恒化器中分阶段捕食者-猎物模型的连贯动力学分析

《Mathematical Biosciences》：Coherent dynamical analysis arising from a stage-structured predator-prey model in a chemostat

　　

# Summary of the Paper

## Core Contribution
A **revised stage-structured predator–prey chemostat model** that includes **juvenile mortality during maturation**, not just washout. The key fix is introducing the survival factor **e???** in the juvenile→adult transition (and the matching juvenile outflow), making the formulation cohort-consistent. This differs from both Blasius et al.'s original equations and the formulation in [14].

## Model Setup
Well-mixed chemostat with constant inflow of limiting nutrient N (concentration N_in). Populations:
- Phytoplankton *P* (prey) — Monod growth in N
- Rotifer predator split into **eggs E**, **juveniles J**, **adults A**
- Egg production: Type III functional response
- Two delays: **θ** (egg development), **τ** (juvenile maturation)

## Main Results

**1. Well-posedness**
An age-cohort balance identifies compatible initial histories; an egg-inclusive balance functional proves global existence, nonnegativity, and dissipativity.

**2. Coexistence & persistence criterion**
With *d* = βe^(?δθ?cτ) and ? = d·F_B(P?)/c:

> Positive equilibrium exists **iff ? > 1**, equivalently:
> **δθ + cτ < log(β·F_B(P?)/c)**

Unique equilibrium:
- P_c = K_B(c·βr_B·e^(?δθ?cτ) ? c)^(1/κ)
- A_c = e^(?cτ)·B_c

**3. Stability analysis** — subtle point
When a delay is varied, the **survival factors shift the equilibrium itself**, so linearization coefficients cannot be held fixed. The characteristic determinant factors into a stable part plus a two-delay characteristic function. Coefficients β?, β?, γ? depend explicitly on delays.

**4. Hopf bifurcation**
- Conditional crossing criteria and a stability-count theorem on the finite coexistence region
- Normal-form calculation determines **direction** of each nondegenerate periodic branch

**5. Numerics** (β=5, δ=0.55, τ=1.8 days default, MATLAB `dde23`)
| Maturation delay | Hopf type | Observed dynamics |
|---|---|---|
| Instantaneous | **Supercritical** | Stable limit cycle |
| τ = 1.8 days | **Subcritical** | Attracting finite-amplitude oscillations (distinct from local unstable Hopf branch) |

Parameter sensitivities quantify threshold displacement.

## Motivation & Significance
Blasius et al.'s year-long chemostat experiments (*Brachionus calyciflorus* / *Monoraphidium minutum*) showed **persistent oscillations over 50+ cycles** — two regimes: coherent (≈π/2 phase lag) and non-coherent, with stochastic reversibility between them. This challenged the view that predator–prey cycles are purely theoretical artifacts requiring external forcing.

---

**Structure:** Sec. 2 model & basic properties → Sec. 3 persistence → Sec. 4 characteristic equation → Sec. 5 normal form → Sec. 6 numerics → Sec. 7 discussion

**Author:** Hui Zhou, School of Mathematics and Statistics, Hefei Normal University
**Funding:** NSFC (12517710, 12331006), CAS Strategic Priority (XDB0900100), National Key R&D (2024YFA1013603, 2024YFA1013600)

Would you like me to go deeper into any part — e.g., the characteristic matrix derivation, the normal-form calculation, or the biological interpretation of the coherent/non-coherent regimes?