An agent-based approach for designing effective protection

This study utilizes an agent-based model to demonstrate that while gradual protection measures have reduced New Zealand dolphin mortality, current efforts remain insufficient to meet conservation goals due to Allee effects, suggesting that a ban on gillnet and trawl fisheries in waters shallower than 100 meters would effectively safeguard most populations.

The Gist

Imagine New Zealand's coastal waters as a giant, bustling neighborhood. In this neighborhood, there are two main groups of residents: the dolphins (specifically Hector's and the critically endangered Māui dolphins) and the fishing boats.

The problem is that the fishing boats are accidentally catching the dolphins in their nets, like a spider catching a butterfly in a web. This has been happening for decades, and the dolphin population is shrinking dangerously.

This paper is like a super-advanced video game simulation created by scientists to figure out how to save these dolphins without shutting down the fishing industry entirely.

Here is the story of their research, broken down simply:

1. The Problem: A Dangerous Neighborhood

The dolphins live in shallow waters close to the shore. The fishing boats (using two types of nets: stationary "gillnets" and moving "trawlers") also work in these same shallow waters.

• Gillnets are like invisible walls of string. Dolphins swim into them, get stuck, and drown.
• Trawlers are like giant vacuum cleaners dragging nets along the bottom. Dolphins are attracted to the fish being stirred up, swim right into the net, and get caught.

Scientists knew this was bad, but they didn't have a perfect way to predict exactly how many dolphins were dying or if the current rules were actually working.

2. The Solution: The "Digital Twin" Neighborhood

Instead of just guessing, the scientists built a computer model (an Agent-Based Model). Think of this as creating a "Digital Twin" of New Zealand's coast.

• The Agents: In this digital world, every dolphin is a character with a personality. They have a home range (a neighborhood they like), they move at specific speeds, and they even have families.
• The Behavior: The model knows that dolphins love shallow water and that they are curious about fishing boats. It also knows that fishing boats move in specific patterns based on real data.
• The "Allee Effect" (The Dating Problem): This is a crucial part of the story. The model includes a rule about finding mates. If the dolphin population gets too small, it becomes hard for a male and female to find each other. It's like living in a tiny town where there are only three people left; it's very hard to find a partner to start a family. This makes it even harder for the population to recover.

3. The Experiment: Testing Different Rules

The scientists ran this digital simulation thousands of times to test different "rules of the road" for fishing. They asked: What happens if we ban fishing in certain areas? What if we ban it everywhere under 100 meters deep?

They compared three scenarios:

1. Current Rules: The protection we have right now.
2. IUCN Rules: A recommendation to ban fishing in all waters shallower than 100 meters (about 330 feet).
3. IUCN+ Rules: The IUCN rules, but with a little extra "buffer zone" added in tricky areas where deep water comes very close to the shore.

4. The Results: The Good, The Bad, and The Ugly

The simulation revealed some surprising truths:

• The Good News: The current protection has helped a little. Fewer dolphins are dying now than in the past.
• The Bad News: The current protection is not enough. The simulation showed that the current rules are like putting a tiny umbrella on a person in a hurricane. The dolphins are still dying at a rate that exceeds national and international safety limits. The "dating problem" (Allee effect) is making it harder for the small populations to bounce back.
• The Ugly News: In some specific areas (like the North Island and the northeast of the South Island), the current rules are practically useless because the fishing boats just move slightly outside the protected zone, or the dolphins wander into the fishing zone.

5. The Verdict: What Needs to Happen?

The computer model gave a clear answer: We need the "IUCN+" plan.

• For most areas: Banning fishing in all waters shallower than 100 meters (the IUCN plan) would almost completely stop the accidental deaths.
• For the most vulnerable areas: We need the IUCN+ plan. This adds a little extra safety buffer in places where the ocean floor drops off quickly. This is the only way to save the critically endangered Māui dolphin and the small Hector's dolphin populations in the northeast.

The Takeaway

Think of the current protection as a fence with big holes in it. The dolphins are smart; they find the holes and swim right into the fishing nets.

This paper says we need to fix the fence. By expanding the protected zones to cover all shallow waters (and adding a little extra buffer in tricky spots), we can stop the accidental deaths. This isn't just about saving a few dolphins; it's about ensuring the "neighborhood" has enough residents to find mates, raise families, and survive for the future.

The scientists used their "digital twin" to prove that if we make these changes, the dolphins can recover. If we don't, the simulation suggests some of these unique populations might disappear forever.

Technical Summary

1. Problem Statement

Fisheries bycatch (entanglement in gillnets and trawls) is the primary threat driving the decline of New Zealand dolphins (Cephalorhynchus hectori), including the critically endangered Māui dolphin and the Hector's dolphin. While area-based protection measures (Marine Protected Areas) have been implemented over the last few decades, their effectiveness remains uncertain.

• Knowledge Gaps: Previous assessments often relied on static habitat models that failed to account for the dynamic movements of dolphins and fishing vessels, the displacement of fishing effort when areas are closed, and the biological reality of small populations (Allee effects).
• Conservation Goals: Current protection levels are failing to meet national (New Zealand) and international (IUCN, US PBR) thresholds for sustainable bycatch. The IUCN has recommended banning gillnets and trawls in waters <100m deep, but the efficacy of this specific boundary for different sub-populations requires rigorous testing.

2. Methodology: Agent-Based Modeling (ABM)

The authors developed a spatially explicit, individual-based Agent-Based Model (ABM) using NetLogo to simulate the interactions between dolphins and fishing fleets. The model was designed and calibrated using Pattern-Oriented Modeling (POM) to ensure emergent behaviors matched real-world data.

Key Model Components:

• Spatial Scale: A 1024 x 1024 patch grid representing New Zealand's coastal waters (1 patch ≈ 1.63 km²).
• Entities:
  • Dolphins: Modeled as "super-agents" representing groups. Attributes include sub-species, group size, age, sex, reproductive status, home range, and movement vectors.
  • Fishing Vessels: Modeled as trawlers (small <15m and medium 16–25m) and gillnetters. Attributes include port of origin, trip duration, speed, and fishing behavior.
  • Gillnets: Static nets with a 24-hour "soak time."
• Behavioral Rules:
  • Dolphin Movement: Dolphins exhibit "wander" behavior within home ranges, constrained by water depth (preference for <100m) and distance from shore. They also exhibit "flocking" behavior, actively following trawlers (attracted by noise and food), which increases bycatch risk.
  • Fishing Effort: Trawlers select fishing grounds based on historical effort data. When protection zones are active, vessels are displaced to adjacent unprotected areas, simulating realistic effort displacement.
  • Bycatch Mechanism: Bycatch occurs when a dolphin group overlaps with a gillnet or a trawler's active net. The probability of capture is calibrated against observer and camera monitoring data.
  • Demographics & Allee Effects: The model includes a mating system where females require a mature male within 5km to conceive. This explicitly models the Allee effect, where low population density reduces reproductive success.

Calibration:

The model was calibrated against 40 years of field data, including:

• Dolphin distribution and abundance surveys.
• Fisheries observer and camera monitoring data (specifically from Banks Peninsula).
• Vessel tracking data (Global Fishing Watch, MPI).
• Parameters were iteratively adjusted until the model's spatial distribution of fishing effort and bycatch rates matched independent government estimates.

3. Key Contributions

• Dynamic Interaction Modeling: Unlike static overlap models, this ABM captures the dynamic feedback loop where dolphins move into/out of protected areas and fishing effort shifts in response to closures.
• Integration of Allee Effects: The model explicitly demonstrates how small population sizes reduce reproductive rates due to mate-finding limitations, a critical factor often ignored in bycatch management.
• Evaluation of Protection Scenarios: The study systematically compares historical protection, current protection, and two future scenarios:
  1. IUCN: Banning fisheries in waters <100m deep.
  2. IUCN+: The IUCN boundary plus a buffer zone in areas where the 100m contour is steep or deep water is close to shore.

4. Results

• Current Protection Inadequacy: While protection has reduced bycatch compared to pre-2020 levels, current measures are insufficient. Bycatch rates for Māui dolphins remain at ~0.6 deaths/year (one death every 1.7 years), exceeding the New Zealand government's Population Sustainability Threshold (PST) and the US Potential Biological Removal (PBR) limit.
• Effectiveness of IUCN Recommendations:
  • Implementing the IUCN (<100m) ban would significantly reduce bycatch for most populations, bringing them close to zero.
  • However, for two specific populations (North Island Māui dolphins and North-east South Island Hector's dolphins), the IUCN boundary is insufficient due to deep-water trenches extending close to shore.
• Superiority of IUCN+: The IUCN+ scenario (adding buffers) effectively eliminates bycatch for all populations, including the two vulnerable groups requiring extra protection.
• Displacement Effects: The model confirmed that closing areas without a sufficient buffer leads to fishing effort displacement, maintaining high bycatch risks in adjacent habitats.
• Allee Effect Impact: Simulations showed that in small populations (e.g., <100 individuals), the reproductive rate drops significantly due to the difficulty of finding mates, accelerating the risk of extinction even if bycatch is reduced.

5. Significance and Implications

• Management Recommendations: The study concludes that New Zealand must adopt the IUCN+ protection level to achieve national goals of zero bycatch and comply with international standards (e.g., US Marine Mammal Protection Act Import Rule).
• Scientific Advancement: This work demonstrates that ABMs are superior tools for conservation planning in complex, spatially dynamic systems. They allow managers to test "what-if" scenarios regarding effort displacement and demographic stochasticity before implementing policy.
• Broader Applicability: The modeling framework is adaptable for other marine species facing bycatch or ship strikes (e.g., Vaquita, Humpback dolphins) and can be extended to assess other anthropogenic threats like noise pollution or habitat modification.
• Policy Impact: The findings provide the empirical basis to reject current "static" protection assumptions and advocate for dynamic, depth-contour-based management that accounts for the specific hydrography of New Zealand's coast.