The field of Evolution Examines How it Got So Good at Evolving

[Chip (OP)]

https://phys.org/news/2025-02-evolution-good-evolving.html

15 Feb, 2025

The field of evolution examines how organisms adapt to their environments over generations, but what about the evolution of evolution itself?

Researchers have long questioned why biological populations are so good at exploiting their environments—a trait called "evolvability." Think, for example, of antimicrobial resistance and the speed with which new viral pathogens change and are able to evade vaccines.

Now, a University of Michigan study shows that perhaps why evolution is so effective is that evolution is itself something that can evolve. The research is published in the Proceedings of the National Academy of Sciences.

"Life is really, really good at solving problems. If you look around, there's so much diversity in life, and that all these things come from a common ancestor seems really surprising to me," said Luis Zaman, an evolutionary biologist at U-M and lead author of the study. "Why is evolution so seemingly creative? It seems like maybe that ability is something that evolved itself."

Whether evolvability itself can evolve is a question, Zaman says, because the major fuel of evolution are the mutations that increase an organism's fitness, increasing their ability to survive in the current environment, the most. But evolvability is not about increasing fitness. Instead, evolvability increases the future potential of an organism's fitness.

"This forward-looking feature of evolvability makes it contentious," Zaman said. "We think it's important. We know it happens. Why it happens and when it happens is something we're less clear about. We were trying to figure out: Can we see the evolution of evolvability in a more realistic computational model?"

To test these ideas, Zaman and his co-authors built a computational model composed of a set of three rewarded logic functions and three toxic logic functions.

You can think of the logic functions as red and blue berries, which are beneficial or poisonous in different environments, the researchers say.

In one environment in the model, red berries are beneficial to a population, but blue berries are poisonous. In the other environment, blue berries are beneficial to the population, but red berries are poisonous. This means that a population can't be "good" in both environments—it can succeed only in one environment or the other, Zaman said.

The researchers then ran a series of scenarios and recorded how evolvability might change over the course of each scenario. In one, the environments remained constant: the population never switched between eating red berries or eating blue berries. In another scenario, the population cycled between having to eat red berries and blue berries.

The researchers found that when they "cycled" between these two environments, the populations in each environment were to be able to jump back and forth between these opposite environments and be successful in both.

In particular, cycling between environments caused the populations to have a thousandfold increase in mutations that would allow them to successfully switch between eating red berries and blue berries in each environment.

Adapting to a mutational neighborhood:

The computational framework the researchers used to examine evolvability is called Avida. When the researchers created scenarios using Avida that cycled between each logic function (represented by red and blue berries), the programs nudged themselves into new mutational neighborhoods.

You can think of the evolved computer programs as pathways of multiple genes made up of computer codes, Zaman says. Each time the environment fluctuates, this pathway needs to be reconfigured to eat the new berries.

"The mutational neighborhood that populations end up occupying—finding through evolution—are places where single mutations are able to reconfigure this pathway," he said.

Mutations occur when one of those computer instructions (genes) within the program (genetic pathway) is changed.

Over time, this reconfigures the pathway, ultimately allowing the population of computer programs to live successfully in a neighborhood where red berry and blue berry specialists live next door.

The researchers also changed the frequency with which they cycled between environments:

They looked at the outcomes of when a population spent one generation in an environment before it changed, compared to 10 generations, compared to 100 generations. They found that if the environment fluctuated too quickly, they didn't see an increase in evolvability. But what was interesting was that even relatively long cycle periods—hundreds of generations—could lead to the evolution of and maintenance of evolvability.

"Once a population has achieved this evolvability, it seems like it didn't get erased by future evolution," Zaman said.

This implies that once evolution evolves to be better at evolution, that evolvability is here to stay.

Key results for three environments. Credit: Proceedings of the National Academy of Sciences (2024). DOI: 10.1073/pnas.2413930121