Quantifying Fish School Fragmentation under Predation Using Stochastic Differential Equations

Junyi Qi; Linh Thi Hoai Nguyen; Masayoshi Harada; Thorkil Casse; Ton Viet Ta

REVIEW 2 major objections 2 minor 3 cited by

Quantifying Fish School Fragmentation under Predation Using Stochastic Differential Equations

T0 review · 2 major / 2 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Fish school fragmentation under predation is strategy-dependent and noise-sensitive, quantified through connected-component metrics in a stochastic differential equation model.

desk verdict What reads like a promising fish-school fragmentation study is attached to a full text about graph perception, so there is nothing here to evaluate yet. read the letter →

arxiv 2508.00953 v1 pith:AD34CK2O submitted 2025-08-01 q-bio.QM math.DSq-bio.PE

Junyi Qi , Thorkil Casse , Masayoshi Harada , Linh Thi Hoai Nguyen , Ton Viet Ta This is my paper · ORCID

classification q-bio.QMmath.DSq-bio.PE

keywords fishschoolfragmentationstochasticdifferentialequationsconnectedcomponentspredationgroupcohesionenvironmentalnoisesensitivityanalysis

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper extends an existing stochastic differential equation model of fish schooling to quantify how schools fragment when attacked by predators. It introduces two measurable indicators—first split time and final component count—derived from graph-theoretic connected components, and runs sensitivity analyses over key model parameters. The central finding is that parameter changes affect fragmentation differently depending on whether the predator attacks the nearest fish or the center of the school, and that high environmental noise by itself makes cohesive schooling difficult. If the model faithfully represents real fish behavior, the framework gives ecologists a quantitative way to compare predation strategies and environmental conditions.

What carries the argument

The machinery is a system of stochastic differential equations governing individual fish positions and velocities, augmented by a predator attack rule (nearest fish or school center). Fragmentation is measured by constructing a spatial graph in which fish are nodes and edges represent proximity, then counting the graph's connected components over time. Two scalar indicators summarize the dynamics: the first split time (when the school first breaks into more than one component) and the final component count (how many groups remain after the attack). These indicators are the ones subjected to parameter sensitivity analysis and noise perturbation.

What would settle it

Track a real fish school under a controlled predator encounter (or simulated predator using a robotic fish) and record the time until the school first splits into two groups and the final number of groups. If the empirical first split times and final component counts do not show the same parameter-dependence ordering across nearest versus center attacks as the model predicts, or if high environmental noise does not reduce cohesion in observed schools, the central claim would be contradicted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the structural fragmentation of a fish school under predation can be captured by tracking the number of connected components in a spatial graph of the school, and that this quantity responds to model parameters in a strategy-dependent way. Under a nearest-attack predator, changes in parameters such as attack range or turning rate have one pattern of influence on the first split time and the final number of components; under a center-attack predator, the same parameter changes produce a different pattern. The paper also reports that increasing environmental noise, modeled as random perturbations in the stochastic differential equations, degrades school cohesion regardless of attack strategy. Together these results support the claim that the SDE-plus-connected-components framework is a structured, quantitative basis for assessing how fish schools respond to predation strategies and environmental noise.

Load-bearing premise

The model's stochastic differential equations and the proximity-graph definition of connected components adequately capture real fish schooling behavior under predation; if the model's parameters do not reflect actual fish movement and attack responses, the sensitivity results and the two indicators will not generalize to real schools.

Editorial extensions

If this is right

Under nearest-attack predation, model parameters influence school fragmentation through a different pathway than under center-attack predation, so field studies should record attack style when interpreting school breakup data.
High environmental noise is predicted to reduce school cohesion even in the absence of predators, implying that turbulent or noisy habitats may have a baseline effect on fish group structure.
The two indicators—first split time and final component count—offer ecologists testable summary statistics that can be compared against video-tracked fish schools.
Sensitivity analyses identify which parameters most strongly control fragmentation, guiding future experiments toward measuring those biological quantities.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Because the model treats connected components in a proximity graph, the same indicators could be applied to other group-living animals, such as bird flocks or ungulate herds, if their movement follows similar stochastic dynamics.
The finding that noise disrupts cohesion suggests a testable prediction: fish schools in naturally noisy environments should show shorter first-split times under equivalent predation pressure, which could be verified with existing acoustic or flow-noise data.
The paper does not address individual-level decision rules; a natural next step is to couple the SDE with behavior rules such as attraction-repulsion thresholds, which would change component dynamics but might preserve the qualitative separation between attack strategies.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The manuscript you'd actually need to review isn't here. The abstract describes an SDE study of fish school fragmentation under predation, building on the authors' earlier model and introducing two quantitative indicators (first split time and final component count) plus a sensitivity comparison of nearest versus center attack. But the full text attached is a completely different paper, "Investigating Crossing Perception in 3D Graph Visualisation" (arXiv:2508.00950). This is not a mislabeled section or a partial overlap; it is a graph-visualization paper with no fish, no SDEs, no simulation code, and no fragmentation metrics.

To give credit where it is earned: the abstract is coherent, and the idea of using graph-theoretic connected components to quantify group fragmentation is a plausible and potentially useful extension of a school-modeling program. The two proposed indicators and the attack-strategy comparison are the kind of things behavioral ecologists would want to know. If the actual manuscript ever shows up, it could be worth a careful look.

But the soft spot is load-bearing: none of the mathematical derivations, parameter choices, simulation protocols, or results can be inspected because the supporting text isn't there. The stress-test note is correct that every claim in the abstract rests on a manuscript not present. This is not a minor formatting error; it makes the submission unverifiable on its own terms. Even taking the abstract at face value, there is no mention of validation against empirical fish school data, so the model is self-referential to the authors' prior SDE work. That concern would matter even in a corrected version, but it is secondary to the missing text.

For peer review: this submission should be desk rejected and returned to the authors. There is no point sending it to referees until the correct manuscript is provided. If the authors resubmit with the actual SDE paper, a serious referee could then assess whether the connectivity metrics add predictive power beyond the existing model and whether the simulation results are robust. As it stands, there is nothing to referee.

Bottom line: not reviewable in this state. A corrected resubmission might be a different story, but this version doesn't deserve referee time.

Referee Report

2 major / 2 minor

Summary. The abstract describes a study of fish school fragmentation under predation using a stochastic differential equation (SDE) model, with graph-theoretic connected components as fragmentation metrics and two quantitative indicators (first split time and final component count). It reports sensitivity analyses over key parameters under nearest and center attack strategies, and an independent analysis of environmental noise effects. However, the supplied full text is not the manuscript described in the abstract; it is a different paper on 3D graph visualization (Investigating Crossing Perception in 3D Graph Visualisation, arXiv:2508.00950). Consequently, the technical content of the claimed SDE study—model equations, simulation protocols, sensitivity analyses, and results—is entirely absent from the submission.

Significance. Should the described study be present and correct, a quantitative SDE framework for fish school fragmentation with explicit indicators (first split time, final component count) could be a useful contribution to computational ecology and behavioral biology, particularly in comparing predation strategies and environmental noise. The abstract's proposed design is interesting and potentially falsifiable. However, as submitted, no significance can be assessed: the manuscript body does not contain the model, derivations, simulations, or validation data. The submission therefore provides no support for its claims.

major comments (2)

[Full Text] The full text of the submission is an unrelated paper on 3D graph visualization, not the fish school SDE study advertised in the abstract. This means the model, the simulation implementation, the sensitivity analysis protocols, and the quantitative results (first split time, final component count) mentioned in the abstract are not available for inspection. The central claims are therefore entirely unverifiable from the submitted material. This is a load-bearing defect that prevents review.
[Abstract] The abstract states that the study 'builds upon our previously proposed SDE-based model' but gives no citation or reference to that prior model, and no model equations appear anywhere in the submission. As a result, even the definitions of the key parameters, the noise term, and the two attack strategies cannot be recovered. A specific reference and at least a summary of the model equations are needed to evaluate the sensitivity claims.

minor comments (2)

[Abstract] The phrase 'our previously proposed SDE-based model' should be accompanied by a citation to the prior work so that readers can locate the model.
[Abstract] The indicators 'first split time' and 'final component count' are introduced without precise definitions; a one-sentence formal definition of each would help readers interpret the sensitivity claims.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity can be demonstrated; the provided full text is a different manuscript, and the abstract's SDE derivation chain is absent.

full rationale

The abstract describes a study building on 'our previously proposed SDE-based model' and reports simulation results on fish school fragmentation. The full text submitted is instead a graph-perception paper, 'Investigating Crossing Perception in 3D Graph Visualisation,' containing no SDE model, no fish-school simulation, and no derivation of the fragmentation indicators. Since none of the claimed equations or parameter choices appear, there is no derivation chain in which a prediction can be shown, by the paper's own equations, to equal its inputs. A mismatch between abstract and full text is a serious verifiability problem, but not a circularity: no specific step reduces to a fit, a renaming, or a self-citation chain. Under the requirement to exhibit the specific reduction, no circular step can be identified, so the score is set to 0.

Assumptions & free parameters 2 free parameters · 2 assumptions · 0 invented entities

No new entities are introduced in the abstract. The central claim rests on the unverified modeling assumptions above and on unspecified key parameters.

free parameters (2)

key model parameters (not explicitly listed in abstract)
The abstract states that sensitivity analyses are performed on key parameters but does not specify which parameters or their fitted values in the accessible text.
environmental noise intensity
Environmental noise is examined for its effect on cohesion, but the abstract does not state the values or distribution used.

assumptions (2)

domain assumption The SDE-based model adequately represents fish school dynamics under predation.
The abstract builds on a previously proposed model; the simulation conclusions depend on this modeling assumption.
domain assumption The number of connected components is a valid measure of school fragmentation.
The paper introduces the number of connected components as a graph-theoretic metric without justification in the abstract.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantifying Fish School Fragmentation under Predation Using Stochastic Differential Equations." pith.science (2026). https://pith.science/paper/AD34CK2O

@misc{pith2026250800953,
  author       = {Pith},
  title        = {Pith review of: Quantifying Fish School Fragmentation under Predation Using Stochastic Differential Equations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AD34CK2O}},
  note         = {Machine review of arXiv:2508.00953}
}

read the original abstract

This study builds upon our previously proposed stochastic differential equation (SDE)-based model to further investigate fish school fragmentation under predation. Specifically, we explore structural dynamics by incorporating graph-theoretic metrics--namely, the number of connected components--to quantify changes in prey school organization. Two quantitative indicators, first split time and final component count, are introduced to assess the timing and extent of group disintegration. Sensitivity analyses are performed on key parameters to evaluate their influence on group stability under nearest attack and center attack strategies. We independently examine the effect of environmental noise on fish school cohesion. Simulation results show that parameter changes impact fish school fragmentation differently under the two predation strategies. High environmental noise also makes it difficult for the school to stay cohesive. This framework provides a structured and quantitative basis for assessing how fish schools respond to different predation strategies and environmental noise levels.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Towards High-Fidelity and Controllable Bioacoustic Generation via Enhanced Diffusion Learning
cs.SD 2025-08 reject novelty 5.0 of 10

BirdDiff combines a multi-band enhancement stage with a DiffWave-based diffusion generator and multimodal conditioning, reporting substantially better bird-call synthesis metrics than DiffWave on a 12-species propriet...
Deep Learning for Automated Identification of Vietnamese Timber Species: A Tool for Ecological Monitoring and Conservation
cs.CV 2025-08 unverdicted novelty 4.0 of 10

ShuffleNetV2 achieves 99.29% accuracy and 99.35% F1-score in classifying ten Vietnamese timber species from images, outperforming heavier models in speed-accuracy trade-off.
Mathematical Models for Fish Schooling
cond-mat.stat-mech 2025-08 conditional novelty 1.0 of 10

A self-review of the authors' SDE fish-schooling models with a single simulation example showing a school reaching food.

Reference graph

Works this paper leans on

1 extracted references · 1 canonical work pages · cited by 3 Pith papers

[1]

Investigating Crossing Perception in 3D Graph Visualisation

Investigating Crossing Perception in 3D Graph Visualisation Ying Zhang /envel⌢pe/orcid University of Konstanz, Germany Niklas Gröne /envel⌢pe University of Konstanz, Germany Karsten Klein /envel⌢pe/orcid University of Konstanz, Germany Giuseppe Liotta /envel⌢pe/orcid University of Perugia, Italy Falk Schreiber /envel⌢pe/orcid University of Konstanz, Germa...

work page Pith review arXiv 2012

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

Investigating Crossing Perception in 3D Graph Visualisation

Investigating Crossing Perception in 3D Graph Visualisation Ying Zhang /envel⌢pe/orcid University of Konstanz, Germany Niklas Gröne /envel⌢pe University of Konstanz, Germany Karsten Klein /envel⌢pe/orcid University of Konstanz, Germany Giuseppe Liotta /envel⌢pe/orcid University of Perugia, Italy Falk Schreiber /envel⌢pe/orcid University of Konstanz, Germa...

work page Pith review arXiv 2012