About This Project

Ask the Scientists

What is the context of this research?

In animal societies, a cohesive group is often much more than the sum of its parts. Decisions such as where to forage, whom to fight, and when to mate, can have dramatic consequences for the entire social group. Individual animals comprise social groups, yet groups themselves exhibit emergent, collective behavior such as flocking, migration, and herding. These emergent behaviors are thought to arise from individuals following simple “rules,” but few studies examine how social relationships may mediate emergent behaviors. The nexus of social behavior and group-level dynamics is a “black box” in behavioral ecology. Through this research, I propose to open this “box” and investigate how social relationships mediate collective movement and herding in a complex primate society.

Recent advances in computational methods have allowed biologist to uncover the rules that govern complex patterns of emergent behavior. However, the role of social relationships in shaping emergent behaviors remains largely unexplored. Network analysis is a powerful tool for understanding interactions between individuals at all levels and across all behaviors. Behavioral ecologists have largely relied on static network analysis to understand social interactions, but these methods ignore information about the order in which interactions occur. Dynamic network algorithms are an extension of traditional static network analysis and allow biologists to understand how socioecological pressures affect the order and patterning of interactions. Here, I will use dynamic network analyses to study the socioecological pressures that shape the collective movement and herding of the gelada (Theropithecus gelada)—a primate that lives in a complex modular society.

In modular societies, core groups of animals split off and rejoin with each other over time to create structured levels. The relationship between intra- and inter-group behaviors is central to understanding the selective pressures that maintain the structure of modular societies and give rise to patterns of emergent behavior, such as collective movement. Collective movement is a self-organizing behavior, where individual decisions about “where to go” and “who to follow” influence motion of the entire social aggregation. Within core groups, individuals might follow dominant individuals or close social partners. This becomes less clear for individuals in modular societies where different levels of association may force individuals to make movement decisions based on the behavior of unknown individuals. Between core groups, local “rules” may predominate an individual’s decision to follow others. It is not known if intragroup properties (roles, status, affiliative relationships) influence intergroup behaviors, e.g. collective movement and herding.

Geladas as a model species

Geladas are an excellent candidate species to examine the relationship between ecological pressures, social relationships, and collective behavior in a modular society. Geladas form complex, stratified social relationships with each other that are easy to quantify through proximity, grooming, and dominance interactions. Additionally, geladas live in an Afroalpine habitat that experiences pronounced seasonal fluctuations in temperature and food availability that may influence movement and association patterns.

Most importantly, geladas live in a modular society with strong fission-fusion dynamics. This society can be organized in to 4 levels: (1) the reproductive unit: a core group of 1-12 related females with their dependent offspring, a dominant “leader” male, and possibly 0-4 subordinate “follower” males that are found together 100% of the time; (2) the team: two to three units that spend at least 90% of their time together; (3) the band: a collection of 5-30 units that spend at least 50% of their time together; and (4) the community: a collection of all the units that share a home range (>100 units) and spend at least some time together. Herds are temporary aggregations of multiple units and all-male “bachelor” groups from several different bands. Herds may persist for a few hours to several days and do not always contain all the units of a given band. Reproductive units overlap spatially within the herd, but bachelors are often found >20 m from the edge. While these levels have been described for nearly 30 years, previous studies of gelada society have ignored how flexible these levels are over time.

What is the significance of this project?

This will be one of the first studies to examine how socioecological factors mediate collective movement and herding in a modular society using dynamic networks. Previous approaches to these problems have relied on either static association networks or one-dimensional measures, such as group size. These methods ignore the ordering and patterning of interactions among conspecifics, both of which may influence collective movement and herding. Through using dynamic networks, my research will transform how biologists understand the adaptive value of association in changing environments. Potential findings will be of great interest to computer scientists and behavioral ecologists interested in the evolution of cooperation, coordination, and complex systems. Moreover, my research will be integral in understanding important evolutionary processes in dynamic systems such as disease and information transmission, population genetic structure, frequency dependent selection, and social niche construction.

My research will help conserve the Afroalpine habitat of the Simiens Mountains National Park (SMNP), Ethiopia. The park is a World Heritage Site “In Danger” (1996) and is home to several threatened and endangered species including the Walia ibex (Capra walie), the Ethiopian wolf (Canis simensis), and the giant lobelia (Lobelia telekii) (IUCN). I will give public talks on conserving Ethiopian biodiversity at a local ecotourism lodge and train Ethiopian nationals as field assistants to collect behavioral, ecological, and GPS data for the duration of this project. I anticipate giving talks to high school biology students in New Jersey and Illinois. Additionally, I maintain a science blog on my website and an “open lab notebook” which details my research to an online audience. When possible, I will use open-source, cross-platform software in all my analysis and will make all code freely available online upon completion of the research project.

What are the goals of the project?

This project combines field-based data with dynamic network algorithms to understand how social relationships mediate collective movement and herding in gelada society. I have three main research questions:

1. What influences intragroup movements?

I expect that intragroup movements will largely be determined by social relationships. Within a core group, “leader” males are dominant to all other individuals, and females form a linear dominance hierarchy. Individuals may exhibit social preferences (through grooming and proximity) for each other that could influence dyadic leadership patterns and cohesiveness during movement. I predict dominant males, followed by dominant females, will exhibit the highest centrality and individual connectivity measures for intragroup leadership networks. I also predict that individuals will prefer to follow close social partners, thus grooming and leadership network metrics will be correlated. Alternatively, energetic costs of female reproduction may strongly influence leadership patterns within core groups. Therefore, I predict that females when lactating (i.e. highest energetic demands) will exhibit high leadership network centrality and connectivity.

2. What influences intergroup movements and herding patterns?

Collective movement of the gelada herd is the result of individuals leading or following extra-group (possibly unknown) individuals. Thus, it is unlikely that social relationships guide decisions at the highest levels of gelada society. Therefore, I expect that intergroup movements will largely be determined by local rules, such as attraction, avoidance, and alignment. The presence of core groups in the same herd may be influenced by both food availability and familiarity (i.e. time co-resident in the herd). I predict core groups of the same “band” to exhibit high group connectivity. I also expect group connectivity to be influenced by food availability, i.e. groups within herds will be less connected (for spatial and leadership networks) as food availability decreases.

3. How do social relationships and local rules interact to shape collective movement?

I expect that both social factors and local rules will interact to shape movements in specific situations. For example, “teams” may maintain proximity within herds, suggesting that individuals are responding to the movements of extra-group team members. Thus social relationships may influence intergroup movements within teams. These patterns may extend to a lesser degree for bands, such that leadership patterns are hierarchically organized within the herd. By comparing the movements within teams to both intragroup movements (dominated by social relationships) and movements between units from different bands in the same herd (dominated by local cues) I can determine how individuals combine different types of information when making movement decisions. Given the spatial overlap of gelada core groups, it is possible that peripheral individuals, i.e. those with low intragroup centrality for affiliative networks, might be highly influential in intergroup movements. Under this hypothesis, I predict peripheral animals will exhibit the highest individual connectivity in a leadership network for the entire gelada herd. Alternatively, intragroup leaders might be highly influential in influencing intergroup movement, especially at large herd sizes. In this scenario, I predict intragroup leaders to exhibit high individual connectivity in dynamic leadership networks for the entire gelada herd.

To complete this research, I will use GPS technology and satellite imagery combined with behavioral data collected by myself and field assistants. In Ethiopia, I am associated with the University of Michigan Gelada Research Project (UMGRP). Scientists of the UMGRP have been studying a population of wild geladas living in the Simien Mountains for nearly 7 years. I previously conducted my dissertation research in this remote location and am familiar with the logistics of working in Ethiopia. Moreover, I am currently affiliated with Princeton University as a postdoctoral researcher in the Ecology and Evolutionary Biology Department. As such I am confident in my ability to complete this proposed project.

With your support, I aim to take a trip to Ethiopia in February or March 2014 to begin my first phase of data collection. Although this initial pilot research will take only a few months, I plan on returning to Ethiopia next year after acquiring additional funds. Although I only have Internet access once per week in Ethiopia (during our weekly “town” trips), I will keep you all informed of the research progress through this site, as well as my personal site and Twitter (@djpappano).