How Chance Drives Evolution express

Original: "Stochastic individual-bases models"
· Anton Bovier, Anna Kraut
arXiv:2607.11520 · 2026-07-13 · CC BY 4.0 · 1 min
Scientists show how population size and mutation rate together shape evolution's path: from smooth changes to sudden leaps.

Evolution flows like water. In a wide river, changes are smooth and slow, while a mountain stream churns with every stone. Population size and mutation rate set the channel: broad and calm, or narrow and rapid-filled.

Scientists built a mathematical model based on the stochastic process theory of Andrey Markov, accounting for time, population size, mutation rate, and effect size. In large groups, evolution moves at a leisurely pace — mutations accumulate over centuries and mostly fade without a trace, like ink dropped in a river. In small groups, it surges: a beneficial mutation can sweep through the entire population in just a few generations, like a stone redirecting a stream.

In the micro-world of bacteria, this is vividly clear: in a trillion-strong colony, a helpful change can simply get lost, while in a test tube with a hundred individuals, it can become the norm within a day.

Such calculations allow us to predict how species will respond to environmental change. By managing the level of chaos in populations, we can slow unwanted evolution, say, in agricultural pests. Biology gains a tool for precise forecasts instead of vague generalities.

The mathematics of random chains that the scientists used was born outside biology: Andrey Markov created it by analyzing the alternation of vowels and consonants in Pushkin's poem 'Eugene Onegin'.

Scientists
Paul DiracEmilio Gino SegrèNancy Grace RomanPeter HiggsRichard FeynmanMarie Curie
Concepts
biological stochastic dynamics Stochastic modeling emergent complexity modeling sample size Markov chain probabilistic model Wright–Fisher model parameter complex system dynamics prognosis species evolution Bose–Einstein statistics Dynamical Systems exponent statistics overfitting scalable parameter fitting benchmark