Source data and reproducibility code for "Defended prey cores redirect predator movement: repulsive prey-taxis, local regime selection, and observable dynamics"

This Zenodo record contains the DOI-locked source data and reproducibility code archive for the manuscript: “Defended prey cores redirect predator movement: repulsive prey-taxis, local regime selection, and observable dynamics” DOI: 10.5281/zenodo.20557136Archive version: v2.2 DOI-locked submission package, figure-numbering consistent edition This archive supports a delayed predator–prey reaction–diffusion model with group-defense-mediated repulsive prey-taxis. It provides the source data, numerical check files, reproducibility scripts, figure-consistency files, and documentation used to reproduce the local Turing–Hopf benchmark, cubic coefficient checks, projected Galerkin-DDE branch-label checks, Euler/RK4 solver comparison, modal-resolution robustness summaries, and normalized observable-signature readouts reported in the manuscript. The archive includes: benchmark parameter files and source-data tables; cubic coefficient and branch-selection diagnostic files; projected Galerkin-DDE check files; Euler/RK4 solver-comparison outputs; modal-resolution robustness summaries; normalized observable-signature files; main-text figure files with numbering consistent with the manuscript; figure-consistency hash files and README documentation; README files, transfer notes, and MANIFEST_v2_2.csv; ZENODO_DOI.txt locking this archive to DOI 10.5281/zenodo.20557136; the reproducibility script run_reproducibility_check.py. The current main-text figures are organized consistently with the manuscript numbering: Figure 1: mechanism–dynamics–observables closure for repulsive prey-taxis;Figure 2: local biological regime map;Figure 3: branch-selection and robustness summary;Figure 4: observable signature composite. The numerical files are intended to support local branch-label reproducibility and observable-signature robustness. They are not a certified convergence proof for the full delayed PDE, a global attractor computation, or a far-from-threshold pattern-selection atlas. Recommended use: Unzip the archive, install the dependencies listed in requirements.txt, and run: python run_reproducibility_check.py The script checks the core numerical reproducibility outputs and verifies the main-text figure consistency files used in the manuscript.