper suggests it may have been a common ancestor with macaques."

Platt's group recently won an additional five years of funding from the National Institute of Mental Health to continue and expand the study. Social network observations are now being done on other troops of monkeys on the island and the blood that has been collected will be subjected to further genetic testing.

Having 75 years of family history, combined with the latest genetic tools and a lot of observational data, is going to open up all sorts of new questions, Platt said. "This is just the first two genes," he said. "We'll hopefully be moving on to sequence the entire genome of each animal" to find even more associations.

"This is the first major part of what will hopefully be a very big puzzle," Brent said.

"Genetic origins of social networks in rhesus macaques," Lauren J. N. Brent, Sarah R. Heilbronner, Julie E. Horvath, Janis Gonzalez-Martinez, Angelina Ruiz-Lambides, Athy G. Robinson, J.H. Pate Skene and Michael Platt. Nature Scientific Reports, 9 January 2013. DOI: 10.1038/srep01042