Pith. sign in

REVIEW 3 major objections 6 minor 36 references

Subcritical Turing bifurcation and the morphogenesis of localised patterns

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Adding source and loss terms to a simple reaction–diffusion system makes its Turing bifurcation subcritical, which turns the onset of pattern formation into a jump to stable localised spots and stripes.

desk verdict A credible but not fully self-contained demonstration that source/loss terms make Turing bifurcations subcritical and produce homoclinic snaking; the decisive coefficient sign is outsourced to a thesis. read the letter →

arxiv 2506.07855 v1 pith:42T5SV5X submitted 2025-06-09 nlin.PS

Víctor Breña-Medina , Alan Champneys This is my paper
classification nlin.PS MSC 35B3635K5737L20
keywords reaction-diffusionsystemsTuringbifurcationsubcriticalhomoclinicsnakinglocalisedpatternsHamiltonian-HopfSchnakenbergmodelanalysis
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 argues that the same mechanism behind homoclinic snaking—spatial reversibility combined with a subcritical Hamiltonian–Hopf bifurcation—arises naturally in reaction–diffusion systems once realistic source and loss terms are included. In the generalised Schnakenberg system the balance of the source term $b$ and removal rate $r$ makes the Turing bifurcation subcritical in an open parameter region. Subcriticality is signalled by $\sigma = q_1 q_3 > 0$ in the Lyapunov–Schmidt reduced bifurcation equation. On long domains this turns the usual Turing instability into a spontaneous jump to finite-amplitude, localised spot and stripe patterns, via homoclinic snaking in one and two dimensions.

What carries the argument

The argument is carried by the reduced bifurcation function $g(z,\mu)=q_1\mu z+q_3z^3+q_5z^5+\cdots$, obtained by Lyapunov–Schmidt reduction of the generalised Schnakenberg system; its coefficients determine $\sigma=q_1q_3$, which fixes the pitchfork criticality. The same normal form appears as the spatial amplitude equation near a Hamiltonian–Hopf bifurcation, so the sign of $\sigma$ connects the Turing instability to the homoclinic snaking mechanism. The system's spatial reversibility supplies the symmetry needed for symmetric localised states, while the absence of variational structure removes asymmetric ladders.

What would settle it

Recompute the Lyapunov–Schmidt coefficients for system (8) independently and evaluate $\sigma=q_1q_3$ along the Turing curve; if the sign differs from the paper, the central claim fails. Alternatively, a direct numerical continuation in a long one-dimensional domain could look for the predicted fold (limit point) and snaking branches only when $\sigma>0$, and their absence for $\sigma<0$ would settle the matter.

Watch

Extended reading notes

Core claim

The paper's central claim is that subcritical Turing bifurcations are not a special feature of models with rational nonlinearities but a generic consequence of balancing production and removal in a pure-power reaction–diffusion system. For system (8), the bifurcation is supercritical when $\sigma = q_1 q_3 < 0$ and subcritical when $\sigma > 0$, with the transition occurring at a codimension-two point in the $(k^2,D_1)$ plane. In the subcritical regime, the long-domain problem is equivalent to a reversible Hamiltonian–Hopf bifurcation, and the normal-form prediction of infinitely many homoclinic orbits is realised as stable multi-pulse and multi-spot states organised in a snaking bifurcation diagram without variational ladders.

Load-bearing premise

The load-bearing premise is that the coefficient calculation giving the sign of $\sigma=q_1q_3$ is correct, since the paper points to an omitted derivation in a thesis; if that sign is wrong, the subcritical-to-snaking connection collapses.

Editorial extensions

If this is right

  • Long-domain reaction–diffusion models with source and loss terms can spontaneously form localised patterns rather than extended ones, making the transition hysteretic and robust to parameter drift.
  • The snaking region in the $(k^2,D_1)$ plane predicts stable even- and odd-spike branches that annihilate in fold bifurcations, so the number of spikes is selected by initial conditions and history.
  • In two dimensions, equal source and loss ($b=r$) is the operative condition for the localised spots and stripes found, indicating a preferred balance for pattern localisation.
  • Without variational structure, asymmetric 'ladder' states are absent, so only symmetric localised states exist, a testable difference from Swift–Hohenberg systems.

Reading between the lines

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

  • If subcriticality is as generic as the paper suggests, many reaction–diffusion systems with explicit production and degradation terms may need re-examination: localised patterns could appear even where the classical supercritical Turing analysis predicts only extended patterns.
  • The $b=r$ restriction seen in the two-dimensional simulations hints at a conservation-like constraint; testing systems with unequal source and loss but different parameter scalings could reveal whether the localisation mechanism needs exact balance or merely a proximity.
  • The mechanism can be probed experimentally by looking for hysteresis in pattern onset: a subcritical Turing system should show a jump to finite-amplitude localised spots when control parameters are swept in opposite directions.
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.

Referee Report

3 major / 6 minor

Summary. This paper proposes that the inclusion of balanced source and loss terms in a two-species reaction-diffusion system (generalized Schnakenberg system, Eq. (8)) makes the Turing bifurcation subcritical in an open parameter region, which in long domains corresponds to a Hamiltonian-Hopf bifurcation leading to homoclinic snaking and stable localized spot/stripe patterns. The sub/supercritical threshold is predicted by the sign of sigma = q1 q3 from a Lyapunov-Schmidt reduction (Eq. (12)). Numerical bifurcation diagrams in 1D show a subcritical pitchfork, a fold, and snaking branches with stable regions, and 2D simulations show isolated spots and a stripe. The main theoretical explanation rests on the coefficients q1, q3, q5, whose derivation is deferred to the first author's PhD thesis [30].

Significance. If the central coefficient calculation is correct, the paper offers a generic and conceptually simple mechanism linking mass non-conservation (source/loss) to subcritical Turing bifurcations and localised pattern formation, with plausible biological implications. The paper includes explicit bifurcation diagrams, stability computations, and parameter-region plots, which are valuable evidence for the phenomenon at selected parameter values. However, the decisive normal-form coefficients are not derived in the manuscript, and the 2D evidence is anecdotal, so the broader claims are not yet fully substantiated.

major comments (3)
  1. [Section III, Eq. (12) and following text] The sign of sigma = q1 q3, which determines the subcritical/supercritical transition and the vertical line at k2 = 1.5226 in Fig. 2(b) and the boundary of the snaking region in Fig. 3(a), is not derived in the manuscript; the reader is referred to the first author's thesis [30] with no formulas for q1, q3, or q5. Because this coefficient is the load-bearing predictor of the claimed open subcritical region, the paper should either include the explicit Lyapunov-Schmidt coefficient formulas (or a supplementary derivation) or provide an independent numerical computation of sigma over the claimed region. The existence of subcritical branches at the specific computed parameter values does not by itself establish the open-region claim.
  2. [Section IV and Abstract] The abstract states that the mechanism leads to localised patterns via the homoclinic snaking mechanism, but the only snaking bifurcation diagram is one-dimensional (Fig. 3). The two-dimensional results are four individual simulations (Fig. 4(a)-(d)) with no continuation, no snaking region, and an explicit acknowledgement that 'a full exploration ... is left for future work' and that no localised patterns were found for b != r. This gap should be filled with at least one two-parameter continuation or snaking diagram in 2D, or the abstract and conclusions should be reworded to avoid implying a demonstrated 2D snaking mechanism.
  3. [Section III, paragraph after Fig. 2(b)] The claim that 'q1 q5 is negative in all the entire parameter region of interest' is not backed by any plot, formula, or table. Since q1 q5 < 0 is needed for the unfolding to predict snaking, this check should be documented, for example by plotting the sign of q1 q5 over the (k2, D1) region shown in Fig. 3(a).
minor comments (6)
  1. [Eq. (4)] The formula for the wavenumber kappa appears inconsistent with the double-root condition in Eq. (5); it should likely involve D1 D2 in the denominator and a factor of 1/2. Please correct or clarify.
  2. [Section II] The statement that 'the condition for a criticality of the Turing bifurcation at the double root is precisely the same as the condition for the criticality of the corresponding Hamiltonian-Hopf' is asserted without derivation; please add a short argument or a more precise reference than [25].
  3. [Section III] The unproven assertion that 'otherwise Turing bifurcations are always supercritical in straightforward Schnakenberg and Gray-Scott systems' should be substantiated with a short calculation or a reference.
  4. [Section IV] The 2D simulations are described only as 'a finite difference method implemented in matlab'; please specify the time-stepping scheme, boundary conditions, grid resolution, and convergence criteria.
  5. [References] Reference [30] is a PhD thesis; please state whether it is openly accessible and provide a stable link if possible.
  6. [Throughout] There are minor grammatical and typographical errors (e.g., 'adn' in Ref. [21] and a missing word in the Introduction); these should be corrected in a final pass.

Circularity Check

1 steps flagged · score 6.0 of 10

Sub/supercritical transition is decided by self-cited q1q3 coefficients not derived in the paper; numerical continuation gives partial, not complete, independent support.

  1. self citation load bearing [Section III, text following Eq. (12) and in the σ discussion before Fig. 2(b)]
    "The calculation of the coefficients qi is straightforward but lengthy, full details are available in [30], we omit the details for brevity. ... In the parameter region under investigation the sub/super-criticality of the pitchfork (Hamiltonian–Hopf) bifurcation is determined by the sign of σ = q1q3, see [24]."

    The paper's central claim—that source/loss balance makes the Turing bifurcation subcritical in an open (k2,D1) region—is carried by the sign of σ=q1q3. No formulas for q1 and q3 are derived or given; the reader is referred to the first author's PhD thesis [30]. The criterion q1q3>0 is standard [24], but the coefficient values, hence the transition line and the snaking-region boundary, are imported from an unpublished self-cited source. An error there would shift or destroy the claimed region and the source/loss explanation. However, the numerical continuation in Fig. 2(b) and snaking branches in Fig. 3(b) independently show subcriticality and localised states at the chosen parameters, so this is load-bearing self-citation, not full equivalence-by-definition.

full rationale

The paper's derivation chain is honest up to the critical point: it derives the linear conditions, the Hamiltonian–Hopf correspondence, and the form of the bifurcation equation (12), and it clearly labels the omitted calculation of the coefficients qi as available only in the first author's thesis [30]. Those coefficients determine the sign of σ = q1q3, which is the paper's handle on the open subcritical region and the predicted snaking boundary. Since the decisive numbers are not independently checkable inside the paper, the self-citation is load-bearing; however, the paper does not present q1q3 as a fitted parameter or define it in terms of the outcome it predicts. The reported numerical continuation (Fig. 2(b)) and homoclinic snaking diagram (Fig. 3(b)) at chosen parameters provide independent confirmation of the phenomenon at those points, and the 2D section is explicitly preliminary: only isolated 2-spot, 4-spot, 8-spot, and stripe examples are shown, with the authors stating that 'A full exploration in the spirit of [13–15] though is left for future work.' No other circular step is identifiable: the equivalence of long-domain Turing accumulation and Hamiltonian–Hopf is a standard cited result, and the physical argument about source and loss robustness is an interpretation, not an equation-level identity. The score therefore reflects a central self-cited coefficient calculation, not a pure definitional or fitting circularity.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central normal-form computation is imported from a self-authored thesis, and the parameter values are illustrative; these are the main inputs the reader must accept without independent verification.

free parameters (1)
  • Illustrative model parameters (b, c, r, h, D1, D2, k2) = 1D: b=c=r=h=1, D1=0.1, D2=10; 2D: c=0.1, h=0.01, D2=10
    Hand-chosen to exhibit the phenomenon; no data fitting or systematic parameter study is provided, so the claimed genericity is not quantified.
assumptions (4)
  • standard math The double-root Turing condition (5) is equivalent to the Hamiltonian-Hopf resonance condition in the L to infinity spatial dynamics (Sec. II).
    This equivalence is a known theorem from spatial dynamics, cited via [24] and [25]; the paper assumes it without proof.
  • domain assumption Reflection symmetry in x ensures the reduced bifurcation function is odd in z, so quadratic terms vanish (Sec. III, after Eq. 11).
    The model has quadratic nonlinear terms, but the Z2 symmetry is used to remove them from the reduced equation; this is standard but essential, and is not explicitly proved.
  • ad hoc to paper The Lyapunov-Schmidt coefficients q1, q3, q5 are as stated, with computation deferred to the first author's thesis [30] (Sec. III, Eq. 12).
    This is the load-bearing normal-form calculation; the paper provides no derivation or data to verify it.
  • domain assumption The numerically identified snaking region in Fig. 3(a) is accurate and complete enough to support the conclusions (Sec. III, Fig. 3).
    Only a sketch of the computation is given; no continuation code or error estimates are provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Subcritical Turing bifurcation and the morphogenesis of localised patterns." pith.science (2026). https://pith.science/paper/42T5SV5X

@misc{pith2026250607855,
  author       = {Pith},
  title        = {Pith review of: Subcritical Turing bifurcation and the morphogenesis of localised patterns},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/42T5SV5X}},
  note         = {Machine review of arXiv:2506.07855}
}
read the original abstract

Subcritical Turing bifurcations of reaction-diffusion systems in large domains lead to spontaneous onset of well-developed localised patterns via the homoclinic snaking mechanism. This phenomenon is shown to occur naturally when balancing source and loss effects are included in a typical reaction-diffusion system, leading to a super/subcritical transition. Implications are discussed for a range of physical problems, arguing that subcriticality leads to naturally robust phase transitions to localised patterns.

Figures

Figures reproduced from arXiv: 2506.07855 by the authors.

Figure 1
Figure 1. FIG. 1. (Color online) The equivalence between (a) the mode [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. (Color online) (a) Dispersion relation of [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. (Color online) (a) The snaking region (shaded) inside [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4. (Color online) Stable patterns for a two-dimensional [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [30]

    Singularities and Groups in Bifurcation Theory, volume 1

    M.GolubitskyandD.Schaeffer. Singularities and Groups in Bifurcation Theory, volume 1. New York: Springer– Verlag, 1985

    work page 1985

  2. [34]

    Plaza, F

    R. Plaza, F. Sánchez–Garduño, P. Padilla, R.A. Barrio and P.K. Maini.J. Dyn. Diff. Equat., 16(4)

    work page

  3. [1]

    snaking region

    We suppose that at parameter value µ = µc there is a double zero eigenvalue of Aκ, corresponding to a large mode number mc, not necessarily an integer. The condition for such a double root is (D2a11 + D1a22)2 = 4D1D2 det A. (5) Ontheonehand, fora L ≫ 1therewillbealargenumber of Turing bifurcations for nearbyκ-values corresponding to κ = mπ/L for integers ...

    work page

  4. [2]

    Kondo and T

    S. Kondo and T. Miura.Science, 329: 616–1620, 2010

    work page 2010

  5. [3]

    J.D. Murray. Mathematical Biology II: spatial models and biomedical applications. Springer–Verlag, New York, 3rd. ed., 2002

    work page 2002

  6. [4]

    Vanag and I.R

    V.K. Vanag and I.R. Epstein.Chaos, 17: 037110, 2007

    work page 2007

  7. [5]

    Turing.Phil

    A.M. Turing.Phil. Trans. Roy. Soc. Lond. B, 641: 37–72, 1952

    work page 1952

  8. [6]

    Purwins, H.U

    H.G. Purwins, H.U. Boedeker, and Sh. Amiranashvili. Advances in Physics, 59(5): 485–701, 2010

    work page 2010

Show all 36 references
  1. [7]

    J.S. Bois, F. Julicher and S.W. Grill Phys. Rev. Lett. 106(2): 028103, 2011

    2011

  2. [8]

    Khain and L.M

    E. Khain and L.M. Sander. Phys. Rev. Lett., 96(18): 188103, 2006

    2006

  3. [9]

    R. Hoyle. Pattern Formation: An Introduction. Cam- bridge University Press, Cambridge, 2006

    2006

  4. [10]

    M. Beck, J. Knobloch, D.J.B. Lloyd, B. Sandstede, T. Wagenknecht,SIAM J. Math. Anal.41(3): 936–972, 2009

    2009

  5. [11]

    Woods and A.R

    P.D. Woods and A.R. ChampneysPhysica D 129(3–4): 147–170, 1999

    1999

  6. [12]

    Burke and E

    J. Burke and E. Knobloch,Phys. Rev. E73(5): 056211, 2006

    2006

  7. [13]

    Burke and E

    J. Burke and E. Knobloch.Chaos, 17: 037102, 2007

    2007

  8. [14]

    Avitabile, D.J.B

    D. Avitabile, D.J.B. Lloyd, J. Burke, E. Knobloch, and B. Sandstede,SIAM J. Appl. Dyn. Syst.9(3): 704–733, 2010

    2010

  9. [15]

    Lloyd, B

    D.J.B. Lloyd, B. Sandstede, D. Avitabile and A.R. Champneys, SIAM J. Appl. Dyn. Syst. , 7(3): 1049–1100, 2008

    2008

  10. [16]

    Lloyd and B

    D.J.B. Lloyd and B. Sandstede,Nonlinearity 22: 485– 524, 2009

    2009

  11. [17]

    Schneider J.F

    T.M. Schneider J.F. Gibson and J. Burke Phys. Rev. Lett., 104(10): 104501, 2010

    2010

  12. [18]

    Beaume, A

    C. Beaume, A. Bergeon, and E. Knobloch.Phys. Fluids, 23(9): 094102, 2011

    2011

  13. [19]

    Gomila, P

    D. Gomila, P. Colet, M.S. Miguel, A.J. Scroggie, and G.–L. Oppo.IEEE J.Q.E., 39(2): 238–244, 2003

    2003

  14. [20]

    Houghton and E

    S.M. Houghton and E. Knobloch Phys. Rev. E 84(1): 016204, 2011

    2011

  15. [21]

    Y. Mori, A. Jilkine and L. Edelstein–Keshet SIAM J. Appl. Math., 71(4): 1402–1427, 2011

    2011

  16. [22]

    Yochelis, Y

    A. Yochelis, Y. Tintut, L.L. Demer adn A. Garfinkel, New J. Phys.10: 055002, 2008

    2008

  17. [23]

    Yochelis and A

    A. Yochelis and A. Garfinkel, Phys. Rev. E 77: 035204(R), 2008

    2008

  18. [24]

    Lloyd and H

    D.J.B. Lloyd and H. O’Farrell, Physica D 253: 23–39, 2013

    2013

  19. [25]

    Iooss, and M.–C

    G. Iooss, and M.–C. Perouème.J. Diff. Eqns., 102(1): 62–88, 1993

    1993

  20. [26]

    Burke and E.Knobloch.Discrete and Continuous Dyn

    J. Burke and E.Knobloch.Discrete and Continuous Dyn. Syst., Supplement 170–180, 2007

    2007

  21. [27]

    Dean, P.C

    A.D. Dean, P.C. Matthews, S.M. Cox, and J.R. King. Nonlinearity, 24(12): 3323–3351, 2011

    2011

  22. [28]

    Kozyreff, and S.J

    G. Kozyreff, and S.J. Chapman.Phys. Rev. Lett., 97(4): 044502, 2006

    2006

  23. [29]

    Payne and C.S

    R.J.H. Payne and C.S. Grierson. PLoS ONE, 4(12): e8337. doi:10.1371/journal.pone.0008337, 2009

    2009 doi

  24. [31]

    Breña–Medina

    V.F. Breña–Medina. PhD Thesis, University of Bristol, 2013

    2013

  25. [32]

    Nishiura and D

    Y. Nishiura and D. UeyamaPhysica D130(1–2): 73–104, 1999

    1999

  26. [33]

    Breña–Medina, D

    V. Breña–Medina, D. Avitabile, A.R. Champneys, and M.J. Ward, arXiv:1403.5318 [nlin.PS]

    arXiv

  27. [35]

    Breña–Medina, A.R

    V. Breña–Medina, A.R. Champneys, C. Grierson, M.J. Ward, SIAM J. Appl. Dyn. Syst.13(1): 210–248, 2014

    2014

  28. [36]

    Dawes.SIAM J

    J.H.P. Dawes.SIAM J. Appl. Dyn. Syst.7(1): 186–206, 2008

    2008

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.