The Sudden Surges That Forge Evolutionary Trees | Hacker News Reader

The Sudden Surges That Forge Evolutionary Trees

quantamagazine.org

37 pointsrbanffy7 comments

This isn't new, we've known for probably a hundred years and evolutionary theory has attempted to explain it by altering Darwin's original theory. Punctuated equilibrium is the term for this.

In evolutionary algorithms, fitness climbs rapidly from a random start, then tends to level off over generations. The "split-and-hit-the gas" dynamic makes sense if you think of a given phenotype occupying a certain phase space. If a sudden divergent mutation throws it into a new phase space, it's sort of resetting the evolutionary clock: fitness in the new dimensions of the space increases more rapidly, amplifying the divergence from the other evolutionary branch.

Nature has one thing most evolutionary algorithms don't: environmental changes. That has a way of stirring up the pot.

Which is not at all to say that you're wrong. A lot of the known major environmental changes were brought upon by other creatures - sometimes creatures that found some weird new way of doing things, and upended the environment with it.

You are 100% right. In nature, the fitness landscape is not fixed, it is dependent on the co-evolution of other organisms at every scale. So the computational model is too simple to be a reliable guide. But in this case, if you assume constant fitness over some delta-t (kind of like calculus), I think it's still a plausible explanation for the divergence phenomenon in particular. Just a hypothesis, though, for sure.

> “[The spike is] a novel contribution of this algorithm that is not

> usually done in phylogenetics,” Douglas said.

They built a model that incorporates a controversial non-standard dynamic and found that it exhibited the very processes that they added.

Showing an effect in computer simulations designed to produce exactly this effect is as bad as showing something "in mice"; in both cases, you generally get stories reporting the "results" with only a brief in-passing mention of the key caveat.

I wonder what causes the saltative branching events? Maybe sudden climate changes or a species somehow entering a new environment? Or it could just be a random 'spark' of mutation that sets off an explosive positive feedback loop?

While those mechanisms are involved the current theory is that usually a population becomes genetically isolated first. The larger and more genetically diverse a population is, the more resistant it is to “evolutionary noise” which keeps the genetics stable because there aren’t many mutations or ecological changes that give one group within a species significant advantages over another. It’s not until the relevant population is reduced that mutations and other factors have strong effects (usually) that can cause speciation.

How that isolation happens varies and can take a single generation or up to hundreds of thousands of years. A polyploid plant, for example, might become genetically isolated within a single generation or a homoploid hybrid within a few generations by losing reproductive compatibility with the rest of its species. Then a mutation might give it significant advantages without making it into the rest of the population or a “sudden” ecological change favors the new population over the old, giving the new one room to grow and outcompete.

Other species are isolated over “short” periods via flooding, rising mountains, changes in the paths of rivers, expansion of a predator’s range, fires, and so on. Anything that can isolate a small group of a species geographically can also create a speciation event.

Reply on news.ycombinator.com