{"id":"9c23b427-3a26-41cb-94c1-b0010db62abe","arxiv_id":"2608.03685","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"For reaction-diffusion-chemotaxis systems, a stable and non-reactive homogeneous equilibrium can still be destabilized by sufficiently strong chemotactic transport.","lead":"This paper shows that, unlike pure diffusion, chemotaxis can destabilize a homogeneous equilibrium even when the local reaction kinetics is stable and non-reactive. The result separates chemotaxis-driven pattern formation from classical Turing instability and comes with a numerical discretization plus two examples.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The non-necessity claim rests on a single predator–prey simulation with a non-conservative ECDF/IMEX discretization, lacking grid convergence and conservative-scheme comparison; the paper itself flags this exact gap.","rationale":"The reader's weakest assumption is precisely the numerical vulnerability of the predator–prey demonstration: non-conservative ECDF plus explicit IMEX Euler, no convergence checks, no conservative-scheme comparison. My stress-test confirms that this is the load-bearing point, and additionally shows that the analytical Proposition 3 cannot carry the non-necessity result alone because its necessary condition r(J* - k^2 C)>0 is automatically satisfied for any nonzero beta*u*. Thus the central claim hinges on the numerical example, which is exactly the part that lacks verification. The paper's own concluding section explicitly lists the preservation of the diffusion–chemotaxis instability region by the fully discrete scheme as an open question, and Remark 1 acknowledges the non-conservative design choice. Therefore the reader's CONDITIONAL verdict is appropriate: the theory is sane but the pattern-formation claim is conditional on numerical verification. I recommend UNCHANGED because my analysis reinforces, rather than alters, the reader's conditional assessment.","tokens_in":19275,"tokens_out":6419,"duration_ms":72111,"concrete_test":"Rerun the predator–prey experiment with c=1, beta=-13.5, and parameters (32) on the same domain using a flux-conservative, positivity-preserving discretization of the chemotaxis term (e.g., Scharfetter–Gummel or the finite-volume scheme of [10]) and refine the grid from N=50,100,200,400 with ht scaled as O(h^2). Also compute the Neumann-Laplacian eigenvalues k_{m,n}^2 on the 100x100 grid and check whether any lie in the interval where h(k^2)<0 for beta=-13.5. If the stationary hexagon/stripe structures do not persist under conservative refinement, or if no admissible discrete mode is linearly unstable, the claim of chemotaxis-driven patterning from a non-reactive equilibrium is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 3 does not by itself establish the central claim. The necessary condition it proves, r(J* - k^2 C) > 0 for some k, is vacuous for any beta*u* != 0: at large k, H(J* - k^2 C) has off-diagonal -beta*u* k^2/2, so its largest eigenvalue is positive. Thus the proposition only shows that a weak-reactance necessary condition fails to be informative; it does not show that reactivity of J* is not necessary. The actual evidence for non-necessity is the predator–prey experiment (Section 5.2, Figure 7), where the observed hexagon-to-stripe structures are produced by the non-conservative ECDF discretization (20) with explicit IMEX Euler. No grid-convergence study, no flux-conservative/upwind comparison, and no check that the continuous unstable band h(k^2)<0 contains an admissible discrete wavenumber are supplied. For strongly negative beta, the non-conservative expansion of the chemotactic flux and the explicit treatment are precisely the regimes known to generate spurious stationary structures. The paper itself acknowledges in Section 6 that the analogous qualitative-fidelity question for chemotaxis remains open, and Remark 1 explicitly prioritizes algebraic compatibility over mass conservation and positivity. Unless the predator–prey patterns persist under refinement with a conservative scheme, the numerical demonstration that reactivity is not necessary is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a general two-species reaction–diffusion–chemotaxis system and revisits the classical Neubert–Caswell–Murray result that reactivity of the homogeneous equilibrium is necessary for Turing instability. It derives the linear instability condition h(k^2)<0, introduces the mode-dependent matrix J_k = J* - k^2(D+C), and proves in Proposition 3 that a necessary condition for instability is reactivity of J* - k^2 C, with an additional condition under which reactivity of J* itself is forced. The main theoretical claim is that for sufficiently strong chemotaxis (|β|u* > 2√(Du Dv)), reactivity of the local kinetic Jacobian is not necessary for chemotaxis-driven instability. This is illustrated by a predator–prey example in which J* is stable and non-reactive but the chemotaxis threshold β* ≈ -11.02 predicts instability. The paper also extends the matrix-oriented ECDF discretization to chemotactic terms and presents numerical simulations for a Schnakenberg model and the predator–prey model.","tokens_in":19631,"tokens_out":12729,"duration_ms":131387,"significance":"The conceptual point is interesting and the explicit threshold formulas are useful: they clarify that non-symmetric transport can create an instability route that bypasses the classical reactivity condition. The extension of the matrix-oriented Sylvester-equation framework to chemotaxis is a practical contribution, and the paper is transparent about the limitations of its discretization. If the numerical demonstration is made rigorous, the paper would provide a clear counter-example to the expectation that the Neubert–Caswell–Murray result extends to taxis models. However, the logical role of Proposition 3 is overstated, and the predator–prey simulations currently lack the verification needed to support the claim that chemotaxis alone generates continuous-PDE patterns.","major_comments":[{"comment":"The first necessary condition of Proposition 3, namely that J* - k^2 C is reactive for some k≠0, is vacuous whenever βu* ≠ 0. Indeed, H(J* - k^2 C) = H(J*) + (βu* k^2/2)[[0,1],[1,0]], so its largest eigenvalue grows like |β|u* k^2/2 as k→∞. Thus the proposition does not establish that reactivity of J* can fail to be necessary; it only shows that a certain weak-reactance condition is automatically satisfied for any nonzero chemotactic coupling. The actual non-necessity rests on the parameter example in Section 5.2, where J* is stable and non-reactive and condition (9) holds. The paper should present that example as an explicit existence result and clarify that Proposition 3 alone is not the proof of the main claim.","section":"Section 3, Proposition 3"},{"comment":"The central numerical demonstration that chemotaxis 'alone' generates patterns is not yet supported. The simulations use the non-conservative ECDF-compatible discretization (20) with explicit IMEX Euler, and no grid-convergence study, no comparison with a flux-conservative or upwind scheme, and no check of the discrete unstable spectrum are provided. For strongly negative β, centered explicit discretizations of taxis fluxes are a known source of spurious stationary structures, and this is precisely the regime used here. The paper itself acknowledges in Section 6 that the analogous qualitative-fidelity question for chemotaxis remains open, and Remark 1 explicitly prioritizes algebraic compatibility over mass conservation and positivity. Consequently, the statement in Section 6 that Figure 7 provides 'direct numerical confirmation' is premature. At minimum, the authors should show that the","section":"Section 5.2, Figure 7"},{"comment":"The linear instability condition (9) is derived for continuous wavenumbers k, but the numerical experiments are performed on the finite domain [0,2]^2 with homogeneous Neumann boundary conditions. The paper does not verify that the instability band h(k^2)<0 contains an admissible discrete Neumann wavenumber for the parameters used. Although one can check for the reported values (e.g., for β=-15 and γ=500 the band contains k=(π/2)√(m^2+n^2) with m,n≥1), this verification is absent. The ECDF discrete operator may have a slightly different spectrum, so the simulation could in principle be probing a mode that is not present in the continuous problem. This check is straightforward and should be added.","section":"Section 2.1 and Section 5.2"}],"minor_comments":[{"comment":"The sentence 'gu(Pe) =<0' contains a typo ('=<' should be '<'), and shortly after, the critical threshold is given as 'β* ≈ in (29)', with the value missing.","section":"Section 5.1.1"},{"comment":"In the sentence listing the simulated β values, the entry '−12.5,∗,−12' contains a stray asterisk that should be removed.","section":"Section 5.2"},{"comment":"The first-derivative matrix P_x is not skew-symmetric, which is directly related to the non-conservative character of the chemotaxis discretization. The paper should briefly state the resulting lack of discrete mass conservation, even if Remark 1 already prioritizes algebraic compatibility.","section":"Section 4, Equation (18)"},{"comment":"The discussion of the 'stability of J* - k^2 C' in the (η,θ) plane is not fully defined, since J* - k^2 C is non-Hermitian and k is a variable. It would help to specify whether 'stability' means spectral abscissa negativity for some/all k, and to state the relevant k-dependence.","section":"Section 2.3"},{"comment":"The definition of a reactive matrix family J_k is a bit unusual: reactivity is usually a property of a single matrix. The wording should make clear that the definition applies to the parameterized family and that the adjective is transferred from the homogeneous equilibrium.","section":"Definition 1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of a mathematical biology / dynamical systems journal and has several self-citations, which is not problematic in itself. The main risk is the numerical evidence: the paper's own Section 6 and Remark 1 concede the exact gaps that the referee considers load-bearing. If the authors can provide a convergence study and a conservative-scheme comparison, or at least clearly separate the analytically established linear-instability claim from the more tentative numerical pattern-formation claim, the paper would be a solid contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Angela Monti's paper claims that reactivity of the homogeneous equilibrium is not necessary for chemotaxis-driven instability. I think the claim is true, and it is a useful correction to how people read Neubert-Caswell-Murray. But the route the paper takes to reach it is more tangled than it looks, and the numerical part needs work before the pattern-formation claim is solid.\n\nWhat is genuinely new: the explicit statement that the Neubert necessary condition fails when chemotaxis is strong, the parameter counterexample in the predator-prey model, and the matrix-oriented ECDF discretization for the chemotactic flux. The linear algebra in Theorem 2 and the threshold formula are correct. For c=1, the computed J* is stable and non-reactive, and β*≈−11.02; taking β=−15 satisfies the known instability condition, so there is a linear counterexample. That part stands. The discretization in (20) and Proposition 4 are also sound as algebraic extensions of D'Autilia et al.\n\nThe soft spots are real but partly avoidable. Proposition 3's first necessary condition is vacuous: J*−k²C is reactive for large k whenever βu* is nonzero, because the off-diagonal of H grows like k². So the statement that for |β|u* > 2√(DuDv) 'the constraint requiring J* to be reactive disappears' does not follow from the proposition. The conclusion is true, but the proof needs to be the explicit linear instability condition (9) satisfied with a non-reactive J*, not a claim that a vacuous necessary condition tells us something. Section 3 overreaches here.\n\nThe numerical demonstration in Section 5.2 is the weakest part. The hexagon-to-stripe patterns are produced with a non-conservative ECDF discretization and IMEX Euler, no grid-convergence study, no comparison with a flux-conservative scheme, and the paper itself flags in Section 6 that qualitative fidelity is open. Remark 1 prioritizes algebraic compatibility over mass conservation. That does not destroy the paper's main message, because the linear counterexample is independent of numerics, but it does make the nonlinear pattern-formation claim conditional.\n\nNovelty is moderate. The instability condition itself is from Cao-Wu [9], and the author's own [15] apparently contains related transient analysis. What is new here is the explicit non-necessity statement and the computational framework; that is enough for a specialized math-bio audience, but the paper should clearly separate what is new from [15].\n\nRecommendation: send to peer review. A serious referee can force the author to fix Section 3's logic and add at least a convergence check. The result, once cleaned up, is worth citing.","headline":"The non-necessity claim is true and worth knowing, but the paper proves it with the wrong argument and then leans on a numerical experiment it hasn't validated.","tokens_in":20069,"tokens_out":3740,"would_cite":true,"duration_ms":39261,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B36","35K57","37N25","92D25"],"pacs":[],"model":"deepseek-v4-flash","headline":"Chemotaxis can destabilize a non-reactive homogeneous equilibrium, so the classical reactivity requirement for Turing instability does not extend to taxis-driven pattern formation.","keywords":["chemotaxis-driven instability","reactivity","Turing instability","pattern formation","reaction-diffusion-chemotaxis","non-symmetric transport","matrix-oriented discretization","prey-taxis predator-prey patterns"],"falsifier":"Refine the predator–prey simulation on the same domain with N=200, 400 and CFL-adjusted time steps, or repeat with a conservative finite-volume flux for the chemotaxis term, and compare the resulting stationary patterns and their dominant wavelengths with the unstable wavenumber band of J_k for each β below β*. If the hexagon/stripe patterns disappear, change wavelength, or fail to persist under refinement, the claim that chemotaxis alone generates these patterns loses its numerical support.","tokens_in":19142,"feed_emoji":"🦠","tokens_out":9586,"duration_ms":85771,"temperature":0.7,"pith_summary":"Pattern formation theory has a classical necessary condition: a homogeneous equilibrium must be reactive for diffusion alone to destabilize it and create Turing patterns. This paper asks whether that condition survives when directed movement (chemotaxis) is added. It shows it does not: for sufficiently strong chemotaxis, a stable, non-reactive equilibrium can be driven unstable by spatial perturbations even though diffusion alone cannot do it. The proof traces the necessary condition to the reactivity of the shifted operator J*−k²C, with the old reactivity condition recovered only when |β|u* ≤ 2√(DuDv). A predator–prey example with a non-reactive coexistence equilibrium and no Turing instability produces stationary hexagon-to-stripe patterns solely from chemotaxis.","feed_headline":"Chemotaxis alone can drive patterns from stable, non-reactive states","feed_subtitle":"Strong directed movement can destabilize a non-reactive equilibrium even when diffusion alone predicts no patterns.","key_machinery":"The linearized wavenumber operator J_k = J* − k²D − k²C carries the argument; the chemotaxis matrix C = [[0,−βu*],[0,0]] breaks symmetry and shifts the Hermitian part. The necessary-condition proof uses the numerical abscissa r(J_k)=λ_max(H(J_k)) and Weyl's inequality, reducing the threshold to positivity of D+H(C), i.e. |β|u*≤2√(DuDv). Numerically, the ECDF-compatible discrete chemotaxis operator C_ECDF(U,V)=U⊙(T_x V+V T_yᵀ)+(P_x U)⊙(P_x V)+(U P_yᵀ)⊙(V P_yᵀ) reduces to the discrete Laplacian for constant mobility, so the implicit diffusion step keeps its Sylvester-equation form with chemotaxis entering only the explicit right-hand side.","core_discovery":"The paper claims that reactivity of the homogeneous equilibrium is not necessary for chemotaxis-driven instability. For a general two-species model with flux −β∇·(u∇v), the mode-k linearized operator is J_k = J* − k²(D+C), with non-symmetric C = [[0,−βu*],[0,0]]. Since H(J_k)=H(J*−k²C)−k²D, Weyl's inequality gives r(J_k) ≤ r(J*−k²C). Thus a necessary condition for instability is reactivity of J*−k²C, not of J*; reactivity of J* is forced only when |β|u* ≤ 2√(DuDv). Beyond that threshold, a stable, non-reactive equilibrium can be destabilized by chemotaxis alone. The paper shows this in a prey-taxis predator–prey model where the coexistence equilibrium is non-reactive and diffusion-stable, an","pith_inferences":["Inference: any non-symmetric transport operator entering the linearized wavenumber matrix should similarly relax the classical reactivity requirement, so cross-diffusion and advective-taxis models may pattern from non-reactive equilibria.","Inference: a spectrum check on the predator–prey simulations—matching dominant wavelengths to the unstable band of J_k—would let a reader verify the continuous instability independently of the discretization.","Inference: if the mechanism is robust, the same non-reactive pattern-forming regime should appear in a conservative finite-volume discretization or a second-order time integrator, which would extend the numerical evidence beyond the ECDF-IMEX pair used here."],"forward_implications":["In systems with strong taxis (|β|u* > 2√(DuDv)), reactivity of the equilibrium can no longer be used to rule out pattern formation; the relevant quantity is reactivity of J* − k²C.","The classical Turing necessary condition is the weak-coupling limit of the new condition, so diffusion-driven and chemotaxis-driven instabilities remain genuinely different mechanisms.","Repulsive prey-taxis can act as the sole generator of spatial patterns: for the predator–prey model, stationary patterns appear only below β*≈−11.02, in a regime where the equilibrium is stable and non-reactive and diffusion is stable.","The ECDF-compatible chemotaxis discretization keeps the implicit diffusion step in Sylvester form, so matrix-oriented time integration remains efficient for taxis-driven pattern simulations."],"supporting_citations":[{"why":"Supplies the classical theorem that reactivity of the homogeneous equilibrium is necessary for Turing instability, the result this paper tests and partially overturns.","marker":"[36]"},{"why":"Supplies the linear instability conditions for reaction–diffusion–chemotaxis systems on which the necessary-condition proof builds.","marker":"[9]"},{"why":"Supplies the characterization of stable non-reactive equilibria in diffusive-chemotaxis settings and the reactivity analysis used in the predator–prey example.","marker":"[15]"},{"why":"Supplies the prey-taxis predator–prey model and the parameter values used for the chemotaxis-only pattern experiment.","marker":"[47]"},{"why":"Supplies the matrix-oriented Sylvester formulation of implicit diffusion steps that the numerical method extends to the chemotactic flux.","marker":"[14]"},{"why":"Supplies the ECDF discrete Laplacian and its Turing-pattern approximation setting, the basis for the compatible chemotaxis discretization.","marker":"[41]"},{"why":"Supplies the extended central difference formulas and boundary closures used to build the discrete first- and second-derivative matrices.","marker":"[4]"},{"why":"Supplies the warning that fully discrete schemes can generate or suppress continuous Turing instabilities, the numerical-fidelity concern left open for the chemotaxis case.","marker":"[32]"},{"why":"Supplies the point that linear instability conditions alone need not guarantee persistent nonlinear patterns, motivating direct pattern simulations.","marker":"[28]"}],"fun_headline_variants":["Chemotaxis breaks reactivity requirement for instability","Patterns emerge without reactivity via chemotaxis","Non-reactive equilibria destabilize with chemotaxis alone","Stable states can pattern via chemotaxis-driven instability"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The predator–prey patterns shown for β<−11.02 are genuine stationary solutions of the continuous PDE and not artifacts of the non-conservative ECDF discretization combined with the explicit IMEX Euler time step; the paper supplies no grid-convergence study, no flux-conservative comparison, and no check that the discrete unstable wavenumbers are realized on the chosen grid.","fun_headline_variants_meta":{"raw":{"variants":["Chemotaxis breaks reactivity requirement for instability","Patterns emerge without reactivity via chemotaxis","Non-reactive equilibria destabilize with chemotaxis alone","Stable states can pattern via chemotaxis-driven instability"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1221,"prompt_tokens":875,"completion_tokens":346,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":619,"completion_tokens_details":{"reasoning_tokens":284}},"tokens_in":619,"tokens_out":346,"duration_ms":3784,"temperature":1.0,"reasoning_tokens":284,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T14:44:25.972850+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Refine the predator–prey simulation on the same domain with N=200, 400 and CFL-adjusted time steps, or repeat with a conservative finite-volume flux for the chemotaxis term, and compare the resulting stationary patterns and their dominant wavelengths with the unstable wavenumber band of J_k for each β below β*. If the hexagon/stripe patterns disappear, change wavelength, or fail to persist under refinement, the claim that chemotaxis alone generates these patterns loses its numerical support.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the classical theorem that reactivity of the homogeneous equilibrium is necessary for Turing instability, the result this paper tests and partially overturns."},{"cited_title":"CAO ANDJ","cited_arxiv_id":null,"evidence_quote":"Supplies the linear instability conditions for reaction–diffusion–chemotaxis systems on which the necessary-condition proof builds."},{"cited_title":"DIELE, A","cited_arxiv_id":null,"evidence_quote":"Supplies the characterization of stable non-reactive equilibria in diffusive-chemotaxis settings and the reactivity analysis used in the predator–prey example."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the prey-taxis predator–prey model and the parameter values used for the chemotaxis-only pattern experiment."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the matrix-oriented Sylvester formulation of implicit diffusion steps that the numerical method extends to the chemotactic flux."},{"cited_title":"SGURA, B","cited_arxiv_id":null,"evidence_quote":"Supplies the ECDF discrete Laplacian and its Turing-pattern approximation setting, the basis for the compatible chemotaxis discretization."},{"cited_title":"AMODIO ANDI","cited_arxiv_id":null,"evidence_quote":"Supplies the extended central difference formulas and boundary closures used to build the discrete first- and second-derivative matrices."},{"cited_title":"Turing-region preservation in matrix-oriented splitting methods for reaction-diffusion systems","cited_arxiv_id":"2606.23211","evidence_quote":"Supplies the warning that fully discrete schemes can generate or suppress continuous Turing instabilities, the numerical-fidelity concern left open for the chemotaxis case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the point that linear instability conditions alone need not guarantee persistent nonlinear patterns, motivating direct pattern simulations."}],"review_version":1}