The Physics of Inheritance: How Evolution Obeys Le Chatelier's Principle
The lizard had barely a week to adapt.
Off the coast of Italy, a population of wall lizards was transplanted to a new island in 1971. The move was an accident — a scientific expedition's unintended experiment. Within a single generation, these reptiles had dramatically altered their bite force, jaw shape, and digestive tract to better consume the harder, more abundant prey available on their new home. Within five years, their color patterns had shifted. Within a decade, researchers could visually distinguish island populations from their mainland ancestors.
Evolution, it turns out, can happen fast. Much faster than Charles Darwin imagined.
This speed puzzled biologists for decades. Classical evolutionary theory, built on the idea of random mutations being slowly filtered by natural selection, predicts change measured in hundreds or thousands of generations. But real organisms — from these lizards to bacteria to humans — often transform far more rapidly than that. Something else must be at work.
The answer, increasingly, lies in a pair of biological superpowers that organisms have possessed for hundreds of millions of years: phenotypic plasticity and genetic assimilation. Plasticity is the ability to change your body in response to environment — the way your muscles grow when you exercise, or how a plant bends toward light. Genetic assimilation is what happens when those environmentally induced changes gradually become inherited, baked into the DNA itself across generations.
Together, these processes offer a shortcut through evolution's usual slow lane. But despite decades of evidence for both, no one had successfully married them into a single, coherent mathematical framework — until now.
A new paper by Tetsuhiro S. Hatakeyama, published on arXiv in July 2026, cracks this problem wide open. By bridging a foundational equation of evolutionary biology with a concept from physics — the mathematics of Le Chatelier's principle — Hatakeyama has created something remarkable: a unified theory that explains how organisms can shift their phenotype rapidly, then slowly lock those changes into their genome. It is, at its heart, a theory about the physics of inheritance. And it carries implications far beyond any single species.
The core insight is this: when an environment changes, organisms don't just sit and wait for mutations. They immediately shift — bending, growing, recalibrating. This plastic response, Hatakeyama argues, creates a kind of internal stress within the organism, a tension between what the phenotype is doing (adapting fast) and what the genotype is doing (lagging behind). Evolution, he shows mathematically, is the slow process of resolving that stress. And the way it resolves it follows a rule familiar to chemists and physicists: systems naturally push back against disturbances. This is Le Chatelier's principle — and it turns out that evolution obeys it too.
The Science
To understand Hatakeyama's contribution, we need to understand the two pillars he builds between.
The first is the Price equation. Named after the British evolutionary biologist George Price, who developed it in the 1970s, this equation is one of the most general formalisms in evolutionary theory. It describes how any trait — from height to beak shape to cooperative behavior — changes across generations. The equation captures both the direct action of natural selection and the indirect effects that come along for the ride, like the genetic consequences of environmental change.
The Price equation has a simple beauty. It says that the change in a trait across one generation equals the covariance between an organism's genetics and its fitness, plus any transmission bias — essentially, whatever isn't explained by genes and selection gets thrown into that second term. It's an accounting identity for evolution, flexible enough to handle everything from altruism to sexual selection to, crucially, phenotypic plasticity.
But the Price equation, in its standard form, treats time as discrete — one generation to the next. To handle plasticity and assimilation, Hatakeyama needed to work in continuous time, watching the dynamics unfold moment by moment rather than in generational snapshots. This is a technical move, but it matters: it lets him write down differential equations that describe how phenotype and genotype evolve together, simultaneously but at different speeds.
The second pillar is timescale separation — a concept borrowed from physics and applied mathematics that describes systems where some things change quickly and others change slowly. A pot of boiling water has molecules moving fast (kinetic energy) and the temperature rising slowly (thermal equilibrium). A swinging pendulum has its position changing quickly and its energy more gradually. In each case, understanding the system requires treating these timescales as decoupled: fast variables reach their equilibrium almost instantly, while slow variables drift toward theirs.
Biological systems are made of timescale separations. Organisms breathe in seconds, grow over months, evolve over millennia. Within a single generation, the phenotype — the actual body that an organism builds — can change rapidly through plasticity: muscles remodel, hormones adjust, gene expression shifts. The genotype — the underlying DNA sequence — changes only when mutations arise and spread, a process that typically takes many generations.
Hatakeyama's key move is to formalize this separation mathematically. He treats the phenotype as a fast variable and the genotype as a slow one, then asks: what happens when an environmental shock hits a system with this structure? What equations govern the coupled dynamics? And what does the long-term outcome look like?
To answer these questions, Hatakeyama worked with a specific model of genotype-phenotype coupling. The genotype encodes information that influences the phenotype, but the phenotype also responds to environmental cues independently of the genotype — that's plasticity. The two are linked but not identical. Hatakeyama modeled this coupling using what dynamical systems theorists call a "landscape" — a mathematical surface describing how the phenotype sits in relation to the underlying genotypic coordinates.
When the environment shifts, the landscape tilts. What was once optimal for survival now isn't. The phenotype responds quickly, sliding down the new landscape to a better position. But the genotype, being slow, hasn't moved yet. The result is a mismatch: the phenotype is where it needs to be, but the genotype is still back where the old environment was optimal. This mismatch, Hatakeyama argues, is what he calls internal genetic stress — a kind of evolutionary tension that builds up when phenotype and genotype are out of alignment.
The mathematics of this stress and its resolution are the heart of the paper.
What They Found
The results fall into several interconnected pieces, each building on the last.
The core equation. By extending the continuous-time Price equation to handle phenotype-genotype coupling with explicit timescale separation, Hatakeyama derives a coupled system of differential equations. The fast equation describes how the phenotype responds to environmental cues and genetic instruction; the slow equation describes how the genotype responds to selection pressures generated by that phenotype. These equations aren't new in isolation, but their combination — and what Hatakeyama extracts from them — is.
Genetic assimilation as relaxation. The first major result is the reframing of genetic assimilation. Rather than treating assimilation as a somewhat mysterious process by which plastic responses become genetically encoded, Hatakeyama shows it is simply the dynamical relaxation of the genotype toward a new equilibrium. The fast phenotypic shift generates stress; the slow genotypic evolution is the system resolving that stress. There is no special mechanism required. The assimilation we observe — traits that were once inducible becoming constitutive — is what physics predicts when a coupled system with two timescales is perturbed and allowed to settle.
This reframing matters enormously. It takes a phenomenon that seemed biologically particular — assimilation is a quirky edge case, right? — and shows it to be a universal prediction of any system where phenotype evolves faster than genotype. If life has these two timescales — and it always does — then assimilation is not optional. It is guaranteed.
Le Chatelier's principle. The next result is where things get genuinely surprising. In chemistry, Le Chatelier's principle states that when a system at equilibrium is disturbed, it will respond in a direction that partially offsets the disturbance. Add heat to a chemical equilibrium, and the system will absorb some of that heat by shifting reactions in the endothermic direction. Increase pressure on a gas mixture, and gases will partially convert to reduce the number of molecules.
Hatakeyama shows that the evolutionary dynamics he derives obey an analogous principle. When the environment shifts, the phenotype moves quickly in one direction. The slow genetic response doesn't just passively follow — it amplifies that initial shift. The genotype evolves to reinforce the plastic change. This is not what you'd naively expect: if the phenotype has already moved to a good position, you might think natural selection on the genotype would be weak. But Hatakeyama shows that the internal stress generated by the phenotype-genotype mismatch creates a systematic pressure that pushes the genotype in the same direction as the phenotype — not opposing it, not randomly drifting, but amplifying it.
This amplification is a direct analogue of Le Chatelier's principle applied to evolution. The fast variable (phenotype) shifts; the slow variable (genotype) responds in the same direction to reduce the stress generated by the shift. The principle holds exactly in Hatakeyama's mathematical framework. Evolution, it turns out, is a physical system — and it obeys physical laws.
Relaxation timescales. One of the most striking quantitative predictions concerns how long assimilation takes. Hatakeyama derives an expression for the relaxation time — the time required for the genotype to catch up to the new phenotypic optimum after an environmental shift. This relaxation time, he shows, depends inversely on the strength of what he calls the restoring force.
The restoring force is the pressure that keeps the phenotype near its genotypic optimum in a stable environment. In a perfectly canalized genotype — one that produces essentially the same phenotype regardless of environmental noise — this force is strong. In a genotype that allows substantial clonal phenotypic fluctuations — variation among genetically identical individuals due to random developmental noise — the restoring force is weak.
The prediction is elegant: the weaker the restoring force, the longer assimilation takes.
Relaxation Time vs. Restoring Force Strength
Relaxation time (arbitrary units) as a function of restoring force strength. The relationship shows an inverse dependence: weaker restoring forces (which can arise from larger clonal phenotypic fluctuations) correspond to longer assimilation timescales.
| Label | Value |
|---|---|
| Strong restoring force | 0.2 |
| Moderate restoring force | 0.5 |
| Weak restoring force | 1.5 |
| Very weak restoring force | 4 |
illustrates this relationship conceptually, showing how relaxation time increases as the restoring force weakens. In organisms with highly robust, canalized development — where a given genotype produces a tight distribution of phenotypes — genetic assimilation happens relatively quickly after a plastic response. In organisms with noisier development, where genetically identical individuals can look quite different, the assimilation process stretches out over more generations.
This prediction is testable. Species with more developmental noise should show slower rates of genetic assimilation following environmental change. Hatakeyama doesn't provide specific empirical data to test this, but the theoretical framework generates clear, falsifiable hypotheses.
The divergence at perfect plasticity. The most counterintuitive result comes at the limit of cost-free, perfectly adaptive plasticity. If an organism can plastically adjust its phenotype instantly and without any metabolic cost — the theoretical ideal of perfect plasticity — then the restoring force on the genotype goes to zero. And in Hatakeyama's framework, when the restoring force goes to zero, the relaxation time diverges. It becomes infinite.
This means that in the limiting case of perfect plasticity, genetic assimilation effectively stalls. The phenotype shifts rapidly to track the environment, the genotype stays where it is, and the two never converge. The stress remains unresolved forever. The plastic response is so good that there's no pressure on the genotype to change.
This is a profound and somewhat paradoxical result: better plasticity leads to slower assimilation. It suggests that there is a sweet spot — plasticity that is good enough to buffer environmental change but costly enough to create selection pressure for genetic encoding. Evolution, in this framework, faces a tradeoff. Perfect plasticity would be wonderful for immediate survival, but it would prevent the long-term genetic accommodation that might lock in adaptive gains.
Why This Changes Things
The implications of Hatakeyama's framework ripple outward in multiple directions.
For evolutionary theory. The unification of phenotypic plasticity and genetic assimilation into a single mathematical formalism is a significant achievement. For decades, these phenomena were studied largely separately — plasticity by developmental biologists, assimilation by evolutionary geneticists — with only vague hand-waving about how they connected. Hatakeyama provides the rigorous bridge. The Price equation, already celebrated for its generality, now does even heavier lifting: it becomes the framework for understanding how the fastest and slowest processes in biology interact.
More importantly, the framework places assimilation squarely within the domain of physics. Le Chatelier's principle is not an analogy or a metaphor — it is a mathematical consequence of the dynamics. Evolution is not just a biological process guided by selection; it is also a physical process obeying thermodynamic-style constraints. This doesn't replace natural selection as an explanation, but it enriches it, adding a layer of predictability that selection alone cannot provide.
The prediction about developmental noise and assimilation rate is particularly valuable. It creates a bridge between developmental biology (which studies how genotypes produce phenotypes) and evolutionary genetics (which studies how allele frequencies change). If the relaxation time for assimilation depends on the canalization of development, then the architecture of development itself becomes a factor in the rate of evolution. This is a testable claim that could be checked against existing data on developmental robustness and evolutionary rates across species.
For understanding rapid evolution. The wall lizards of the Adriatic, the Italian wall lizards of the Pod Mrcaru experiment, and countless other cases of rapid evolution have long challenged the notion that evolution is slow. Hatakeyama's framework explains why: phenotypic plasticity buys time, allowing organisms to survive long enough for genetic assimilation to catch up. The plastic shift happens quickly, the genetic response follows more slowly, and the net result is rapid phenotypic change that appears, on short timescales, to be faster than mutation and selection alone could produce.
This is not just a theoretical point. It has practical consequences for how we think about evolution in a changing world. Species facing rapid climate change don't have time to wait for beneficial mutations to arise and spread. But they do have plasticity. Hatakeyama's framework suggests that plasticity will trigger a genetic response in the same direction — amplifying the initial shift rather than opposing it. Evolution under climate change may thus be faster than classical models predict, because organisms are already partway toward the new optimum when selection pressure kicks in.
For the study of development and evolution. The Baldwin effect — named after the psychologist James Baldwin, who hypothesized in 1896 that individually acquired traits could influence the course of evolution — has had a complicated history. Often dismissed or marginalized, it has experienced periodic revivals, most recently with the growth of evolutionary developmental biology (evo-devo). Hatakeyama's work can be seen as a formal, quantitative vindication of Baldwin's intuition. Plastic traits can indeed become genetic traits over time, not through some mystical influence of the phenotype on the genotype, but through the straightforward dynamics of coupled differential equations with timescale separation.
The framework also speaks to the debate about genetic accommodation — the process by which the genetic basis of a trait changes as that trait evolves. Hatakeyama's internal genetic stress provides a concrete mechanism: the mismatch between phenotype and genotype creates selection pressure that drives genetic accommodation in a specific direction. This is more than a story about mutations and selection; it is a dynamical story about how two variables interact to generate a trajectory.
For cancer and somatic evolution. While Hatakeyama's paper focuses on germline evolution across generations, the framework may have applications within organisms as well. Cancer evolution involves the accumulation of genetic changes in cell populations, but cells also exhibit phenotypic plasticity — switching between states, responding to microenvironmental cues. The same timescale separation between fast cellular phenotype and slower genetic change may apply. If so, Hatakeyama's framework might offer predictions about how tumors evolve, how drug resistance arises, and what determines the rate of genetic change in somatic cell populations.
For synthetic biology and artificial life. If assimilation follows predictable laws — if it is a universal response law for systems with phenotype-genotype timescale separation — then it should be observable in any system with those properties. Synthetic biologists building artificial genetic circuits or directed evolution systems might be able to engineer or exploit assimilation dynamics. Similarly, researchers working on artificial life simulations could test whether digital organisms exhibit the same Le Chatelier-style behavior.
What's Next
Hatakeyama's paper opens as many questions as it answers — which is, perhaps, the mark of a productive theoretical contribution.
Empirical tests. The prediction that developmental canalization affects assimilation rate is, in principle, testable. Comparative studies across species could examine whether organisms with more canalized development show faster rates of genetic assimilation following environmental perturbations. Long-term experiments with fast-evolving organisms — bacteria, fruit flies, digital organisms — could track assimilation dynamics directly and compare relaxation times to theoretical predictions.
Hatakeyama doesn't provide specific numerical predictions about what the relaxation time should be in any particular system — that would require knowing the restoring force constant and other parameters. But the framework generates testable qualitative predictions: if you experimentally manipulate the degree of developmental noise in a population, you should see corresponding changes in assimilation rate. This is within the reach of current experimental techniques.
Connecting to quantitative genetics. The framework needs to be connected more tightly to the standard tools of quantitative genetics — heritability, breeding values, the breeder's equation. These tools are used by plant and animal breeders, by evolutionary biologists studying natural populations, and by anyone trying to predict how traits will change under selection. If Hatakeyama's dynamical systems approach can be translated into the language of quantitative genetics — or if it can be shown to contain the breeder's equation as a special case — it will become more accessible to the empirical community.
Extending to multiple traits and fluctuating environments. The paper deals with a single trait in a single environment shift. Real evolution involves many traits simultaneously, and environments don't shift once and stay changed — they fluctuate, sometimes unpredictably. How does the framework extend to multi-trait landscapes, where the fast and slow variables interact across many dimensions? How does repeated environmental fluctuation affect the assimilation dynamics? These questions will require substantial additional work.
The cost of plasticity. Hatakeyama's result about perfect plasticity leading to stalled assimilation raises a deep question about why organisms don't evolve perfectly cost-free plasticity. If plasticity triggers assimilation and assimilation is beneficial, why wouldn't evolution push plasticity toward its ideal limit? The answer presumably lies in the costs of plasticity — the metabolic, computational, and developmental expenses of maintaining a flexible phenotype. But formalizing these costs within the framework is a next step.
Alternative models of genotype-phenotype mapping. The framework assumes a particular structure for how genotype maps to phenotype — a landscape with a single optimum, smooth and well-behaved. Real genotype-phenotype maps are more complex: they involve vast networks of gene interactions, developmental cascades, and epigenetic effects. Extending the framework to more realistic GP maps — particularly the multiverse of phenotypes that arises from gene regulatory networks — is an important direction.
Relation to the extended evolutionary synthesis. The broader movement in evolutionary biology sometimes called the extended evolutionary synthesis argues that processes like phenotypic plasticity, niche construction, and developmental bias are not just consequences of genetic evolution but active drivers of it. Hatakeyama's framework provides a mathematical instantiation of this perspective: plasticity doesn't just respond to genetic evolution; it actively shapes the genetic trajectory through the internal stress mechanism. This may help bridge the sometimes-fractious debates between "classical" and "extended" evolutionary frameworks.
Biology has long searched for the kind of unity that physics enjoys. Newton's laws apply to apples and planets; thermodynamics applies to engines and refrigerators. When the same mathematics describes disparate phenomena, we gain predictive power and deeper understanding.
Hatakeyama's paper is a step toward that unity for evolutionary biology. By showing that phenotypic plasticity and genetic assimilation are two aspects of a single dynamical process — a process that obeys Le Chatelier's principle — he has given us a small piece of a universal law. "Evolutionary systems in which rapid phenotypic responses precede slower genetic change" is a description of almost every biological system we know. The response law Hatakeyama proposes should apply, in some form, to bacteria evolving antibiotic resistance, to lizards adapting to new islands, to human populations shifting their physiology as diets and climates change.
The wall lizards that colonized Pod Mrcaru didn't know they were testing a theory. They simply did what life has always done: respond quickly, then slowly follow. But the mathematics of that response — the stress and the relaxation, the fast shift and the slow amplification — turns out to be written in the same language as physics. Evolution, it seems, is not just a tinkerer. It is also a system in equilibrium, pushed and pulled by forces it cannot see, settling toward a balance it cannot anticipate. The principle that governs that settling is as old as thermodynamics and as new as this paper. Le Chatelier, meet Darwin.