Lattice Brownian bees with cooperative reproduction: steady states, collapse, and spreading

This paper extends the "Brownian bees" model to cooperative reproduction on a lattice, using a hydrodynamic free-boundary formalism to characterize steady states, stability, and the transition between diffusive spreading and finite-time collapse as the cooperation order $k$ varies.

The Gist

Imagine a bustling city made of tiny people called "Brownian bees." These bees are constantly wandering around randomly, bumping into each other and moving in all directions. This is a model for how populations grow and spread in nature.

In the classic version of this story, any bee can have a baby. But there's a strict rule to keep the city from getting too crowded: as soon as a new baby is born, the bee that is currently farthest away from the city center gets kicked out of the city. This keeps the total number of bees exactly the same.

This paper asks a fascinating question: What happens if the bees need to work together to have a baby?

Instead of just one bee making a baby, what if they need a group of $k$ bees to gather in the same spot to reproduce?

• If $k=2$, two bees need to meet.
• If $k=3$, three bees need to meet.
• If $k=4$, four bees need to meet, and so on.

The researchers found that the number of bees required to reproduce ($k$) can completely change the fate of the city. Here is the breakdown of their findings:

1. The "Sweet Spot" (When $k = 1$ or $k = 2$)

The Analogy: Think of a stable, healthy town.
When only one or two bees are needed to reproduce, the city finds a perfect balance. If you poke the city or push the bees around, they naturally settle back into that perfect shape. The population is stable. It's like a well-tuned engine that runs smoothly forever.

2. The "Tipping Point" (When $k = 3$)

The Analogy: A tightrope walker.
When three bees are needed to make a baby, the system becomes incredibly sensitive. It's like walking a tightrope.

• If the bees are too eager to reproduce: The city collapses. The bees rush toward the center, crowding together until they all pile up on top of each other in a tiny, dense point. This happens in a finite amount of time.
• If the bees are too slow to reproduce: The city spreads out forever. The bees drift away from the center, getting thinner and thinner, like a drop of ink spreading in a glass of water.
• The Perfect Balance: There is one specific, magical ratio of "how fast they wander" vs. "how fast they reproduce" where the city can stay in a steady state. But even then, there isn't just one shape; there is a whole family of possible shapes they could take, all equally valid.

3. The "Instability Zone" (When $k = 4$ or more)

The Analogy: A house of cards that is already falling over.
When four or more bees are needed to reproduce, the "stable city" shape is a lie. It looks stable for a moment, but it is actually unstable.

• If the city starts slightly too small: It collapses. The bees rush to the center, and the population density spikes wildly. The researchers found that this collapse happens in a very specific, predictable way: the center gets incredibly dense while the edges thin out, creating a "core" of bees surrounded by a thin "skin" where the rules of movement change.
• If the city starts slightly too big: It spreads out. The bees drift apart. Because it's so hard to get four bees to meet, the reproduction stops mattering, and the bees just act like random walkers.

The "Edge" Effect

One of the coolest discoveries in the paper is about what happens during the collapse (when $k=4$).
Imagine the bees rushing to the center. The middle of the group is so crowded that the "reproduction" (making babies) is the only thing that matters. But right at the very edge of the group, the bees are so spread out that "diffusion" (wandering) is the only thing that matters.
The researchers had to use a special mathematical technique called "matched asymptotics" to describe this. Think of it like describing a storm: you need one set of rules to describe the violent eye of the storm (where the bees are crashing together) and a completely different set of rules to describe the calm, thin rim on the outside. The paper shows how these two different worlds fit together perfectly.

Summary

The paper tells us that nature has a strong preference for simple reproduction.

• Simple reproduction ($k=1, 2$): Leads to stable, robust communities that can recover from shocks.
• Complex cooperation ($k \ge 4$): Leads to instability. The community either implodes into a singularity or dissolves into nothingness.
• The middle ground ($k=3$): Is a fragile, critical state where the outcome depends entirely on the exact balance of speed and reproduction.

The researchers confirmed all these predictions by running computer simulations of hundreds of thousands of individual bees, showing that the math perfectly matches the behavior of the microscopic particles.

Technical Summary

Technical Summary: Lattice Brownian Bees with Cooperative Reproduction

Problem Statement
This work extends the "Brownian bees" model—a system of $N$ symmetric random walkers where the population size is fixed by removing the particle farthest from the origin at every birth event—to include cooperative reproduction. In the standard model, reproduction is binary ($A \to 2A$). Here, the authors investigate $k$-body cooperative reproduction ($kA \to (k+1)A$), where a successful reproductive event requires the encounter of $k$ individuals at the same lattice site. The study focuses on the hydrodynamic (HD) limit ($N \to \infty$) to determine how the order of cooperation $k$ affects the existence, stability, and long-time dynamics of the population density.

Methodology
The authors formulate a hydrodynamic free-boundary problem for the macroscopic density $u(x,t)$ in the limit $N \to \infty$. The system is governed by a reaction-diffusion equation with a nonlinear source term representing cooperative reproduction:
$$\partial_t u = D \partial_{xx} u + \lambda u^k$$
subject to absorbing boundary conditions at the edges of a symmetric compact support $|x| < L(t)$, where $L(t)$ is determined implicitly by the mass conservation constraint $\int_{-L(t)}^{L(t)} u(x,t) dx = 1$.

The analysis proceeds through three primary methods:

1. Analytical Steady-State Solutions: Derivation of exact density profiles using Lane-Emden equations and elliptic functions.
2. Linear Stability Analysis: Perturbation of steady states to derive a Schrödinger-type eigenvalue problem, determining stability based on the sign of the leading eigenvalue $\sigma$.
3. Asymptotic and Numerical Analysis:
   • For $k=3$ (marginal case), analysis of self-similar collapse and spreading solutions.
   • For $k \ge 4$, development of a matched-asymptotic expansion to describe the collapse dynamics, identifying a separation between a reproduction-dominated bulk and a diffusion-dominated boundary layer.
   • Numerical solutions of the HD free-boundary problem and Monte Carlo (MC) simulations of the underlying microscopic lattice model to validate analytical predictions.

Key Contributions and Results

• Steady-State Existence and Uniqueness:

  • For $k < 3$, a unique steady-state density profile exists for all ratios of reproduction to diffusion rates ($\alpha = \lambda/D$).
  • For $k = 3$, a steady state exists only at a single critical ratio $\alpha_c = \pi^2/2$. At this critical point, a continuous family of steady states exists, parametrized by the peak density.
  • For $k > 3$, a unique steady-state profile exists mathematically but is shown to be linearly unstable.

• Stability Transition at $k=3$:

  • The system undergoes a linear stability transition at $k=3$. Steady states are linearly stable for $k \le 2$ and unstable for $k \ge 4$.
  • The instability for $k > 3$ is proven analytically using a mass-slope criterion, showing that the leading even eigenvalue becomes positive.

• Dynamical Regimes:

  • $k=2$: The system relaxes to a stable steady state.
  • $k=3$ (Marginal): The dynamics are controlled by the ratio $\alpha$.
    • If $\alpha < \alpha_c$, the population undergoes asymptotically self-similar diffusive spreading ($L(t) \sim t^{1/2}$), though the cubic reproduction term remains quantitatively relevant to the shape of the profile.
    • If $\alpha > \alpha_c$, the population undergoes a finite-time, self-similar collapse to the origin ($L(t) \sim (T-t)^{1/2}$).
  • $k \ge 4$: The unstable steady state acts as a separatrix.
    • Initial conditions narrower than the steady state lead to a finite-time collapse. For $k=4$, this collapse exhibits a scale separation: the bulk is dominated by reproduction ($u \sim (T-t)^{-1/3}$), while a narrow boundary layer is dominated by diffusion. The collapse rate scales as $L(t) \sim (T-t)^{1/3}$.
    • Initial conditions broader than the steady state lead to diffusive spreading. While the bulk density approaches a Gaussian, the boundary position $L(t)$ exhibits a logarithmic correction to standard diffusion, scaling as $L(t) \sim \sqrt{2Dt \ln t}$.

Significance
The paper demonstrates that the order of cooperative reproduction $k$ fundamentally alters the spatial structure and dynamical regimes of the population. The authors identify $k=3$ as the mass-critical exponent in one dimension, separating regimes of stable self-organization ($k \le 2$) from regimes of instability leading to either collapse or unbounded spreading ($k \ge 4$).

The work highlights that minimal models of population dynamics with spatial selection encode a strong preference for low-order reproduction. For $k \le 2$, the system self-organizes into a robust, stable macroscopic collective. For $k \ge 4$, no such stable configuration exists; the population either collapses to a point (resolved only by finite-size effects not captured by the continuous HD theory) or spreads diffusively. This finding offers a theoretical echo to the empirical prevalence of binary reproduction in biological systems, suggesting that higher-order cooperative mechanisms may be dynamically unstable in the absence of other regulatory constraints. The results are quantitatively confirmed by both HD numerical solutions and microscopic Monte Carlo simulations.