🔢 mathematics

Size-Selective Threshold Harvesting under Nonlocal Crowding and Exogenous Recruitment

This paper formulates and analyzes an infinite-horizon bioeconomic optimal control problem for a size-structured fish population with exogenous recruitment and nonlocal crowding, proving that the optimal harvesting strategy is a bang-bang threshold policy that aligns economic maximization with long-term biological viability, as demonstrated through a numerical case study on Atlantic cod.

Original authors: Jiguang Yu, Louis Shuo Wang, Ye Liang

Published 2026-03-31

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jiguang Yu, Louis Shuo Wang, Ye Liang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a massive, underwater city where the residents are fish. In this city, the fish grow bigger as they age, but they also face two big challenges: competition (too many fish fighting for food) and fishing nets (humans trying to catch them).

For a long time, scientists and fishermen have tried to figure out the perfect rule: How many fish should we catch, and how big should they be? If we catch too many, the city collapses. If we catch too few, we lose money.

This paper is like a sophisticated traffic control system for that underwater city. It uses advanced math to find the "Goldilocks" zone where we make the most money without destroying the fish population.

Here is the breakdown of their ideas using simple analogies:

1. The "Crowded Room" Effect (Nonlocal Crowding)

Usually, we think of fish growing at a steady pace. But in reality, if a room gets too crowded, everyone moves slower and gets sick easier.

The Analogy: Imagine a gym. If only a few people are there, you can run fast on the treadmill. If the gym is packed, you have to slow down, and you might trip or get sick from the stress.
The Paper's Twist: The authors realized that the entire population affects every single fish. They created a "Crowding Index" (let's call it the Room Temperature). If the Room Temperature is high, fish grow slower and die faster. This "temperature" changes based on how many fish are in the water at that moment.

2. The "External Waterfall" (Exogenous Recruitment)

Most fish models assume that baby fish are born from the adult fish currently in the water. But this paper looks at a different scenario: Stocked Fisheries.

The Analogy: Imagine a river that flows into a lake. Usually, the lake's water level depends on how much rain falls inside the lake. But in this model, imagine a giant hose (a hatchery or a nursery upstream) is constantly pouring new baby fish into the lake, regardless of how many adult fish are there.
Why it matters: This changes the rules. The population isn't just surviving on its own; it's being "fed" from the outside. The scientists had to invent a new way to measure if the fish are healthy enough to survive this specific setup. They called this the "Intrinsic Replacement Index." Think of it as a "Health Score." If the score is above 1, the fish are doing great. If it's below 1, they are struggling, even if the hose is keeping them alive.

3. The "All-or-Nothing" Rule (Bang-Bang Threshold)

The biggest question is: How do we catch the fish? Should we catch a little bit of everything? Or only the big ones?

The Math Magic: The authors used a famous mathematical tool (Pontryagin's Maximum Principle) to solve this. They proved that the best strategy isn't a gentle mix. It's a "Bang-Bang" switch.
The Analogy: Imagine a light switch. It's either OFF (don't catch) or ON (catch as much as you can). There is no "dimmer switch."
The Result: The math showed that the optimal strategy is a Size Threshold.
- Fish smaller than the line: Leave them alone! (Switch is OFF). Let them grow.
- Fish larger than the line: Catch them all you can! (Switch is ON).
- This is exactly how "Minimum Landing Size" laws work in the real world, but the paper proves mathematically why this is the most profitable way to do it.

4. The Atlantic Cod Case Study

To prove their theory works, they simulated a fishery for Atlantic Cod (the fish famous for the "Great Collapse" in the 1990s).

The Setup: They fed their computer model real-world data: how fast cod grow, how much they cost, and how crowded they get.
The Discovery: They found a "Magic Number."
- If you catch cod smaller than 66.45 cm, you make less money in the long run because you're killing fish before they get big and valuable.
- If you only catch cod larger than 66.45 cm, you maximize your profit.
The Happy Ending: At this specific size (66.45 cm), the fishery makes the most money AND the "Health Score" (Intrinsic Replacement Index) stays high. The fish population remains healthy and sustainable.

The Big Takeaway

This paper is a victory for precision management.

In the past, we might have said, "Let's just limit the total number of fish we catch." But this paper says, "No, that's too blunt. We need to be surgeons, not sledgehammers."

By using a strict rule that says "Catch only the big ones, protect the little ones," we can:

Make more money (because big fish are worth more).
Save the fish (because the little ones grow up to replace the big ones).
Handle the crowd (because the model accounts for how crowded the ocean gets).

It's like a video game strategy: Don't kill the weak enemies immediately; let them level up, then take out the high-level ones for the big points, ensuring the game never ends.

1. Problem Statement

The paper addresses the bioeconomic optimization of fisheries management, specifically focusing on size-structured fish populations where individual growth and mortality are influenced by nonlocal crowding (density dependence acting on the aggregate population) and where recruitment is driven by exogenous inflow (e.g., hatchery stocking or subsidized stocks) rather than endogenous reproduction.

The core challenge is to determine an optimal size-selective harvesting policy that maximizes the infinite-horizon discounted economic revenue while maintaining the biological viability of the stock. Unlike standard models that treat fish as a homogeneous biomass, this problem requires managing a distributed state variable (population density across sizes) subject to a nonlinear partial differential equation (PDE) with nonlocal feedback.

2. Methodology

A. Mathematical Modeling

The authors formulate a nonlinear McKendrick–von Foerster transport equation to describe the population dynamics $x(t, l)$ , where $t$ is time and $l$ is body length:
$\frac{\partial x}{\partial t} + \frac{\partial}{\partial l}(g(E, l)x) = -(\mu(E, l) + u)x$

State Variable: $x(t, l)$ is the density of fish of size $l$ at time $t$ .
Nonlocal Crowding Index ( $E$ ): Defined as $E(t) = \int_{l_0}^{l_m} \chi(l)x(t, l) dl$ $E (t) = \int_{l_{0}}^{l_{m}} χ (l) x (t, l) d l$ . This index feeds back into the system, affecting:
- Growth ( $g$ ): Slows down as crowding increases.
- Natural Mortality ( $\mu$ ): Increases with crowding.
Control Variable ( $u$ ): Size-selective fishing mortality, constrained by $0 \le u(t, l) \le u_{max}$ .
Boundary Condition: Recruitment is strictly exogenous: $g(E, l_0)x(t, l_0) = p(t)$ , where $p(t)$ is a prescribed inflow flux. This distinguishes the model from traditional endogenous birth-law models.

B. Steady-State and Viability Analysis

Stationary Profile: The authors derive the explicit stationary solution for the no-harvest case ( $u \equiv 0$ ) and establish the existence and uniqueness of the steady-state crowding level $E^*$ by solving a scalar nonlinear closure equation $F(E) = 0$ .
Intrinsic Replacement Index ( $R(E)$ ): A novel metric is introduced to assess biological viability in exogenously forced systems. Unlike traditional persistence thresholds (where $R>1$ implies survival), here $R(E)$ serves as a diagnostic tool. It measures the stock's intrinsic capacity to replace itself if the external inflow were removed. The authors impose $R(E) \ge 1$ as a biological viability target.

C. Optimal Control Synthesis

Objective: Maximize the infinite-horizon discounted revenue:
$J[u] = \int_0^\infty e^{-rt} \int_{l_0}^{l_m} c(l) u(t, l) x(t, l) \, dl \, dt$
Pontryagin Maximum Principle: The authors derive formal first-order necessary conditions using a Lagrangian approach. This yields a coupled state-adjoint system.
- The adjoint equation ( $\lambda$ ) represents the shadow value of a marginal fish.
- Due to the nonlocal nature of $E$ , the adjoint equation contains a complex integral term.
Weak-Coupling Approximation: To make the problem tractable, the authors apply a weak-coupling approximation, neglecting the nonlocal integral terms in the adjoint equation (assuming density sensitivities are small relative to transport and mortality).
Switching Rule: The optimal control is determined by a switching function $S(t, l) = c(l) - \lambda(t, l)$ .
Threshold Policy Proof: Under a single-crossing assumption (market value $c(l)$ increases with size, and the shadow value $\lambda$ decreases or grows slower), the authors mathematically prove that the optimal control is a bang-bang threshold policy:
$u^*(l) = \begin{cases} 0 & \text{if } l < l^* \\ u_{max} & \text{if } l > l^* \end{cases}$
This implies a strict minimum landing size $l^*$ .

3. Key Contributions

Novel Model Formulation: Integration of nonlocal crowding effects with strictly exogenous recruitment in a size-structured PDE framework, specifically tailored for enhancement fisheries or heavily subsidized stocks.
Biological Viability Diagnostic: Introduction of the Intrinsic Replacement Index ( $R(E)$ ) for exogenously forced systems, clarifying its role as a viability measure rather than a binary persistence threshold.
Analytical Proof of Threshold Structure: Rigorous derivation (via Pontryagin principles and single-crossing assumptions) proving that the optimal size-selective harvesting strategy for this class of nonlinear PDEs is a bang-bang threshold policy.
Bridging Theory and Practice: A calibrated numerical case study demonstrating how theoretical optimal control translates to specific management parameters (minimum landing sizes) for real-world fisheries.

4. Results (Numerical Case Study)

The authors calibrated the model to the Atlantic cod (Gadus morhua) fishery using parameters for growth (von Bertalanffy), mortality, and market value.

Baseline: The no-harvest steady state yielded a crowding level $E^* \approx 103,108$ and an intrinsic replacement index $R(E^*) \approx 1.47$ , indicating a healthy, viable stock.
Optimal Threshold: The numerical optimization of the discounted revenue function identified an optimal minimum harvest size of $l^*_{opt} = 66.45$ cm.
Economic-Biological Alignment:
- At $l^*_{opt}$ , the maximum discounted revenue was achieved.
- Crucially, the resulting steady state maintained a replacement index of $R \approx 1.98$ , well above the viability target of 1.
- This demonstrates that maximizing economic return does not necessarily compromise biological viability in this framework; the optimal policy naturally protects smaller, less valuable fish while harvesting larger ones.
Dynamics: Simulations showed that the population converges to the new stationary regime within 8–10 years under the optimal threshold policy.

5. Significance and Implications

Management Relevance: The study provides a mathematical justification for minimum landing size regulations, a common but often heuristic tool in fisheries. It shows that such rules emerge naturally from the trade-off between immediate harvest value and the future value (shadow price) of allowing fish to grow.
Handling Exogenous Recruitment: By modeling recruitment as exogenous, the paper offers a robust framework for managing stock enhancement programs (e.g., hatchery releases) where natural reproduction is not the primary driver of stock levels.
Nonlocal Effects: The inclusion of nonlocal crowding highlights that individual growth and survival are functions of the entire population structure, not just local density, leading to more realistic predictions of stock responses to harvesting.
Limitations & Future Work: The authors acknowledge that the derivation of the adjoint system is formal (not fully rigorous in infinite dimensions) and that the weak-coupling approximation simplifies the nonlocal feedback. Future work should focus on endogenous recruitment laws and stochastic elements.

In conclusion, the paper successfully demonstrates that size-structured threshold harvesting is an optimal strategy that can simultaneously achieve economic maximization and long-term biological sustainability, even in complex systems with nonlocal density dependence and external recruitment inputs.