Fish School Cohesion Simulation
ISEF Category: Animal Sciences
Ready to Turn This Idea Into a Real Project?
This guide was put together with the help of AI research tools to give you a solid starting point. But a competitive science fair project lives in the details: refining your research question, fine-tuning your variables, analyzing your data, and presenting your findings like a seasoned scientist.
For next steps tailored to your interests, skill level, and timeline, work one-on-one with a MehtA+ mentor. Learn more about MehtA+ Science & Engineering Research Mentorship →
Subcategory: Animal Behavior · Difficulty: Intermediate · Setup: Home Setup · Time: 1 to 2 Months
The Hook
Fish schools can move like one mind, but each fish only follows local cues. A small drop in sensing can spread through the group fast. That makes this topic great for simulation, because you can change one rule and watch the whole school change shape. You can test whether impaired sensing weakens cohesion before you ever touch a tank.
What Is It?
An agent-based simulation treats each fish as a simple actor with rules. One fish senses nearby neighbors, adjusts its direction, and tries to stay with the group. When you slow down sensing or shrink the sensing range, the whole school can loosen, split, or drift.
That helps with microplastic research because you can model damage without running a live exposure test. You are not proving what plastic does in a tank. You are asking whether a specific sensory change can reproduce the movement patterns seen in public zebrafish datasets.
Why This Is a Good Topic
This is a strong science fair topic because the question is clear, measurable, and built for public data. You can compare baseline and impaired schools, then score cohesion with metrics like nearest-neighbor distance, polarization, or split rate. The project links animal behavior, pollution, and modeling, and you can finish a real analysis with only a laptop and open datasets.
Research Questions
- How does reduced neighbor-sensing range change school cohesion in an agent-based zebrafish model?
- What is the effect of sensory response delay on average nearest-neighbor distance?
- Does calibrating the model with public zebrafish trajectories improve its match to observed schooling patterns?
- To what extent do impaired agents increase school splitting and rejoining events?
- Which impairment setting best reproduces the movement statistics of microplastic-exposed zebrafish datasets?
- How does group size change the sensitivity of cohesion to sensory impairment?
Basic Materials
- Laptop or desktop computer with at least 8 GB RAM.
- Python 3.11 and Jupyter Notebook.
- Public zebrafish trajectory dataset.
- Spreadsheet software such as Google Sheets or LibreOffice Calc.
- Free plotting package such as Matplotlib.
- Text editor such as VS Code.
Advanced Materials
- Access to additional zebrafish trajectory datasets from a university lab.
- High-performance workstation or university computing cluster.
- Python scientific stack with pandas, NumPy, SciPy, and Seaborn.
- MATLAB or R for cross-checking the analysis.
- Behavioral tracking exports from a zebrafish facility.
- Version-controlled research repository with Git.
Software & Tools
- Python: Runs the simulation and data analysis.
- Jupyter Notebook: Lets you test assumptions and keep code, charts, and notes together.
- pandas: Organizes trajectory tables and summary metrics.
- NumPy: Handles fast array math for agent updates and metric calculations.
- Matplotlib: Plots cohesion, polarization, and parameter sweeps.
Experiment Steps
- Define the behavior rules your fish agents will follow, and pick the cohesion metric you will score first.
- Choose one sensory change to model, such as shorter sensing range, weaker response, or delayed reaction.
- Calibrate the baseline model against public zebrafish trajectories before adding impairment.
- Plan validation tests that compare simulated and real schools on more than one statistic.
- Run a controlled parameter sweep to see which impairment settings shift cohesion the most.
- Compare your best-fitting model against a few alternate explanations, like random noise or speed change.
Common Pitfalls
- Using only one cohesion metric, which can hide changes in school shape or polarization.
- Letting the impaired model and the baseline model use different starting positions, which makes comparison unfair.
- Calibrating on the same trajectories you later report as a result, which inflates fit quality.
- Treating microplastic exposure as a single catch-all effect, which blurs whether sensing range or response delay matters.
- Ignoring frame-rate or coordinate-scale differences across datasets, which breaks calibration and comparison.
What Makes This Competitive
A stronger version of this project does more than make a pretty simulation. It tests several sensory-impairment rules, holds one dataset out for validation, and reports which rule best matches real trajectories. You can raise the level again by adding uncertainty bands, sensitivity analysis, and a comparison between exposed and control schools. That kind of careful modeling shows you can separate a good story from a good test.
Project Variations
- Model vision loss instead of general sensory loss, then compare the cohesion shift.
- Swap zebrafish for another schooling species dataset, such as guppies or minnows, to see whether the same rules hold.
- Test whether the model fits better when impairment slows turning response instead of shrinking sensing range.
Learn More
- PubMed: Search review articles on zebrafish schooling, social behavior, and microplastic exposure.
- Dryad: Search for open zebrafish trajectory datasets and behavior supplements.
- Zenodo: Find open fish-tracking datasets and code examples.
- NCBI Bookshelf: Read free background chapters on animal behavior, data analysis, and experimental design.
- MIT OpenCourseWare: Look for free material on modeling, simulation, and agent-based systems.