My research has long focused on understanding how simple processes can produce the amazing levels of complexity and diversity we see in nature. This past week, we had a paper appear in PLoS Biology that I am particularly pleased with. In it, we use digital organisms to explore how the interactions between hosts and parasites can promote the evolution of new complex traits, even when those traits would otherwise be costly.
Zaman L, Meyer JR, Devangam S, Bryson DM, Lenski RE, and Ofria C (2014) Coevolution Drives the Emergence of Complex Traits and Promotes Evolvability. PLoS Biology 12(12): e1002023. doi:10.1371/journal.pbio.1002023.
Researchers have long understood that coevolution produces rapid evolutionary changes: parasites race to find new mechanisms to infect hosts and in turn those hosts are pressed to keep evolving new defenses, just to survive. This effect was dubbed the red-queen hypothesis, based on the Red Queen’s famous line to Alice in Through the Looking Glass: “Now, here, you see, it takes all the running you can do, to keep in the same place.”
Evolutionary algorithms frequently use the Red Queen effect to promote rapid exploration and increase the probability that better solutions will be discovered. In such cases, it’s not surprising when complexity rises in association with new traits that provide a big fitness boost.
More surprising is the fact that coevolution can still produce complex traits even when they come at a fitness cost. Normally, coevolution is thought to be helpful since it produces more diverse populations (we’ve also seen such diversity increases in Avida ). More diversity means that the populations try out more possible solutions at a time, any one of which might be more beneficial than anything discovered before. However, in the experiments described in our paper, the only advantage of a more complex trait is that it forces a parasite to evolve the same trait in order to infect this host; therefore it is harder for a parasite to evolve the infection mechanism when hosts rely exclusively on more complex traits.
I’m excited to explore the interesting implications of this result on evolutionary computation systems. When coevolutionary dynamics are introduced into a system, the selective pressures on organisms change beyond simply forcing hosts to continually move around on the fitness landscape. If some portions of the landscape are trickier for parasites to access, they inherently have more of an advantage than their direct fitness might suggest. Such effects may either help or hurt the hunt for the best solutions depending on the system’s details, but more study is certainly warranted.
The other interesting result that we present in this paper is that the genetic architecture of the organisms evolved in conjunction with parasites resulted in a different distribution of mutational effects. Specifically, coevolved organisms were substantially more likely to gain mutations that replace an old task with a new one, without changing the total number tasks that an organism does. This new distribution of mutational effects are far more likely to aid an organism in escaping from parasites – organisms need to perform at least one task to obtain resources, but doing multiple tasks creates more opportunities for parasites to infect them. The evolution of this new genetic architecture is surprising because the value of a certain distribution of mutational effects takes several generations to be felt; that is, the real benefits are only realized once mutations actually occur.
Such delayed benefits are why so many types of evolvability are difficult to see. In our own work, we’ve previously used changing environments to try to encourage organisms to raise their mutations rates  or evolve sexual recombination , but in both cases we failed to show any strong improvement in evolutionary potential. As such, we were excited to see that parasites can produce an increase in the evolvability of hosts. We currently have studies underway to see if parasites can similarly push asexual populations to become sexual. Our initial results are already intriguing and we hope to have something exciting to report in the near future.
We’d love to hear your opinions on this work! Also, stay tuned next week when Emily Dolson will post her seven-part series on work-life balance: “An Academic Christmas Carol”.
 Zaman L, Devangam S, and Ofria C (2011) Rapid Host-Parasite Co-Evolution Drives the Production and Maintenance of Diversity in Digital Organisms, Proceedings of the 2011 Genetic and Evolutionary Computation Conference.
 Clune J, Misevic D, Ofria C, Lenski RE, Elena SF, and Sanjuan R (2008) Natural Selection Fails to Optimize Mutation Rates for Long-Term Adaptation on Rugged Fitness Landscapes, PLoS Computational Biology, 4(9): e1000187. doi:10.1371/journal.pcbi.1000187
 Misevic D, Ofria C, and Lenski RE (2010) Experiments with Digital Organisms on the Origin and Maintenance of Sex in Changing Environments, Journal of Heredity, 101(supp 1):S46-54. doi:10.1093/jhered/esq017.