Wild marmots’ social networks reveal controversial evolutionary theory in action
- Written by Conner Philson, Executive Director, UCSB Natural Reserve System, University of California, Santa Barbara
It probably feels obvious that having a close friend can influence your well-being. But do the groups that you’re a part of also affect your well-being? For example, does the culture of your work colleagues influence your productivity?
It may seem like the answer is also an obvious “yes.” But the idea that a group’s composition or structure can affect the individuals in it has been among the most controversial ideas in biology.
This phenomenon, called multilevel selection, is an extension of natural selection[1]: the process by which organisms with traits better suited to their environment are more likely to survive and reproduce. Over generations, these advantageous traits – behavioral, morphological or physiological – become more common in the population.
In the traditional view of how evolution works, natural selection acts on an individual organism’s traits. For instance, mammals with more friends typically live longer lives and have more offspring[2]. The trait under selection in this case is the number of social connections.
Multilevel selection proposes that at the same time selection is happening on the traits of individuals, selection also acts on the traits of groups. Here’s an example: Living in a more social and interconnected group may be beneficial for the members of that group, meaning the group’s traits are under selection. In nature, this means individuals in well-connected groups may live longer lives and have more offspring because well-connected groups may be better at finding limited resources or detecting predators. The traits of the group as a whole are what’s under selection in this case.
Multilevel selection could even select for traits that seem at odds at the individual and group levels. For instance, it could mean that selection favors individuals that are more reserved while at the same time favoring groups that are very social, or vice versa.
Multilevel selection has been a controversial idea[3] since Charles Darwin first suggested that groups likely affect individuals in his 1871 book “The Descent of Man[4].”
The only evidence for multilevel selection acting simultaneously on individuals’ social relationships and on social groups comes from laboratory experiments[5]. Experiments like these are vital to the scientific process, but without evidence for multilevel selection in wild animals, the 154-year-old debate rages on. As two field biologists interested in the evolution of behavior[6], we investigated multilevel selection in the wild by studying yellow-bellied marmots[7].
Our newly published study[8] provides support for this contested concept, suggesting that the structure of the groups marmots are members of may matter for survival just as much as, if not more than, the friendly one-on-one relationships they have with other marmots.
Spying on marmots’ social lives
It’s taken a century and a half to answer the question of multilevel selection because you need an incredible amount of data to have an adequate sample size to address it.
Scientists at the Rocky Mountain Biological Laboratory[9] in Crested Butte, Colorado, have been studying the marmots nearby since 1962. This research[10] is the second-longest study of individually identifiable wild mammals in the world.
Each year, the team ensures that all marmots are individually marked. We trap them so we can give them unique ear tags and paint a mark on their back that lets us identify them from afar. Then trained “marmoteers,” as we call them, spend about 1,000 hours a year watching these chunky cat-sized rodents through binoculars and spotting scopes.
Since 2003, the team has paid particular attention to the marmots’ social interactions and relationships. Our analysis of multilevel selection was based on 42,369 unique affiliative social interactions – behaviors such as playing and grooming – between 1,294 individuals from 180 social groups, with group sizes ranging from two to 35 marmots. We also tracked how long marmots lived – up to 16 years in some cases – and how many offspring individual animals had each year.
Using this data, we mapped out the marmots’ social networks. Our goal was to identify how many social relationships each marmot had, who was connected to whom, and the overall structure of each group.
Understanding all these marmot connections let us ask two crucial questions. First, how do social relationships affect individual survival and reproduction – that is, what individual traits are under selection? Second, how do social groups affect individual survival and reproduction – in other words, what group traits are under selection?
Importantly, we didn’t ask these two questions in isolation – we asked them at the same time. After all, marmots are influenced simultaneously by both their social relationships and the social groups they’re part of. Our statistical approach, which researchers call contextual analysis[12], tells us how much social relationships and social groups matter relative to each other.
New evidence changes the debate
It can be tricky to distinguish how group-level selection differs from traditional individual-level selection. It’s like a more complex version of thinking about the relationships that affect an individual. Instead of just your own behavior affecting you, your group – a product of many individuals – is affecting you.
Our new analysis shows that there is indeed multilevel selection for social behavior in the wild[13]. We found that not only do both social relationships and social groups affect individual animals’ survival and reproduction, but social groups matter just as much, if not more. We calculated the selection gradient, a measure of how strong the selection is on a trait, to be 0.76 for individual traits, while for group traits it was 1.03.
Four juvenile yellow-bellied marmots play together.
D.T. Blumstein
Interestingly, the type of impact on survival and reproduction wasn’t always the same across the two levels. In some cases, selection favored marmots with fewer social relationships while favoring marmots living in more social and connected groups. In human terms, think of an introvert at a really bustling party.
Evolution and multilevel selection are complex natural processes, so these types of complicated findings are not unexpected.
Multilevel selection is relevant for human groups, too, which come in many forms, whether friend groups, local communities, businesses we frequent[14] or work at, economies[15] or even entire nations[16]. Our marmot study suggests it’s not uniquely human for groups at every level to have consequences for individual success.
References
- ^ natural selection (www.britannica.com)
- ^ longer lives and have more offspring (doi.org)
- ^ controversial idea (doi.org)
- ^ The Descent of Man (darwin-online.org.uk)
- ^ laboratory experiments (doi.org)
- ^ interested in the evolution of behavior (scholar.google.com)
- ^ studying yellow-bellied marmots (scholar.google.com)
- ^ Our newly published study (doi.org)
- ^ Rocky Mountain Biological Laboratory (www.rmbl.org)
- ^ This research (sites.lifesci.ucla.edu)
- ^ Maldonado-Chaparro et al, Behavioral Ecology, 2015. (doi.org)
- ^ contextual analysis (doi.org)
- ^ multilevel selection for social behavior in the wild (doi.org)
- ^ businesses we frequent (doi.org)
- ^ economies (doi.org)
- ^ nations (doi.org)
Authors: Conner Philson, Executive Director, UCSB Natural Reserve System, University of California, Santa Barbara
