⚛️ biophysics

Dynamics of fluctuating populations in multi-state switching environments

This paper investigates how microbial strain competition dynamics, including population statistics and fixation outcomes, are influenced by multi-state stochastic environmental switching with intermediate resource levels, revealing how gradual fluctuations differ from traditional binary feast-famine models.

Original authors: Mobilia, M.

Published 2026-08-12

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Mobilia, M.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine a bustling city where the rules of survival change every day. Sometimes the streets are packed with food and resources, and everyone thrives; other times, a sudden drought leaves everyone starving. This is the reality for microbes—tiny, single-celled organisms that live in environments that are constantly shifting. Scientists have long studied how these tiny populations survive, focusing on two main forces: the "noise" of their own numbers (if a group is small, a few random deaths can wipe them out) and the "noise" of their environment (sudden changes in food or temperature). For decades, the standard story was simple: the environment flips a switch between two states, like a light bulb that is either fully "On" (feast) or fully "Off" (famine). But real life is rarely that black and white. In the messy, real world, conditions often fade in and out, passing through many shades of gray before reaching the extremes.

This paper, written by Mauro Mobilia, dives into that gray area. It asks a simple but profound question: What happens to a competition between two types of microbes if the environment doesn't just snap between "feast" and "famine," but slowly drifts through a series of intermediate steps? Think of it like a dimmer switch instead of a light switch. The researchers used computer simulations and math to watch how a slightly slower-growing microbe (the underdog) fares against a faster-growing one when the food supply changes gradually through multiple stages, rather than jumping abruptly. They discovered that the speed of these changes and the specific "shape" of the food supply curve can completely flip the odds of who wins, revealing a much richer and more complex story of survival than the old two-state models suggested.

The Dimmer Switch of Survival

In the world of microbes, life is a constant race. Imagine two teams of runners: Team Fast (the "F" strain) and Team Slow (the "S" strain). Team Fast is naturally quicker, but Team Slow is tough. They are running on a track that changes length and surface conditions. In the old, simplified view of biology, the track would instantly snap between a short, easy sprint (a "feast" with plenty of food) and a long, grueling marathon (a "famine" with almost no food). Scientists used to think this binary switch was enough to explain how these populations evolved.

But the real world is more like a dimmer switch. The lights don't just go from bright to dark; they pass through a dozen shades of gray. The environment might go from "super abundant" to "moderately good" to "okay" to "barely enough" before hitting "starvation." This paper explores what happens when the environment moves through these intermediate states. The researchers set up a digital playground where they could control exactly how many "steps" the environment took between the best and worst conditions. They could make the environment switch between 3 states (famine, middle, feast), 5 states, or even more, and they could control how fast the switch happened.

The Race Against the Clock

The core of the experiment was to see which team would eventually take over the whole population—a process called "fixation." In a stable world, Team Fast would almost always win because they are better at grabbing resources. Team Slow would fade away. But in a fluctuating world, things get tricky.

The researchers found that the outcome depends heavily on two things: how fast the environment changes and how the food is distributed across the different states.

1. The Slow-Motion World:
When the environment changes very slowly (like a dimmer switch being turned over the course of hours), the population has time to adjust to each specific level of food. If the environment spends a lot of time in the "harsh" states (low food), the population size shrinks. When a population is small, random chance plays a huge role. It's like a coin toss: if you flip a coin 10 times, you might get 8 heads just by luck. If you flip it 1,000 times, you'll get close to 50/50. In these small, starving populations, the "underdog" Team Slow has a better chance of surviving just by luck, even though they are slower. The simulations showed that if the environment lingers in these tough, intermediate states, Team Slow can actually win more often than in the simple two-state model.

2. The Fast-Forward World:
When the environment flips rapidly (like a strobe light), the population can't react to each individual change. Instead, it feels an "average" environment. In this fast-switching scenario, the population behaves as if it's living in a single, constant world with a specific amount of food. The researchers found that if the environment spends more time in the "mild" or "feast" states, the population stays large, and Team Fast wins easily. However, if the "average" environment is harsh, the population stays small, and Team Slow gets a fighting chance.

3. The Middle Ground:
The most interesting results happened in the "intermediate" switching regime, where the environment changes at a speed similar to how fast the microbes reproduce. Here, the population size fluctuates wildly, creating "bottlenecks"—moments where the population crashes down to a tiny number before bouncing back. These bottlenecks are crucial. They act like a sieve, randomly filtering out individuals. The simulations showed that in these conditions, the presence of intermediate states (the gray areas) makes the population dynamics much more complex than the simple "feast-or-famine" model predicted.

The Surprising Twist

The most exciting finding is that adding more intermediate states doesn't just make the math harder; it changes the winner.

In the traditional two-state model (just feast and famine), the odds of the slow strain winning were predictable based on how often the environment was harsh. But in the multi-state world, the researchers found that the distribution of the food levels matters immensely. If the "middle" states are actually quite harsh (closer to famine than feast), the slow strain has a much better chance of winning. It's as if the dimmer switch is stuck in the "low" zone for a long time, keeping the population small and giving the underdog a lucky break.

Conversely, if the environment spends most of its time in the "mild" or "feast" zones, the fast strain dominates. The paper suggests that the old binary models might be missing the mark because they don't account for these gradual transitions. By ignoring the "gray" states, we might be underestimating how often the underdog can survive and take over in real-world scenarios.

The Takeaway

This study doesn't claim to have solved the mystery of evolution, but it does offer a powerful new lens. It suggests that nature's "dimmer switches" are just as important as its "on/off" switches. The frequency of environmental changes and the specific shape of the resource curve determine whether the fast and furious win, or whether the slow and steady (sometimes) take the crown.

The researchers used computer simulations to map out these scenarios, showing that the path to survival is rarely a straight line. Whether a microbe survives a feast-famine cycle depends on the rhythm of the changes and the specific steps it takes through the dark and the light. It's a reminder that in the chaotic, fluctuating world of microbes, the details of the journey matter just as much as the destination.

Technical Summary: Dynamics of Fluctuating Populations in Multi-State Switching Environments

Problem Statement
Microbial populations typically evolve in volatile environments where abiotic factors (e.g., nutrient availability, temperature) fluctuate over time. While traditional theoretical and experimental studies often model these fluctuations using binary (two-state) switching models—representing abrupt transitions between "feast" (abundant resources) and "famine" (scarce resources)—recent experimental evidence suggests that real-world feast-famine cycles exhibit more complex temporal dynamics. Specifically, environmental deterioration and recovery often occur through intermediate states rather than abrupt jumps. This paper addresses the gap in understanding how these intermediate environmental states influence the competition between microbial strains, particularly how the frequency and amplitude of fluctuations affect population size statistics and fixation probabilities.

Methodology
The author proposes a class of multi-state stochastic switching models to describe the eco-evolutionary dynamics of two competing strains: a faster-growing type ( $F$ ) and a slightly slower-growing type ( $S$ ) with a selective disadvantage $s$ . The population size $N(t)$ is limited by a time-varying carrying capacity $K(t)$ , which serves as a coarse-grained proxy for nutrient availability.

Model Structure: The environment is modeled as a continuous-time Markov chain with $2n+1$ states (where $n \geq 1$ ), ranging from the harshest state ( $i=-n$ , famine, capacity $K_-$ ) to the mildest state ( $i=n$ , feast, capacity $K_+$ ), with $2n-1$ intermediate states. Transitions occur between nearest-neighbor states with rates $\nu_{2n+1}^\pm$ and $\omega_{2n+1}^\pm$ .
Population Dynamics: The population evolves via a birth-death process where birth rates depend on individual fitness relative to the average population fitness, and death rates are logistic, limited by the instantaneous carrying capacity $K(t)$ .
Analytical and Computational Approaches:
- Simulations: Extensive stochastic simulations using the Gillespie algorithm were performed to track population trajectories, fixation events, and population size distributions (PSD).
- Analytical Approximations:
  - Slow/Fast Switching Limits: In the limits of very slow ( $\nu \ll s$ ) and very fast ( $\nu \gg s$ ) switching, the author derives "quenched" and "annealed" approximations, respectively, averaging Moran fixation probabilities over the stationary distribution of carrying capacities.
  - N-PDMP Approximation: For the intermediate switching regime ( $\nu \sim s$ ), where analytical solutions are difficult, the author employs a Piecewise Deterministic Markov Process (N-PDMP) approximation. This method treats the population size as a deterministic variable evolving between environmental switches, averaged over the rescaled switching rates to account for the number of switches occurring on the evolutionary timescale.

Key Contributions and Results

Population Size Distribution (PSD):
- The study characterizes the PSD under different switching regimes. In the slow switching regime, the PSD is multimodal, exhibiting $2n+1$ peaks corresponding to the distinct carrying capacities of the environmental states. In the fast switching regime, the PSD becomes unimodal, centered around an effective carrying capacity $K_{2n+1}$ (the harmonic mean of capacities weighted by the stationary distribution).
- Unlike binary models, multi-state models show that intermediate states significantly broaden the PSD and alter the probability of the population size falling between the extreme values $K_-$ and $K_+$ .
- The N-PDMP approximation accurately reproduces the peaks of the PSD in slow switching and the unimodal shape in fast switching, though it underestimates the width of the distribution in slow switching due to the neglect of demographic noise.
Fixation Probability of the Slow Strain ( $\phi$ ):
- The paper investigates the probability that the slower strain $S$ fixes (takes over) the population. This probability is highly sensitive to the distribution of carrying capacities and the switching rate.
- Comparison with Binary Models: The presence of intermediate states can either enhance or hinder the fixation of the slow strain compared to binary switching, depending on the specific parameters (number of states $n$ , median capacity $K_0$ , and environmental bias $\delta$ ).
- Environmental Bias: When the environment is biased towards harsh states ( $\delta < 0$ ), the fixation probability of the slow strain tends to be higher in multi-state environments than in binary ones, particularly if the median capacity $K_0$ is low. Conversely, bias towards mild states ( $\delta > 0$ ) generally reduces the fixation probability of the slow strain in multi-state settings.
- Regime Dependence: The relationship between multi-state and binary fixation probabilities is not monotonic. For certain parameter sets, the multi-state fixation probability exceeds the binary case in slow/intermediate regimes but falls below it in fast regimes, or vice versa.
Analytical Validity:
- The derived approximations for slow and fast switching limits align well with simulation data.
- The N-PDMP-based approximation (Eq. 19) is shown to be a robust tool for predicting fixation probabilities across a broad range of switching rates, including the intermediate regime where ecological and evolutionary timescales overlap.

Significance and Claims
The paper claims that multi-state stochastic switching models provide a necessary and versatile framework for investigating eco-evolutionary dynamics that cannot be captured by traditional binary models. By treating the carrying capacity as a coarse-grained representation of nutrient levels with intermediate states, the author demonstrates that:

Environmental Complexity Matters: The statistical structure of environmental fluctuations (frequency and amplitude) significantly shapes population dynamics. The existence of intermediate states introduces richer dynamical scenarios, such as non-monotonic dependencies of fixation probability on switching rates.
Coarse-Graining Utility: These models successfully bridge the gap between microscopic nutrient dynamics and macroscopic population outcomes without explicitly modeling nutrient uptake kinetics.
Predictive Power: The developed analytical tools (specifically the N-PDMP approximation) allow for the efficient characterization of fixation properties in complex, fluctuating environments, offering scaling insights that are difficult to obtain through simulation alone.

The author concludes that while binary models offer mathematical tractability, they fail to capture the nuances of gradual environmental changes observed in experiments. The multi-state approach presented here offers a more realistic description of feast-famine cycles, highlighting how the interplay between environmental variability and demographic noise determines the evolutionary fate of competing strains.

The Dimmer Switch of Survival

The Race Against the Clock

The Surprising Twist

The Takeaway

More like this