Comparative insights into complex adaptive systems: from cells to societies

Corina E. Tarnita @ Princeton

Abstract:
My research examines the organization and emergent properties of complex adaptive systems at multiple scales, from single cells to entire ecosystems. Central to our approach is the development of general theoretical frameworks. Simultaneously, we use empirical data to identify and catalog patterns in nature and, within the general frameworks, we develop models whose predictions we attempt to empirically test using eco-evolutionary experiments, molecular and genomic analyses, and field manipulations.

My approach is mainly theoretical and combines evolutionary dynamics, evolutionary game theory and elements of network theory but my lab works in collaboration with experimental and field ecologists, molecular biologists and evolutionary biologists to integrate modeling and empirical work. The questions I'm interested in range from multicellularity to social behaviors in bacteria, insects or humans, to the effects of population structure and spatial patterns on evolutionary and ecological dynamics, to mutualistic interactions especially in the context of multi-species networks of symbionts.

Bio:
Corina is a Professor of Ecology and Evolutionary Biology at Princeton University. She joined Princeton in February 2013 after being a Junior Fellow at the Harvard Society of Fellows from 2010 to 2012. She obtained her B.A.('06), M.A.('08) and PhD ('09) in Mathematics from Harvard University. She is an Alfred P. Sloan Research Fellow, a Kavli Frontiers of Science Fellow of the National Academy of Sciences, and an Early Career Fellow of the Ecological Society of America. Her work is centered around the emergence of complex behavior out of simple interactions, across spatial and temporal scales.

Focus: complexity of living organisms and life overall
- Evolutionary substrate: Biochemistry, Biology, Culture, Technology
- Major transitions in evolution: RNA, DNA, Cells, Brains
Looking at hierarchical major transitions: smaller units grouping into larger units, which become a new dominant unit of organization
- Start: small groups of larger individuals
- Enabled by some change in small groups that allowed them to persist or displace their smaller ancestor
- E.g. multi-cellular organisms arising from single-celled, ant colonies arising from more independent individuals
Base Model
- We have a set of individuals who are stressed
- They need to coordinate to survive
- Need mechanism to control/eliminate free-riders
The evolution of Eusociality: social organization observed in ants/wasps/bees
- Why do ant workers support other ants and rear their queen’s children?
- This is a very successful strategy as ants are extremely numerous
- But this strategy is not common across species
- If it is so effective, why not?
- Ideas
  - Kin selection: help closer relatives more than further relatives
    - Worker ants pass their genes on via their mother
    - But they’re not that related to the males
    - Why would they sacrifice reproduction?
  - Another idea: if a mother was able to make some daughters stay behind to help, would that be favored by natural selection?
    - Option 1: Mother leaves eggs and they develop on their own
    - Option 2: Mutation that causes some daughters to stay at the nest (may or may not reproduce)
    - Assumptions on costs and benefits of sociality: reproductive rate and expected lifetime

- - - Social mother is less fit when she’s on her own (because it needs to maintain eggs on her own)
    - Fitness rises with size of colony
    - Eusociality is bad if:
      - Too few staying daughters (not useful)
      - Too many staying daughters (colony doesn’t spread)
      - Favored by natural selection in middle range of colony migration probability
      - Critically, eusociality can evolve only if benefits kick in at VERY small group sizes

Experiment on real colonies of ants
- Ooceraea biroi
  - Queen-less
  - All ants lay eggs, then all become workers to take care of the eggs
  - Reproduce asexually (everyone is clones)
  - Genetics of individuals are very tightly controlled
  - Size of colony can be controlled by number of larvae
- Goal: understand how group size affects group dynamics
- Result: colony of size 6 individual has same growth as larger-size colonies
- Dynamics: there was a rudimentary division of labor
  - When there are <=4 ants, each ant does all tasks and keeps switching
  - When there are >=6 ants each ant does more of a single job
    - Each ant has a probability distribution of its task that becomes much more concentrated on one task
    - Distributions of different ants concentrate on different tasks
  - How do they decide on their jobs?
    - Individuals are not genetically identical and are more sensitive to different stimuli (larvae asking for help, trash piling up, etc.)
    - For ants transcriptomics and neuroanatomy affect sensitivity of different odorant receptors
    - Evolutionary studies show that basic genes that affect multi-cellularity evolved before animals and fungi evolved
  - Division of labor ensures that tasks don’t get neglected (each task gets more dedicated attention) and so eggs develop faster
  - Modeling studies that simulated larger group sizes show that this mechanism doesn’t improve growth beyond 20.
  - But, now we have small groups running around. What if they evolve the ability to sense each other?
    - Add interaction bias: individuals who specialized in the same task tend to stay together and focus more on a common task
    - Models of this mechanism show that division of labor increases dramatically
  - Observation: the above mechanisms have been proposed for driving political polarization