Pith. sign in

REVIEW 4 major objections 4 minor 22 references

Periodicity in delayed self-regulation is a predator-prey process

T0 review · 4 major / 4 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 Periodic solutions of delayed self-regulation DDEs reduce to planar predator-prey ODEs on each periodic branch.

desk verdict A big structural result with a genuine, likely repairable gap in the proof of the nesting theorem. read the letter →

arxiv 2506.07660 v1 pith:PX5C3MUW submitted 2025-06-09 math.DS math.CA

Alejandro López-Nieto This is my paper
classification math.DSmath.CA MSC 34K1334K1837G1534C25
keywords delaydifferentialequationsmonotonedelayedfeedbackperiodicorbitspredator-preysystemstime-amplitudeparametrizationPoincaré-BendixsontheoryHopfbifurcationintegrableplanarODEs
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

The paper claims that periodic oscillations in scalar delay equations with monotone delayed feedback are, branch by branch, two-dimensional predator-prey cycles. It proves that every connected component of periodic solutions is an annulus with global time and amplitude coordinates, so each amplitude determines exactly one periodic orbit per branch. On the annulus, the two-point projection $(x(t), x(t-1))$ satisfies an integrable planar ODE with a predator-prey feedback condition, and the full set of periodic solutions is encoded by two functions of amplitude: the period map and the delay map. If correct, this gives a complete two-dimensional reduction of periodicity in this class of delay equations and explains how periodic branches are born at Hopf and homoclinic points and how they are related by time rescalings.

What carries the argument

The central machinery is the extended DDE with the delay $r$ as a state variable, together with the two-point projection $P(\phi) = (\phi(0), \phi(-1))$ that embeds periodic orbits as Jordan curves. The time-amplitude parametrization $\beta(t,a) = (x_t^a; r(a))$ is the global chart on each branch, the amplitude function $\alpha$ is a first integral on the cyclicity component $O$, and the period map $p(a)$ together with the delay map $r(a)$ encode the branch shape. The time rescaling symmetry $x((1+mp)t)$ generates all branch copies and yields the delay-independent nesting theorem that separates cyclicity components.

What would settle it

Track the first periodic branch of the Hutchinson equation by numerical continuation: if the branch ever contains two periodic solutions with the same amplitude but different delays, or if the projected loops $(x(t), x(t-1))$ from two amplitudes on one branch intersect without being identical, the annulus time-amplitude picture would fail.

Watch

Extended reading notes

Core claim

For the scalar delay equation $\dot x(t) = r f(x(t), x(t-1))$ with monotone delayed feedback, the paper shows that each connected periodic branch $B$ of the extended system, where $r$ is a constant state variable, is an annulus: the map $(t,a) \mapsto (x_t^a; r(a))$ is a global diffeomorphism, and the amplitude $a$ labels unique periodic orbits. The projection $\bar P(\phi; r) = (\phi(0), \phi(-1))$ embeds $B$ onto a cyclicity component $O$, and the projected curves solve the planar ODEs $(\dot u, \dot v) = r(\alpha(u,v)) (f(u,v), g(u,v))$ with $\partial_1 g \, \partial_2 f < 0$, making $\alpha$ a first integral equal to the amplitude. Corollary 2.6 then characterizes every periodic solution by two time maps $p(a)$ and $r(a)$: any periodic solution satisfies $\alpha(x(t), x(t-1)) = a$, the pair $(x(t), x(t-1))$ solves the planar ODEs with $\dot x(t-1) = r(a) g(x(t), x(t-1))$, and $v(t) = u(t - r(\alpha))$ with minimal period $p(\alpha) r(\alpha)$.

Load-bearing premise

The load-bearing premise is the Poincaré–Bendixson structure of these delay equations: every periodic orbit must project to a non-self-intersecting loop whose maximum and minimum occur once per period and are crossed transversely; if any periodic orbit violated this, the time-amplitude coordinates and the amplitude ordering of branches would collapse.

Editorial extensions

If this is right

  • Every connected periodic branch has no amplitude turning points: each amplitude labels exactly one periodic orbit, so isolated loops of periodic solutions cannot form.
  • The full set of periodic solutions is governed by the cyclicity set $C$: an amplitude first integral $\alpha$ and two time maps $p(a)$, $r(a)$ determine periods and delays, and all branch copies satisfy $\hat r = (1+mp)r$ and $\hat p = p/(1+mp)$.
  • New periodic solutions appear in only two ways: at branch boundaries through Hopf or homoclinic bifurcations, or in the interior at saddle-node bifurcations located at critical values of the delay map $r(a)$.
  • The slow branch, whose period map satisfies $p>2$, contains the slowly oscillating periodic solutions; these are the only periodic solutions that can be exponentially attracting and can accumulate on homoclinic orbits.
  • Because the delay is a time scaling in the reduced planar ODEs, periodicity in the infinite-dimensional DDE is governed by a finite-dimensional integrable structure, unifying delayed self-regulation with predator-prey feedback.

Reading between the lines

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

  • If the reduction is correct, standard tools for planar integrable systems, such as period-map signatures, could be used to read off the connection graph of a DDE from $r(a)$ and $p(a)$, mirroring known results for rotating waves on a circle.
  • The sign of $r'(a)$ may provide a computable proxy for how the unstable dimension of the periodic orbits changes along a branch, up to orientation; the paper only suggests the correlation and leaves the orientation undetermined.
  • Near homoclinic amplitudes, the rescaled branch copies with $m \neq 0$ should exhibit spike-like, temporally localized solutions; direct numerical simulation of the paper's QRT example near the figure-eight could confirm this behavior.
  • A natural next question is which subsets of $\mathbb{R}^2$ are realizable as cyclicity sets; if realizability is constrained, that would sharply restrict the possible global bifurcation diagrams of monotone delayed feedback DDEs.
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

4 major / 4 minor

Summary. The paper studies scalar delay differential equations ẋ(t) = r f(x(t), x(t-1)) with monotone delayed feedback. Its main claim is that every connected component ("branch") of the periodic set in the extended phase space C × R is an annulus admitting global time-amplitude coordinates (Theorems 2.1–2.2), that the projection (x(t), x(t-1)) satisfies integrable planar ODEs with a predator-prey sign condition (Theorem 2.3), and that distinct branches have either disjoint projections or are related by the time-rescaling symmetry (Theorem 2.5), leading to a complete two-time-map description of all periodic solutions (Corollary 2.6). The local continuation arguments and the global branch projection are developed in detail. However, the proof of Theorem 2.5 contains a major unproved perturbation step, an apparent misuse of the Poincaré–Bendixson theorem, and an unjustified application of Lemma 5.2, so the claimed complete characterization of all periodic solutions is not established as written.

Significance. If correct, the paper would give a striking reduction: periodicity in an infinite-dimensional monotone delayed-feedback DDE would be fully described by planar integrable ODEs with known period and delay maps, generalizing the author's symmetric-feedback examples. Strengths include the careful local continuation setup (Lemmas 3.4–3.5, 4.3–4.5), the explicit treatment of hyperbolic and nonhyperbolic cases, and the concrete examples (enharmonic oscillator, QRT Hamiltonian) that illustrate the claimed structure. The flaw in Theorem 2.5 is load-bearing for Corollary 2.6; without a repaired proof of delay-independent nesting, the completeness claim is unsupported.

major comments (4)
  1. [Section 6, Eq. (6.11)] The proof of Theorem 2.5 reduces to the generic Hopf case by replacing f with \tilde f = f + ε ι(u,v)(v-a) in Eq. (6.11) and asserting that "our previous analysis holds" for small ε. This is not shown. Theorems 2.1–2.3 and Lemmas 3.4–3.5 are proved for the fixed nonlinearity f; no persistence result states that the perturbed DDE has branches B_ε and \hat B_ε whose cyclicity components intersect at an infimum a_ε → a, nor that the Hopf relation \hat r_ε(a_ε) = (1 + m p_ε(a_ε)) r_ε(a_ε) survives the limit ε → 0. The degenerate alternative ∂1f = -∂2f with double ν = 0 is exactly what the perturbation removes, so without a continuity argument the original system may lie in that degenerate case. Since Eq. (2.13) and Corollary 2.6 rest on this proof, the gap is load-bearing.
  2. [Section 6, after Eq. (6.8)] The proof claims that if x_a(t) is nonconstant and \hat x_a(t) is constant, then "the existence of a nonconstant periodic solution x_a(t) contradicts the Poincaré–Bendixson theorem [14, Theorem 2.1]". This is not a consequence of [14]: monotone delayed-feedback DDEs such as the Hutchinson equation (1.1) have both an equilibrium and a nonconstant periodic orbit. In addition, x_a(t) is obtained as a C^1 limit at the infimum of the amplitude interval, so it need not belong to the branch B; the assertion that it "admits a local continuation for smaller values of the amplitude a" requires proving that the continuation lies in B (and in \hat B) and has amplitude below the infimum, which is not supplied.
  3. [Section 6, Lemma 5.2] The constancy of p(a)r(a) and \hat p(a)\hat r(a) near the infimum is attributed to Lemma 5.2, but Lemma 5.2 is stated for a single fixed orbit γ* and its own b-continuation γ̃_b under the hypothesis Pγ* ∩ Pγ̃_b ≠ ∅ for all b in an interval. In Theorem 2.5, B and \hat B are two distinct branches, and the assumption \bar P B ∩ \bar P \hat B ≠ ∅ only gives an intersection of some two curves; it does not imply that a fixed projected orbit intersects an entire one-parameter family through an interval. The hypotheses of Lemma 5.2 are therefore not verified, so the constancy of the period-delay products and the subsequent Arzelà–Ascoli limit are not justified.
  4. [Section 6, Eq. (6.12)] The displayed characteristic equations are written as ν = r(a)(∂1f(a,a) + e^{-ν}∂2f(a,a)) and \hat ν = \hat r(a)(∂1f(a,a) + e^{-\hat ν}∂2f(a,a)). This does not match the characteristic equation (6.10), which has iν on the left and e^{-iν} in the exponential. Even after correcting this typo, the step from two Hopf characteristic equations to \hat r(a) = (1 + m p(a))r(a) is not derived; one has to rule out the possibility of distinct frequencies solving the same characteristic equation and to prove the relation holds for all a in the branches, not only at the infimum. The closing statement that "uniqueness of the Hopf branch completes the proof" is an assertion, not a demonstrated argument.
minor comments (4)
  1. [Section 5, first paragraph] The word "Converserly" should be "Conversely".
  2. [Section 2, Example 2.9] The sentence "the phase portrait is of (2.25) is of double-well" contains a duplicated "is of" and should be rephrased.
  3. [Section 3, proof of Lemma 3.4] The Poincaré return map and the semiflow are both denoted S (e.g., Eq. (3.26) versus (3.1)); using different symbols would improve readability.
  4. [Section 7, Eq. (7.7)] The integral is written as ∫_0^t yT(t)ℓ(t) dt with the same letter t for the upper limit and the integration variable; it should be ∫_0^t yT(s)ℓ(s) ds.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation found: main theorems are proved from stated monotone-feedback assumptions plus external Poincaré–Bendixson/zero-number results; the few self-citations are illustrative and non-load-bearing.

full rationale

The central claims (Theorems 2.1–2.3, 2.5, Corollary 2.6) are not obtained by fitting a parameter and calling it a prediction, nor by defining the conclusion into the hypotheses. The local continuation in Lemmas 3.4–3.5 is an implicit-function/center-manifold argument from the linearized DDE; the amplitude reparametrization in Corollary 4.2 replaces the local parameter b by the maximum of the solution only after proving ∂_b x_b(0)≠0 (Lemmas 3.5 and 4.4); the global embedding in Theorem 4.1 rests on the Jordan-curve/nesting arguments of Section 5, which invoke external results [13,14,15,16]. The g-function in Theorem 2.3 is constructed from the time/amplitude chart (Eq. 4.36), and the inequality ∂_1 g ∂_2 f<0 is derived, not assumed (Eqs. 4.38–4.39). The author's prior work [10,11] appears only in motivational examples (Eq. 1.8–1.10, Example 2.8), a Sobolev-semiflow remark (Remark 3.7), and a concluding remark on unstable dimension; none of these supports Theorem 2.5 or Corollary 2.6. I therefore find no step in which a prediction is equivalent to an input by construction. One correctness proviso, explicitly flagged: in the proof of Theorem 2.5, Eq. (6.11) perturbs f to \tilde f and asserts that 'our previous analysis holds' without proving persistence of the original branches B and \hat B through the perturbation or continuity of relation (6.13) as ε→0. That is an unproven perturbation step, not a circular reduction; the theorem's conclusion is not assumed by that sentence.

Assumptions & free parameters 0 free parameters · 9 assumptions · 0 invented entities

The central claim rests on established theorems (Poincaré-Bendixson for monotone delayed feedback, zero numbers, Floquet theory, center manifolds) plus the structural assumption r * d2 f != 0. No data are fitted. The only suspicious added premise is the perturbation of f in the proof of Theorem 2.5.

assumptions (9)
  • standard math Poincaré-Bendixson theorem for monotone delayed feedback systems: the projection P(phi) = (phi(0), phi(-1)) C^k-embeds every periodic orbit as a Jordan curve, with the same-delay nesting property (dichotomy (1.7)).
    Invoked in Section 1 (Eq 1.5-1.7) and used throughout, especially Lemmas 3.1, 5.1, and 5.2. The entire planar reduction presupposes this theorem.
  • standard math Zero number and Floquet theory for scalar DDEs: all zeros of the derivative of a periodic solution are simple, the critical Floquet multiplier has a two-dimensional generalized eigenspace, and the minimal period lies in one of the intervals J_n = (2/n, 2/(n-1)).
    Used in Lemma 3.1, Proposition 3.3, and Lemma 3.6 via [13, Theorem 5.1] and [15]. The local continuation in Section 3 depends on this spectral structure.
  • standard math Nesting property for monotone cyclic feedback systems ([16, Proposition 3.2]).
    Used in Lemma 5.2 for rational-period solutions via the M-vector reduction; this is an external result.
  • standard math C^k dependence of the DDE semiflow on initial data and parameters, and Floquet theory as in [5].
    Needed for the Poincaré map, implicit function theorem applications in Lemmas 3.4 and 3.5, and for hyperbolic and nonhyperbolic continuation.
  • standard math Center manifold theorem for fixed points of C^k maps on Banach and Sobolev spaces as in [7].
    Used in the nonhyperbolic continuation proof (Lemma 3.5). Remark 3.7 acknowledges the need to move to Sobolev space to obtain smooth cutoffs.
  • standard math Formal adjoint theory for linear DDEs (Hale 1977, Henry 1970), including the bilinear form (7.4) and spectral duality.
    Used in Section 7 to construct the projection P_Psi and prove P_Psi(rho_{p*}) is nonzero in Lemma 3.6.
  • standard math Arzelà-Ascoli theorem and convergence of normalized period-one solutions.
    Used in the proof of Theorem 2.5 to pass to limits of periodic solutions at the boundary of the amplitude range.
  • domain assumption The standing structural assumption f in C^k, k >= 2, with r * d2 f != 0 (monotone delayed feedback), and existence of the periodic branches considered.
    Stated in Section 1 (Eq 1.2). All theorems are conditional on this assumption.
  • ad hoc to paper The perturbed nonlinearity tilde f = f + epsilon * iota(u,v)(v - a) in Theorem 2.5 has periodic branches B and hat B analogous to the original equation.
    This is the questionable premise: the proof of Theorem 2.5 relies on a perturbation of f but does not verify that the original intersecting branches persist or that the previous analysis applies to the perturbed DDE.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Periodicity in delayed self-regulation is a predator-prey process." pith.science (2026). https://pith.science/paper/PX5C3MUW

@misc{pith2026250607660,
  author       = {Pith},
  title        = {Pith review of: Periodicity in delayed self-regulation is a predator-prey process},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PX5C3MUW}},
  note         = {Machine review of arXiv:2506.07660}
}
abstract

We unify two different periodicity mechanisms: delayed self-regulation and planar predator-prey feedback. We consider scalar delay differential equations $\dot x(t) = rf(x(t), x(t - 1))$ where $f$ is monotone in the delayed component. Due to a Poincar\'{e}--Bendixson theorem for monotone delayed feedback systems, the typical global dynamics present periodic orbits as the delay parameter $r$ increases. In this article, we show that, as we vary the delay, each connected component of periodic orbits is an annulus with global coordinates given by the time and the amplitude of the corresponding periodic solutions. On each annulus, the variables $x(t)$ and $x(t - 1)$ solve an integrable ordinary differential equation that satisfies a predator-prey feedback relation. Moreover, we completely characterize the set of periodic solutions of the delay differential equation in terms of two time maps generated by the underlying predator-prey system.

Figures

Figures reproduced from arXiv: 2506.07660 by the authors.

Figure 1
Figure 1. Left: Numerical approximation of the integral curves of the planar ODEs (2.10) for the Hutchinson DDE (1.1). The cyclicity component O consists of one annulus (0,∞) 2 \ {(1, 1)}. Right: The three branches are the graphs of the delay map r(a) (solid) and the time rescaling symmetries (2.11) given by (1 + p(a))r(a) (dashed), and (1 − p(a))r(a) (dotted). All periodic branches connect to a Hopf bifurcation at amplitude … view at source ↗
Figure 2
Figure 2. Left: Numerical approximation of the integral curves of three branches of periodic solutions of the DDE (1.2) with the QRT nonlinearity (2.23). The cyclicity set C consists of three connected components in blue, green, and orange separated by a homoclinic figure-eight. Right: The delay map r(a) for the three branches shows that the periodic solutions appear by Hopf bifurcation at amplitudes −1 and 0, and by a homocl… view at source ↗
Figure 3
Figure 3. Left: Crossing with two tangent intersection points (open dots) where the Jordan curve P γ∗ changes the connected components defined by P γ†. Right: Tangency, although the curves intersect (open dot), P γ∗ is fully contained in the closure of the interior component of P γ†. in particular, we have that (5.4) τ (t − p∗) = τ (t) − p† and τ (t − 1) = τ (t) − 1 − mp†, for some m ∈ Z. Differentiating x ∗ (t), we obtain x˙… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Left: Delay maps r˜(b) associated to periodic branches of the extended DDE (2.1) that appear by the rescaling symmetry (2.11). Right: If p˜(b)˜r(b) is not constant, then the height R(m) of the branches grows to infinity. Thus, it produces intersections of periodic orbi…
Figure 5
Figure 5. Figure 5: Possible relative configurations of the projections Pγ˜b if Pγ˜b1 and Pγ˜b2 have a tangency (open dot). Left: If Pγ˜b lies in the interior of Pγ˜b1 , then deform￾ing it into Pγ˜b2 yields a crossing. The same happens if Pγ˜b lies in the outside of Pγ˜b2 instead. Right: …
Figure 6
Figure 6. Figure 6: Triangles formed by the projections P γai and the nullcline f −1 (0). Left: If the projected orbits cross, then P γa∗ must cross one of them. Right: If P γa∗ crosses neither P γa1 nor P γa2 , then it leaves the triangle through a tangency at the top tip (open dot). Nex…
Figure 7
Figure 7. Figure 7: Plots of x¨ ∗ (t) (solid) and X(t) (dashed) over the interval (n1p∗,(n1 + 1)p∗). Left: If p∗ ∈ Jn with an odd n, then the distribution of extrema of X(t) is given by (7.47). Together with X(1) = X(p∗) = 0, we see that X(q + n1p∗) 6= X(q + 1), yielding a contradiction t…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages

  1. [14]

    Mallet-Paret and G

    J. Mallet-Paret and G. R. Sell. The Poincar´ e–Bendixso n theorem for mono- tone cyclic feedback systems with delay. J. Differ. Equ. , 125(2):441–489, 1996

    work page 1996

  2. [1]

    Chow and J

    S.-N. Chow and J. Mallet-Paret. The Fuller index and glob al Hopf bifurcation. J. Differ. Equ. , 29(1):66–85, 1978

    work page 1978

  3. [2]

    J. J. Duistermaat. Discrete Integrable Systems: QRT Maps and Elliptic Sur- faces. Springer, New York, 2010

    work page 2010

  4. [3]

    Fiedler, C

    B. Fiedler, C. Rocha, and M. Wolfrum. Heteroclinic orbit s between rotat- ing waves of semilinear parabolic equations on the circle. J. Differ. Equ. , 201(1):99–138, 2004

    work page 2004

  5. [4]

    J. K. Hale. Theory of Functional Differential Equations , volume 3 of Applied Mathematical Sciences. Springer-Verlag, New York, 1977

    work page 1977

  6. [5]

    J. K. Hale and S. M. Verduyn-Lunel. Introduction to Functional Differential Equations, volume 99 of Applied Mathematical Sciences . Springer-Verlag, New York, 1993

    work page 1993

  7. [6]

    D. Henry. The adjoint of a linear functional differential e quation and bound- ary value problems. J. Differ. Equ. , 9(1):55–66, 1970

    work page 1970

  8. [7]

    M. W. Hirsch, C. C. Pugh, and M. Shub. Invariant Manifolds , volume 583 of Lecture Notes in Mathematics . Springer-Verlag, Berlin and Heidelberg, 1977

    work page 1977

Show all 22 references
  1. [8]

    G. E. Hutchinson. Circular causal systems in ecology. Ann. N. Y. Acad. Sci. , 50(4):221–246, 1948

    1948

  2. [9]

    J. Lewis. Autoinhibition with transcriptional delay: A simple mechanism for the zebrafish somitogenesis oscillator. Curr. Biol. , 13(16):1398–1408, 2003

    2003

  3. [10]

    L´ opez-Nieto

    A. L´ opez-Nieto. Enharmonic Motion: Towards the Global Dynamics of Neg- ative Delayed Feedback. PhD thesis, Freie Univ. Berlin, 2023

    2023

  4. [11]

    L´ opez-Nieto

    A. L´ opez-Nieto. Global bifurcation of periodic solut ions in delay equations with symmetric monotone feedback, 2024. Submitted. 40

    2024

  5. [12]

    M. C. Mackey and L. Glass. Oscillation and chaos in physi ological control systems. Science, 197(4300):287–289, 1977

    1977

  6. [13]

    Mallet-Paret and R

    J. Mallet-Paret and R. D. Nussbaum. Tensor products, po sitive linear oper- ators, and delay-differential equations. J. Dyn. Differ. Equ. , 25(4):843–905, 2013

    2013

  7. [15]

    Mallet-Paret and G

    J. Mallet-Paret and G. R. Sell. Systems of differential de lay equations: Flo- quet multipliers and discrete Lyapunov functions. J. Differ. Equ. , 125(2):385– 440, 1996

    1996

  8. [16]

    Mallet-Paret and H

    J. Mallet-Paret and H. L. Smith. The Poincar´ e–Bendixs on theorem for mono- tone cyclic feedback systems. J. Dyn. Differ. Equ. , 2(4):367–421, 1990

    1990

  9. [17]

    Nishiguchi

    J. Nishiguchi. C 1-smooth dependence on initial conditions and delay: spaces of initial histories of Sobolev type, and differentiability o f translation in Lp. Electron. J. Qual. Theory Differ. Equ. , 2019(91):1–32, 2019

    2019

  10. [18]

    C. Rocha. Realization of period maps of planar hamilton ian systems. J. Dyn. Differ. Equ. , 19(3):571–591, 2007

    2007

  11. [19]

    M. J. Su´ arez and P. S. Schopf. A delayed action oscillat or for ENSO. J. Atmos. Sci. , 45(21):3283–3287, 1988

    1988

  12. [20]

    G. Vas. Configurations of periodic orbits for equations with delayed positive feedback. J. Differ. Equ. , 262(3):1850–1896, 2017

    2017

  13. [21]

    Yanchuk, S

    S. Yanchuk, S. Ruschel, J. Sieber, and M. Wolfrum. Tempo ral dissipative solitons in time-delay feedback systems. Phys. Rev. Lett. , 123(5):053901, 2019

    2019

  14. [22]

    Yoshioka-Kobayashi, M

    K. Yoshioka-Kobayashi, M. Matsumiya, Y. Niino, A. Isom ura, H. Kori, A. Miyawaki, and R. Kageyama. Coupling delay controls synch ronized oscil- lation in the segmentation clock. Nature, 580(7801):119–123, 2020. 41

    2020

Pith tools

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