# Hypercycle: A Principle of Natural Self-Organization Lecture Notes

Lessons: 1 · Slides: 11

## Reading Guide

- [Lesson 5 | Topic: Selection, Species, Will](#lecture-lesson-1)

---

<a id="lecture-lesson-1"></a>

## Lesson 5 | Topic: Selection, Species, Will

<a id="lecture-page-1"></a>

### Selection, Species, Will

*Page 1*

> **Page Objective:** Explains why open catalytic chains and branched catalytic networks cannot integrate information: under constant organization their fixed points migrate to corners or edges of the concentration simplex, giving partial competition or a 'once-for-ever' selection in which the last member of the chain finally dominates. Ring closure changes the qualitative behavior. Hypercyclic growth functions are defined through the exponent matrix P and classified into elementary (p = 2) and compound (p = n) types, with the exponent restriction Σ p_ik = p and cyclic symmetry. Fixed-point analysis locates a central fixed point at x̄0 = (c0/n)(1,…,1)ᵀ and n corner fixed points, and normal-mode analysis shows the center changing character with dimension: focus for n = 2, spiral sink for n = 3, center for n = 4, and saddles with spiral components for n ≥ 5. Lyapunov functions prove asymptotic stability of the central fixed point for n = 2, 3 and 4, while decomposition into globally invariant subsystems (flowing edge 2A, fixed-point edge 2B, face 3A) describes broken hypercycles. Numerical integration supplies evidence for a stable limit cycle for n ≥ 5, so that higher-dimensional hypercycles sustain cooperative, wave-like coherent growth of all members rather than a pure state.

**Teaching Points**

- Analyze catalytic chains under constant organization and derive the six (and generally 2n) fixed points of the three-membered and n-membered systems.
- State the rate-constant and total-concentration conditions k1 ≫ kj and c0 > Σ_{j=2}^{n} (k1 − kj)/kj for a stable stationary state, and explain why the chain cannot stabilize a joint function.
- Show how branched catalytic couplings behave and why their stable fixed point migrates toward one of the competing corners.
- Define hypercyclic growth functions via the exponent matrix P, the homogeneity condition Σ p_ik = p, and cyclic symmetry, and distinguish elementary from compound hypercycles.

Page Objective

This slide explains why open catalytic chains and branched catalytic networks cannot integrate information under constant organization, and why closing a catalytic ring changes the qualitative behavior. It then introduces hypercyclic growth functions, their exponent-matrix description, and the fixed-point and stability picture for different numbers of members. The overarching contrast is between partial competition or once-for-ever selection in open chains and cooperative, wave-like coherent growth in higher-dimensional hypercycles.

Teaching Points

1. Open catalytic chains under constant organization. For a three-membered chain, fixed-point analysis yields six fixed points;

for an n-membered chain, there are generally 2n fixed points. These lie on corners or edges of the concentration simplex. As parameters change, the stable fixed point migrates toward a corner or edge, so the system expresses partial competition rather than a joint function. The last member of the chain eventually dominates, giving a once-for-ever selection. A stable stationary state may appear only under restrictive rate and concentration conditions, such as k1 much larger than kj and c0 greater than the sum over j = 2 to n of (k1 minus kj) divided by kj. Even then, the chain cannot stabilize a joint function shared by all members.

2. Branched catalytic couplings. When the chain branches, the stable fixed point migrates toward one of the competing corners. The branch points do not create a balanced internal equilibrium;

instead they sharpen competition among alternatives. This helps explain why open networks have limited capacity to integrate information: their asymptotic states are partial, not cooperative.

3. Ring closure. Closing the catalytic chain into a cycle changes the qualitative behavior. The relevant object becomes the hypercycle, whose growth functions are defined through the exponent matrix P. Each entry p_ik specifies the catalytic influence of species k on species i. The homogeneity condition is that the sum over k of p_ik equals p. Cyclic symmetry is required. Hypercycles are classified as elementary when p equals 2 and compound when p equals n.

4. Fixed-point structure. For the symmetric hypercycle, fixed-point analysis locates a central fixed point at x0 = (c0/n)(1,..., 1)^T and n corner fixed points. The central point represents equal concentrations of all members;

the corner points represent pure states in which one member dominates.

5. Normal modes and dimension. The character of the central fixed point depends on n. For n = 2 it is a focus. For n = 3 it is a spiral sink. For n = 4 it is a center. For n greater than or equal to 5 it becomes a saddle with spiral components. This dimension dependence is central: low-dimensional hypercycles can settle to the central cooperative state, whereas higher-dimensional ones generically do not.

6. Stability results. Lyapunov functions prove asymptotic stability of the central fixed point for n = 2, 3, and 4. For higher dimensions, the central fixed point is not asymptotically stable in the same way. Decomposition into globally invariant subsystems helps organize the analysis of broken hypercycles: the flowing edge 2A, the fixed-point edge 2B, and the face 3A describe how trajectories can be trapped or channeled when cyclic symmetry is broken.

7. Higher-dimensional behavior. Numerical integration supplies evidence for a stable limit cycle for n greater than or equal to 5. Thus higher-dimensional hypercycles sustain cooperative, wave-like coherent growth of all members rather than settling to a pure state. The cycle is not a static balance;

it is an organized oscillation that preserves participation of all species.

Check for Understanding

Ask a student to explain in their own words why an open catalytic chain tends toward a corner of the concentration simplex, while a ring-closed hypercycle can preserve all members. Listen for mention of fixed-point migration, partial competition, ring closure, the central fixed point, and dimension-dependent stability. If the student only says that the hypercycle is more complex, questions them to contrast the asymptotic states: pure dominance versus coherent coexistence or limit-cycle behavior.

Transition

Once students can state why open chains cannot integrate information and why ring closure changes the stable outcome, move to the next page, where the focus can shift to examples, applications, or a summary of hypercycle dynamics.

<a id="lecture-page-2"></a>

### The Integration Problem in Molecular Selection

*Page 2*

> **Page Objective:** Learners should understand why molecular selection and evolution require special properties and why simple competitive systems cannot integrate information.

**Teaching Points**

- Open chains and branched networks can produce stable or migrating fixed points, but those outcomes are asymmetric dominance or corner selection, not reciprocal integration. Ask whether all members jointly contribute to function, or whether one component wins.
- For an n-membered hypercycle, remember the central fixed point at x̄<sub>0</sub> = (c<sub>0</sub>/n)(1,...,1)ᵀ and n corner fixed points. Normal-mode character changes with dimension: focus for n = 2, spiral sink for n = 3, center for n = 4, and saddles with spiral components for n ≥ 5.
- An integrated unit must remain competitive with its mutants. If it cannot, it may lose its specific information during further evolution. Use this as a filter when evaluating any proposed integration architecture.

Page Objective

Learners should understand why molecular selection and evolution require specific molecular properties, and why simple competitive systems cannot integrate information. They should be able to compare open catalytic chains, branched catalytic networks, and closed hypercycles as candidate integration architectures.

Teaching Sequence

1. Start from the molecular basis of selection.

2. State the integration problem and the ligation constraint.

3. Compare three candidate architectures.

4. Define hypercyclic growth functions.

5. Review fixed points and stability of hypercycles.

6. Apply the integration criteria to judge whether an architecture supports joint function rather than dominance.

Teaching Points

Selection and evolution are not global attributes of arbitrary matter. They occur only when molecules have specified properties such as replication, catalytic influence, and heritable variation. These properties make differential survival and reproduction possible. Without them, the integration problem does not arise.

How can independently competitive molecules be combined into a larger functional unit without ligating them into one large replicative unit?

Ligation is constrained by the threshold relation for v_max: a single long replicator must keep its replication rate above the error threshold, so fusion is not generally available. Any alternative architecture must satisfy four criteria: joint function, information retention, competitiveness against mutants, and stability against parasitic or dominant subcomponents.

Open catalytic chains, branched catalytic networks, and closed hypercycles.

Under constant organization, an n-membered chain has 2n fixed points. A stable stationary state requires strong asymmetry k_1 >> k_j (j = 2,...,n) and total concentration c_0 > sum_{j=2}^n (k_1 - k_j)/k_j. This yields asymmetric dominance, not reciprocal integration;

the chain cannot stabilize a joint function.

In branched coupling, the stable fixed point migrates toward one competing corner of the concentration simplex. The system tends toward dominance by one component rather than balanced joint function.

Cyclic catalytic coupling defines hypercyclic growth functions through the exponent matrix P. The homogeneity condition is sum p_ik = p, with cyclic symmetry. Elementary hypercycles have p = 2;

compound hypercycles have p = n. These are the remaining candidate for integrating competitive molecules while retaining selection.

Fixed points and stability of hypercycles: For an n-membered hypercycle under constant organization, fixed-point analysis locates a central fixed point at x̄_0 = (c_0/n)(1,...,1)^T and n corner fixed points. Normal-mode analysis shows the center changing character with dimension: focus for n = 2, spiral sink for n = 3, center for n = 4, and saddles with spiral components for n ≥ 5. Lyapunov functions prove asymptotic stability of the central fixed point for n = 2 and 3;

the four-membered case is a neutral center. Decomposition into globally invariant subsystems—flowing edge 2A, fixed-point edge 2B, face 3A—describes broken hypercycles. Numerical integration gives evidence for a stable limit cycle for n ≥ 5, so higher-dimensional hypercycles sustain cooperative, wave-like coherent growth of all members rather than a pure state.

An architecture succeeds only if the integrated unit remains competitive with its mutants; otherwise it cannot evolve further without losing its specific information. Joint function, not dominance, is the target.

**Key Terms**

- replication
- catalytic influence
- heritable variation
- v_max threshold
- constant organization
- open catalytic chain
- branched catalytic network
- closed hypercycle
- exponent matrix P
- homogeneity condition
- cyclic symmetry
- elementary hypercycle
- compound hypercycle
- central fixed point
- corner fixed point
- limit cycle
- broken hypercycle.

> **Visual Cues**
>
> - Use the four-panel concept map to contrast independent competitors, open catalytic chains, branched networks, and closed hypercycles. Highlight that the first lacks integration, the second and third produce asymmetric dominance or corner dominance, and only the cyclic hypercycle offers a route to retain selection while coupling members. The hypercycle panel links the exponent matrix P, homogeneity sum p_ik = p, and the elementary/compound distinction.

**Practice Guidance**

specified molecular properties such as replication, catalysis, and heritable variation;

evolution is not generic. Q2 Suggested answer: the v_max threshold limits how large a competitive replicator can be. Q3 Suggested answer: strong asymmetry k_1 >> k_j together with c_0 > sum_{j=2}^n (k_1 - k_j)/k_j. Q4 Suggested answer: the fixed point migrates toward one of the competing corners. Q5 Suggested answer: an exponent matrix P with homogeneity sum p_ik = p, cyclic symmetry, and elementary/compound types. If learners choose an incorrect option, direct them back to the distinction between dominance and joint function, or to the formal definition of hypercyclic growth.

**Class Activity**

Present a small prebiotic population of competing replicators and ask students to evaluate the three candidate architectures. Have them decide whether each architecture satisfies joint function, information retention, competitiveness with mutants, and stability against parasites. Ask them to justify their choice using fixed-point behavior rather than intuition.

**Check for Understanding**

Ask students to explain in their own words why open chains and branched networks fail to produce reciprocal integration;to state the central fixed point formula and the n corner fixed points;

and to describe how normal-mode character changes with n. Then ask why competitiveness with mutants is a necessary filter for any integrated unit.

> **Transition**
>
> With the integration criteria and hypercycle stability results established, the next step is to apply them to broken hypercycles and to determine when higher-dimensional hypercycles sustain cooperative growth rather than collapsing to a dominant component.

<a id="lecture-page-3"></a>

### Fixed Points of Open Catalytic Chains

*Page 3*

> **Page Objective:** Learners should understand how constant organization turns open catalytic chains into systems whose fixed points lie on the boundary of the concentration simplex.

> **Page Objective**
>
> Understand how constant organization fixes the total concentration in an open catalytic chain and forces all fixed points onto the boundary of the concentration simplex. Derive the six fixed points for a three-membered chain and the general 2n count for an n-membered open chain.

> **Teaching Sequence**
>
> 1. Begin with constant organization and the concentration simplex.
>
> 2. Explain why an open chain has no all-positive interior fixed point.
>
> 3. Count the boundary fixed points for n = 3 and for general n.
>
> 4. Work through the A → B → C example.
>
> 5. Use the simplex diagram to visualize corner, edge, and interior states.
>
> 6. Check understanding with the quiz and case study.

**Teaching Points**

- Under constant organization, the total concentration c0 is held fixed. Normalized concentrations are ξ_i = x_i / c0, satisfying ξ_i ≥ 0 and Σ ξ_i = 1. The allowed state space is the concentration simplex S_n. For n = 3, S_3 is a triangle
- corners represent pure single-species states, edges represent two-species mixtures with one species absent, and the interior represents all three species present. An open catalytic chain lacks a closing link from the last member back to the first (for example, C → A is missing). Because of this, the fixed-point conditions cannot be satisfied with every ξ_i > 0. Any fixed point must have at least one species absent, so fixed points lie on the boundary of S_n. The all-positive interior state is not a fixed point
- it is transient. For n = 3, there are six boundary fixed points: three pure corner states (A, B, C) and three two-species edge equilibria (AB, BC, CA). For a general n-membered open chain, the structural count is 2n boundary fixed points. The 2n result is an enumeration of boundary fixed points only. It does not determine which fixed points are stable or how the system approaches them. Stability and dynamical behavior require separate analysis. In the A → B → C example, solving the normalized fixed-point conditions gives exactly the six boundary fixed points. Under the chain dynamics, trajectories migrate toward the boundary, and the last member of the chain can eventually dominate, leading to a once-for-ever selection.

**Key Terms**

- total concentration c0 is held fixed. ξ_i = x_i / c0. the set of ξ with ξ_i ≥ 0 and Σ ξ_i = 1. one species at ξ = 1 and all others at zero. exactly two species present and at least one absent. a chain with no closing link from the last member back to the first.

> **Visual Cues**
>
> - The diagram shows the concentration simplex S_3 for an open chain A → B → C. Red circles at the corners A, B, and C mark corner fixed points. Blue circles on the edges AB, BC, and CA mark edge fixed points. A dashed circle in the interior is labeled as transient (not fixed). Arrows indicate trajectories moving from the interior toward the boundary. The caption notes that an n-membered open chain has 2n fixed points.

**Practice Guidance**

Correct answer is 'The total concentration c0'. Feedback: Constant organization fixes c0, so normalized concentrations satisfy Σ ξ_i = 1. It is a global constraint, not a constraint on a single species or rate constant.

Correct answer is 'ξ_i ≥ 0 and Σ ξ_i = 1'. These conditions define the concentration simplex S_n.

Correct answer is 'Because the chain has no closing catalytic link returning the last member to the first'. Without cyclic closure, the fixed-point conditions cannot be satisfied with all ξ_i > 0.

Correct answer is 'Three pure corner states and three two-species edge equilibria'. The corners are A, B, C; the edge equilibria lie on AB, BC, and CA.

Correct answer is '2n fixed points'. The general structural count for an n-membered open chain is 2n boundary fixed points.

Case Study Guidance:

Set the normalized rate equations to zero (dξ_A/dt = dξ_B/dt = dξ_C/dt = 0) along with ξ_A + ξ_B + ξ_C = 1. Because the chain is open (no C → A), there is no balanced cyclic flow that would allow all three species to be positive simultaneously. Therefore any fixed point must have at least one ξ_i = 0, placing it on the boundary.

The six boundary fixed points are: corner A (1,0,0);corner B (0,1,0);corner C (0,0,1);edge AB with ξ_C = 0 and ξ_A, ξ_B > 0;edge BC with ξ_A = 0 and ξ_B, ξ_C > 0;

edge CA with ξ_B = 0 and ξ_C, ξ_A > 0. The central interior state with all three species positive is not a fixed point.

**Check for Understanding**

Ask learners to explain in their own words why an open catalytic chain cannot have an all-positive fixed point.

Ask them to sketch the simplex for n = 3 and mark the six fixed points.

Ask them to state the general 2n count and clarify that it does not imply stability.

> **Transition**
>
> Once learners can identify and count the boundary fixed points of an open chain, proceed to the stability analysis of these fixed points and the dynamics that select particular boundary states.

<a id="lecture-page-4"></a>

### Stability Conditions and the Failure of Joint Function

*Page 4*

> **Page Objective:** Let learners test and consolidate the current concept.

**Teaching Points**

- Memorize: k<sub>1</sub> ≫ kⱼ for all j > 1, and c<sub>0</sub> > Σⱼ₌<sub>2</sub>ⁿ (k<sub>1</sub> − kⱼ)/kⱼ. Both are necessary for a stable positive stationary state.
- A stable stationary state in an open chain is not a hypercycle. Integration requires cyclic feedback so that all components affect each other and the system acts as a joint functional unit.
- In this model, selective advantage means a higher rate constant. A molecule with larger k dominates the stationary state, but that dominance does not produce cooperation or a joint function.

Page Objective

Establish the two conditions for a stable positive stationary state in an open catalytic chain and clarify why such stability does not amount to cooperative integration. Learners should leave able to state the rate-constant condition and the total-concentration threshold, compute the threshold in simple cases, and explain why an open chain remains first-member-dominated rather than becoming a hypercyclic unit.

Teaching Sequence

Begin with the central distinction: stability of a stationary state is not the same as integration. In an open catalytic chain under constant organization, a positive stationary state can be stable while the chain still lacks cyclic feedback. That missing feedback is the key architectural limitation.

Teaching Points

1. Rate-constant condition. The first member must dominate all downstream members: k₁ ≫ kⱼ for every j > 1. If downstream rate constants are comparable to or larger than k₁, the required positive stationary state cannot be maintained in the same way.

2. Total-concentration condition. The total concentration must exceed the threshold: c₀ > Σⱼ₌₂ⁿ (k₁ − kⱼ)/kⱼ. Below this threshold, the implied downstream concentrations become nonpositive, so no positive stationary state exists in the concentration simplex. Stronger dominance by the first member—larger k₁/kⱼ ratios—raises the required c₀. This is counterintuitive at first: making the first member more dominant makes the concentration requirement harder, not easier.

3. Worked example n = 2. For k₁ = 100 and k₂ = 1, the threshold is c₀,min = (100 − 1)/1 = 99. If c₀ = 100, then x₁ = 99 and x₂ = 1. If c₀ < 99, no positive stationary state exists. This gives a concrete numerical anchor for the general inequality.

4. Stability does not imply integration. Even when both conditions hold, the state remains an open, first-member-dominated chain. Downstream species do not return to influence the first member. In this sense, the chain lacks cyclic feedback, and perturbations do not engage the whole chain as a joint functional unit. Stability is therefore not equivalent to integration into a hypercyclic collective.

5. Selective advantage at the molecular level. Here selective advantage means a higher rate constant. A molecule with larger k outcompetes others and shifts the stationary state toward its own dominance. That competitive advantage does not by itself produce cooperation or a joint function.

Visual Cues

Use the plot of c₀,min versus k₁/k₂. The feasible region shows where a positive stable stationary state is possible;

the lower region shows where no positive stationary state exists. Emphasize the caption: the feasible region does not imply cooperative integration. The threshold line and example point c₀ = 100 with c₀,min = 99 make the inequality visible.

Practice Cues

necessary conditions. Correct choice: k₁ ≫ kⱼ and c₀ > Σ (k₁ − kⱼ)/kⱼ. Feedback: the first member must dominate, and total concentration must exceed the threshold. If incorrect, review both inequalities.

threshold for k₁ = 100 and k₂ = 1. Correct: 99. Feedback: c₀,min = (100 − 1)/1 = 99, so c₀ must be greater than 99.

why stability fails to represent cooperative integration. Correct: because the chain lacks cyclic feedback among its components. Feedback: no return loops from downstream species to the first member, so perturbations do not engage the whole chain as a unit.

effect of increasing k₁/k₂ on c₀. Correct: it increases c₀. Feedback: the threshold formula shows larger k₁ relative to kⱼ raises the required total concentration.

selective advantage meaning. Correct: a higher rate constant that lets the molecule dominate the stationary state. Feedback: advantage means domination in this model, not cooperation or joint function.

Three-Membered Open Chain

Set n = 3, k₁ = 100, k₂ = 1, k₃ = 1, and c₀ = 150. Compute c₀,min = (100 − 1)/1 + (100 − 1)/1 = 99 + 99 = 198. Since c₀ = 150 is less than 198, the condition c₀ > c₀,min is not satisfied, so no positive stationary state exists;

downstream concentrations would become nonpositive.

Now raise c₀ to 250. Then a stable positive stationary state exists. But the chain remains open and first-member-dominated: k₁ ≫ k₂, k₃. There is no cyclic feedback from downstream members back to the first member. Therefore, perturbations do not engage the whole chain as a cooperative unit. Stability of the stationary state is not equivalent to integration into a hypercyclic collective, so the chain cannot stabilize a joint function.

Check for Understanding

Ask learners to state the two conditions from memory, compute c₀,min for a two-member chain, and explain in one sentence why an open chain with a stable stationary state is still not a hypercycle. Listen for the ideas of first-member dominance, total-concentration threshold, and absence of cyclic feedback.

Transition

Once learners can distinguish stability from integration, move on to settings where cyclic feedback is present. That contrast prepares the next concept: how closed or hypercyclic organization allows all components to affect one another and the system to act as a joint functional unit.

<a id="lecture-page-5"></a>

### Branched Catalytic Networks and Once-for-Ever Selection

*Page 5*

> **Page Objective:** Learners should understand how branched catalytic couplings behave and why their stable fixed point migrates toward one competing corner.

> **Page Objective**
>
> By the end of this slide, learners should explain how a branched catalytic network behaves under constant organization and why its stable fixed point migrates to one competing corner rather than an interior cooperative state.

> **Teaching Sequence**
>
> a branched network has one catalyst or catalytic unit feeding multiple competing replicators, but it has no ring closure and no mutual dependence. This is not a hypercycle, so it cannot integrate multiple functions into a joint catalytic organization.
>
> 2. Present the selection equation: ẋ_i = x_i(A_i − Φ), with Φ = Σ_j x_j A_j. Explain that x_i is branch concentration, A_i is selective value or productivity, and Φ is the concentration-weighted mean productivity. A branch grows only when A_i exceeds Φ. This comparison is the key to predicting motion on the simplex.
>
> 3. Move to fixed points and corner selection: the concentration simplex S_n has pure states at its corners. For a branched, non-cyclic network, interior coexistence is generally unstable or saddle-like when branches compete. Linearization near a corner gives a Jacobian. If every eigenvalue has negative real part, the corner attracts;
>
> if at least one has positive real part, it repels. The branch with the higher A_i drives x_i toward 1 while others decline toward 0. Fluctuations or small rate differences can tip the system, but once an advantage appears, the fixed point migrates to that branch's corner. Losing competitors go extinct.
>
> once-for-ever selection. This is a winner-take-all commitment, not cooperative integration. The stable fixed point migrates to one competing corner and remains there. Because branched coupling lacks closure and mutual dependence, it fails to sustain joint function.
>
> 5. Work the source example, the two-member branched and parasite case: ċ_A = k_A c_A² − (c_A/c_0)(k_A c_A² + k c_A x), with c_0 = c_A + x, and ẋ = k c_A x − (x/c_0)(k_A c_A² + k c_A x). This leads to ẋ > 0 for c_0 k/k_A. The system has two fixed points at the corners of S_2, illustrating corner selection and the absence of stable interior coexistence. Use this example to show how total concentration and rate constants control whether the parasite grows or declines.
>
> 6. Use the practice questions to check whether learners can connect the formula, corner stability, and once-for-ever selection.

**Teaching Points**

- one catalyst supports multiple competitors, but the coupling is open. There is no closure and no mutual dependence among all members. Therefore it is not a hypercycle and cannot integrate multiple functions into a joint catalytic organization. Constant organization fixes total concentration. The branch dynamics follow ẋ_i = x_i(A_i − Φ), with Φ = Σ_j x_j A_j. Here Φ is the weighted mean of all selective values, so a branch grows only if its A_i exceeds Φ. at a pure-state corner, linearize and inspect the Jacobian eigenvalues. Negative real parts mean the corner attracts
- positive real parts mean it repels. The winning branch is the one whose corner is stable. Branching is not a hypercycle: without ring closure and mutual dependence, the system cannot integrate functions. Expect partial competition or once-for-ever selection, not cooperative coexistence.

> **Visual Cues**
>
> - The reaction diagram shows Catalyst A feeding Branch 1 (B1) and Branch 2 (B2), with rates k1 and k2. The caption emphasizes one catalyst feeding two competing replicators, no ring closure, and no mutual dependence. The simplex S2 shows corners x1 = 1 and x2 = 1, an interior coexistence point marked unstable or saddle, and a moving dot labeled stable fixed point toward the winning branch. Reinforce that higher A_i wins and losing branches go extinct.

**Practice Guidance**

1. Under constant organization, where does the stable fixed point migrate as one branch gains a selective advantage?

it migrates to a corner of the simplex. Rationale: open branching lacks a stable interior fixed point, so selection drives the system to a pure state. If learners choose a limit cycle or interior point, redirect them to the corner-selection logic.

2. In ẋ_i = x_i(A_i − Φ), what does Φ represent?

the average productivity Σ_j x_j A_j. It is the concentration-weighted mean, so a branch grows only if A_i exceeds Φ. Misconception: Φ is not the largest Jacobian eigenvalue, total concentration, or a single rate constant.

3. For the two-member branched and parasite case, which condition gives ẋ > 0? Correct: c_0 k/k_A, ẋ < 0 and the parasite declines. Have learners check the inequality directly.

4. In linearization near a corner, which Jacobian eigenvalue pattern makes that corner stable?

every eigenvalue has a negative real part. Negative real parts mean small perturbations decay. If any positive real part appears, the corner is unstable.

5. Why do branched catalytic couplings fail to integrate multiple functions into a cooperative hypercycle?

they lack closure and mutual dependence among all members. Without these, the outcome is partial competition or once-for-ever selection, not cooperative coexistence.

**Class Activity**

Use the branched parasite competition scenario. A constant-organization reactor contains catalyst A supporting a cooperative branch B and a self-reproductive parasite P. Dynamics follow ẋ_i = x_i(A_i − Φ). Ask learners to determine the sign of ẋ for c_0 < k/k_A and explain parasite invasion;

then ask them to predict long-term composition if the parasite has a higher selective value than the cooperative branch. Expected reasoning: the higher-selective-value branch has a locally stable corner;

linearization gives negative real parts there, while the competing corner is unstable or saddle-like. The system commits to the dominant branch and the loser goes extinct. This is once-for-ever selection, not hypercyclic cooperation, because there is no closure or mutual dependence to sustain joint function.

**Check for Understanding**

Ask learners to state in one sentence why branched networks cannot settle at an interior cooperative state. A strong answer mentions open coupling, absence of ring closure and mutual dependence, unstable or saddle-like interior coexistence, and migration of the stable fixed point to a pure-state corner. If time allows, have them compare this with a hypercycle in one additional sentence.

> **Transition**
>
> Once learners can connect the selection equation, corner stability, and once-for-ever selection, move to the next topic on ring closure and hypercyclic organization, where cyclic closure and mutual dependence change the qualitative behavior.

**Source Basis**

Source coverage (pages 44–54).

<a id="lecture-page-6"></a>

### Ring Closure and Hypercyclic Growth Functions

*Page 6*

> **Page Objective:** Learners should understand how closing an open chain into a ring defines a hypercycle and changes the qualitative behavior of the system.

**Teaching Points**

- In an n-membered hypercycle the catalyst of species i is species i − 1, read mod n. Write the index of the <em>predecessor</em>, not the successor, and remember that the first species is catalyzed by the last — that wrap-around is exactly what 'ring closure' means.
- Row i of P lists the catalysts that appear in the growth function of species i; the diagonal 1s are self-replication and the off-diagonal 1s are the preceding-member (i − 1) couplings. A cyclic 1 in row 1, column n is the ring-closure signature — if that entry is missing, you are looking at an open chain.
- An open chain selects a single winner at a corner of the simplex, whereas a closed hypercycle couples all members and sustains cooperative, wave-like growth. Also keep the arrow meaning straight: it denotes catalysis, not the stoichiometric conversion of one species into another.

Page Objective

Explain how closing an open catalytic chain into a ring creates a hypercycle and why this changes the qualitative dynamics from a one-way cascade to cooperative, information-integrating growth.

Teaching Sequence

1. Start with the open chain A → B → C: catalytic influence flows forward only. Under constant organization, no interior attracting state exists;

fixed points migrate to the corners of the concentration simplex, so the terminal member tends to dominate. This is once-for-ever selection, not a joint function.

2. Add the missing feedback edge C → A. This is ring closure. The topology becomes A → B → C → A, a hypercycle: a cyclic catalytic arrangement whose members are cooperatively coupled and can integrate information rather than merely pass it along.

3. Formalize with the exponent matrix P. Row i lists which concentrations appear in the growth function of species i. Diagonal entries P_ii = 1 represent self-replication. Off-diagonal entries P_i,i−1 = 1 represent catalysis by the preceding member. The cyclic entry P_1,n = 1 is the ring-closure signature: the last species catalyzes the first. If that entry is absent, the network is an open chain, not a hypercycle.

4. Read indices modulo n. In an n-membered hypercycle, species i is catalyzed by species i − 1 (mod n). Thus the first species is catalyzed by the last; that wrap-around is exactly what ring closure means.

5. Growth terms are nonlinear and second order: x_i x_{i−1}. Under constant organization, a dilution/flux term Φ balances the total concentration, so relative concentrations evolve on the simplex. For n = 3: dx_A/dt = x_A(k_A + k_AC x_C) − x_A Φ, dx_B/dt = x_B(k_B + k_BA x_A) − x_B Φ, dx_C/dt = x_C(k_C + k_CB x_B) − x_C Φ, with indices read mod 3. The C → A term completes the cycle.

6. Hypercyclic growth functions satisfy the homogeneity condition Σ_k p_ik = p and cyclic symmetry. They are classified by that exponent into elementary (p = 2) and compound (p = n) types. The qualitative point is that the closed loop can sustain coherent, wave-like growth of all members rather than driving a single corner of the simplex to dominance.

Teaching Points

Keep the distinction clear between an open chain and a closed ring. An open chain is a linear cascade with no feedback;

a closed ring is a hypercycle with cooperative coupling. The arrow denotes catalysis, not stoichiometric conversion of one species into another. Also keep the predecessor direction straight: species i depends on species i − 1 modulo n, so the first species depends on the last.

Visual Cues

Use the before-and-after diagram: left side shows open chain A → B → C with a dashed missing C → A edge and the note “linear cascade — no feedback”;

right side shows the closed ring A → B → C → A with arrows for catalysis. Point out the 3 × 3 matrix P for n = 3, especially P_1,3 = 1 highlighted as the ring-closure entry. Reinforce that the diagonal 1s mark self-replication.

Practice Cues

Quiz 1 asks for the definition of a hypercycle. Correct Suggested answer: a cyclic catalytic arrangement in which each species is coupled to its neighbours, the loop closes, and cooperative information integration becomes possible. Feedback: look for “cyclic” and “cooperative coupling”;

branching or member count is not enough.

Quiz 2 asks which edge completes A → B → C into a ring. Correct Suggested answer: C → A. Feedback: ring closure adds the missing feedback edge from the terminal member back to the initial member.

Quiz 3 asks which species catalyzes species i when indices are read modulo n. Correct Suggested answer: species i − 1. Feedback: follow the loop direction; i − 1 wraps around at the closure edge.

Quiz 4 asks for the form of the cross-catalytic growth term. Correct Suggested answer: a second-order product k_i x_i x_{i−1}. Feedback: predecessor-dependent coupling multiplies species i by its predecessor, not a linear or constant term.

Quiz 5 asks what P_1,n = 1 signifies. Correct Suggested answer: the last species catalyzes the first, which is the ring-closure edge. Feedback: row 1 contains a dependency on species n; remove it and the matrix describes an open chain.

Practice Guidance for the Case Study

In the four-membered ring, before closure R1 lacks predecessor input from the end of the chain. After adding R4 → R1, each growth function takes the form x_i(k_i + k_{i,i−1}x_{i−1}) − x_i Φ with indices mod 4. The network becomes a hypercycle because catalytic influence returns to its starting point, coupling all four units cooperatively. The exponent matrix for (R1, R2, R3, R4) is P = [[1,0,0,1],[1,1,0,0],[0,1,1,0],[0,0,1,1]]. The cyclic entry P_1,4 = 1 is ring closure, representing R4 catalyzing R1. Since each growth term is second order, p = 2, so the hypercycle is elementary;

a compound hypercycle would instead have p = n = 4.

Check for Understanding

Ask learners to state in their own words what single edge is missing in an open chain and where that edge appears in P. Then ask them to explain why closure converts a winner-take-all cascade into a cooperative cycle. Listen for “cyclic coupling,” “predecessor i − 1 mod n,” and “P_1,n = 1.”

Transition

Once learners can read the closure edge from both the diagram and the exponent matrix, move on to fixed-point and stability analysis of hypercycles, where dimension determines whether the central fixed point behaves as a focus, spiral sink, center, or saddle with spiral components.

<a id="lecture-page-7"></a>

### Exponent Matrix and Hypercycle Classification

*Page 7*

> **Page Objective:** Learners should understand how the exponent matrix defines hypercyclic growth functions and distinguishes elementary from compound hypercycles.

**Teaching Points**

- Always verify the homogeneity condition ∑<sub>k</sub> p<sub>ik</sub> = p before classifying a hypercycle. If row sums differ, the growth functions are not homogeneous of a single degree.
- Cyclic symmetry means P is circulant: each row is a cyclic shift of the previous row. Test this by shifting indices simultaneously, not by transposing or permuting rows alone.
- The number of species n does not determine elementary vs compound. Look at p and the number of nonzero entries per row. A 4-species cycle with p = 2 is elementary; with p = 4 it can be compound.

Page Objective

By the end of this page, learners should be able to read the exponent matrix P of a hypercycle, check its two structural conditions, and use it to distinguish elementary from compound hypercycles.

Teaching Points

The catalytic architecture of a hypercyclic growth function is encoded in a single matrix P = (p_ik). For n species, P is an n × n matrix of nonnegative integers. The entry p_ik gives the exponent, or molecularity, with which species k contributes to the growth function of species i. A zero entry means no catalytic influence;

a positive entry gives the order of that influence.

Two conditions make P a hypercycle matrix rather than a general catalytic network. First, the homogeneity condition requires each row to sum to the same integer p: sum_k p_ik = p for every i. All species therefore share one total growth degree. If row sums differ, the growth functions are not homogeneous of a single degree. Second, cyclic symmetry requires invariance under a simultaneous cyclic relabeling of indices: i → i+1 and k → k+1 mod n, so p_ik = p_{i+1,k+1} mod n. P is therefore circulant, or block-circulant in compound cases. The catalytic pattern is identical for every member up to rotation.

Elementary hypercycles are the simplest closed cyclic case, usually with p = 2. Each growth function depends on the species itself and its immediate cyclic predecessor. For n = 4, p_ii = 1 and p_{i,i-1} = 1 with indices mod 4, so every row sums to 2. The matrix is sparse: exactly two 1s per row.

Compound hypercycles have a higher-order cyclic organization, for example p = n or multiple nonzero cyclic contributions per row. For n = 4 and p = 4, each row can contain four unit cyclic contributions, giving quartic homogeneous growth functions while retaining cyclic symmetry. The matrix is denser, but remains circulant.

elementary versus compound is read from the value of p and the sparsity or pattern of P, not merely from the number of species. A four-membered cycle can be elementary (p = 2) or compound (p = 4), depending on the catalytic exponents.

Visual Cues

Use the matrix diagram to contrast the two types. The elementary panel for n = 4 and p = 2 shows a sparse matrix with two 1s per row and row sums equal to 2. The compound panel for n = 4 and p = 4 shows a dense matrix with all entries equal to 1 and row sums equal to 4. In both panels, point out the annotated row sums and the circular index key that marks the cyclic relabeling i → i+1 mod n. The diagram makes the invariance condition visible: shifting both row and column indices leaves the matrix pattern unchanged.

Practice Cues

Q1. What does p_ik represent?

Guide learners to the exponent or molecularity of species k in the growth function of species i. It is not a rate constant and not a probability. If they choose otherwise, send them back to the definition of P as an exponent matrix.

Q2. When is a hypercyclic growth function homogeneous of degree p?

The correct condition is that every row of P sums to p, so all species share the same total molecularity. Feedback should reinforce the row-sum condition, not diagonal or column properties.

Q3. What does cyclic symmetry mean? The correct answer is that P is unchanged by a cyclic shift of the indices, making P circulant. Clarify that this is a simultaneous shift of row and column indices, not transposition or diagonality.

Q4. For n = 4 and p = 2, what is the row pattern? Each row has two unit entries: self and predecessor, so it sums to 2. If learners think every entry is 1 or that only one entry is 2, remind them that p = 2 is a row sum, not a single entry.

Q5. How do you distinguish elementary from compound hypercycles?

Use the degree p and the sparsity or pattern of P, not the number of species alone. A 4-species cycle can be elementary at p = 2 or compound at p = 4. Feedback should point to the classification criterion.

Class Activity

Designing a Five-Membered Hypercycle. Present the scenario: model a prebiotic hypercycle with n = 5 species, where P is 5 × 5, nonnegative integer, and must satisfy homogeneity and cyclic symmetry. Ask learners to compare an elementary design and a compound design.

Answer guidance for the elementary design: for p = 2, set p_ii = 1 and p_{i,i-1} = 1 with indices mod 5, and all other entries 0. Each row has two 1s, so sum_k p_ik = 2. Cyclic symmetry holds because shifting i → i+1 and k → k+1 maps the self-entry to the next row's self-entry and the predecessor entry to the next row's predecessor entry, so p_ik = p_{i+1,k+1} mod 5.

Answer guidance for the compound design: for p = 5, set all entries p_ik = 1. Each row sums to 5, so the growth functions are homogeneous of degree 5. The matrix becomes completely dense while remaining circulant. This contrasts with the sparse elementary matrix that has only two 1s per row.

Check for Understanding

Ask a learner to explain, in their own words, how the exponent matrix defines a hypercyclic growth function and how to tell elementary from compound. A strong answer will mention p_ik as an exponent, the row-sum condition sum_k p_ik = p, cyclic symmetry under simultaneous index shift, and classification by p and sparsity pattern. If the learner omits row sums or confuses cyclic symmetry with transposition, revisit the corresponding section and the matrix diagram.

Transition

Before moving on, confirm that learners can read an exponent matrix, verify homogeneity, test cyclic symmetry, and classify the hypercycle. The next step can apply these checks to new matrices or use them to interpret higher-dimensional hypercycle behavior.

<a id="lecture-page-8"></a>

### Central and Corner Fixed Points of the Hypercycle

*Page 8*

> **Page Objective:** Learners should understand the fixed-point structure of a hypercycle, including the central fixed point and n corner fixed points.

**Teaching Points**

- If you forget the formula, rebuild it geometrically: the central fixed point x̄<sub>0</sub> = (c<sub>0</sub>/n)(1, …, 1)ᵀ is the barycentre of the n pure-state vertices, so every coordinate is the arithmetic mean c<sub>0</sub>/n. For n = 3 with c<sub>0</sub> = 0.9 this immediately gives (0.3, 0.3, 0.3); check that the n equal shares always sum back to c<sub>0</sub>.
- A common misconception is that a corner point is 'not an equilibrium' because it lies on the boundary of the simplex. It is a genuine fixed point: one species holds all of c<sub>0</sub> while the others are zero, and every dxᵢ/dt vanishes. Remember the geometric picture — the central point is interior, the n corners are vertices, and both satisfy the steady-state condition.
- For every n-membered hypercycle, the fixed-point set on the simplex is exactly 1 central point plus n corner points. After listing them, always run two checks: each coordinate set must sum to c<sub>0</sub>, and each point must satisfy dxᵢ/dt = 0 for all species. Consider separately that the corners exist regardless of n, whereas whether the central point actually attracts trajectories depends on n.

> **Page Objective**
>
> Understand the fixed-point structure of an n-membered hypercycle under constant organization: one central fixed point with equal concentrations and n corner fixed points corresponding to pure states. Distinguish existence from attraction: corners always exist;
>
> whether the central point attracts trajectories depends on n and will be examined later.

> **Teaching Sequence**
>
> 1. Start from constant organization: total concentration is held at c0, so states lie on the simplex x1+...+xn=c0. A point is a fixed point only when dxi/dt=0 for every species simultaneously; the sum constraint alone only marks admissible states.
>
> 2. Derive the central fixed point: xbar0=(c0/n)(1,...,1)^T. Every species carries c0/n, the coordinates sum to c0, and the point lies in the interior. It is permutation-symmetric, representing full cooperation with no species favoured.
>
> 3. Introduce the corner fixed points: there are n of them, one for each species. At a corner, that species holds all of c0 and all others are zero. These are pure states at the vertices of the simplex, representing dominance or competitive exclusion by a single member.
>
> the simplex is a triangle. The central fixed point is (c0/3,c0/3,c0/3);
>
> the corners are (c0,0,0), (0,c0,0), (0,0,c0). The fixed-point set is 1+n points in total. If c0=0.9, the central point is (0.3,0.3,0.3) and the corners are (0.9,0,0), (0,0.9,0), (0,0,0.9). Always check both conditions: coordinates sum to c0 and dxi/dt=0 for all species.
>
> the central fixed point is the barycentre (arithmetic average) of the n vertices, which explains the c0/n coordinates. Corners are genuine steady states even though they lie on the boundary;
>
> the boundary is part of the state space. The structural contrast is interior versus boundary and symmetry versus symmetry-breaking, not existence versus non-existence.

> **Visual Cues**
>
> - Use the n=3 simplex triangle. Mark the three vertices as species 1: (c0,0,0), species 2: (0,c0,0), species 3: (0,0,c0), and mark the central dot as xbar0=(c0/3,c0/3,c0/3). The dashed arrows in the diagram can be read as possible trajectories or flow near the corners, helping students see that corners are boundary equilibria while the centre is an interior equilibrium. Emphasize that the centre is the average of the vertices. Practice Cues and Answer Guidance: At the central fixed point each species carries c0/n. Guidance: equal shares must sum to c0. Correct feedback: xbar0=(c0/n)(1,...,1)^T. Incorrect questions: revisit the simplex constraint and symmetry
> - divide c0 by n. There are exactly n corner fixed points, one per species. Guidance: count the simplex vertices. Correct feedback: each species can be sole survivor. Corners lie at the vertices, where one species equals c0 and all others are zero. Guidance: a pure state means a single species present at c0. Correct feedback: extreme points of the simplex. A fixed point requires dxi/dt=0 for every species simultaneously. Guidance: the sum constraint only defines admissible states
> - corners also satisfy the steady-state condition. Correct feedback: the steady-state condition is stronger than admissibility. For n=3 and c0=3, the corner is (3,0,0). Guidance: one coordinate c0, others zero
> - (1,1,1) is central, (1.5,1.5,0) is on an edge. Correct feedback: only a vertex is a pure state. Case Study / Practice Guidance: For n=3, c0=0.9, the fixed points are central (0.3,0.3,0.3) and corners (0.9,0,0), (0,0.9,0), (0,0,0.9). Classify central as interior equal-concentration symmetric
> - corners as boundary pure-state symmetry-breaking. Verify each sums to 0.9 and all rates vanish. If a student claims corners are not genuine equilibria because they are on the boundary, correct this: for (0.9,0,0), absent species have zero growth and the surviving species is stationary, so dxi/dt=0 for all i. Full credit requires the enumeration, classification, checks, and the structural contrast.

**Check for Understanding**

Ask students to (a) write the central fixed point formula and explain why each coordinate is c0/n, (b) state how many corner fixed points exist and where they lie, (c) give the two verification checks for any candidate fixed point. A quick oral questions: Why is a simplex vertex a legitimate equilibrium even though it is on the boundary?

Listen for zero growth of absent species and stationarity of the survivor.

> **Transition**
>
> Now that the fixed-point set is established, the next step is to test attraction and stability. The corners exist for every n, but whether the central cooperative point actually attracts trajectories depends on n and on the normal modes;
>
> this leads into the stability analysis and the behaviour for n=2, 3, 4 and higher n.

**Source Basis**

pages 44-54.

<a id="lecture-page-9"></a>

### Normal Modes and Dimensional Character

*Page 9*

> **Page Objective:** Learners should understand how the character of the central fixed point changes with dimension n in normal-mode analysis.

**Teaching Points**

- Remember: the real part of eigenvalues determines stability (negative = stable, positive = unstable, zero = neutral), and the imaginary part indicates oscillatory/spiral behavior. For the hypercycle, n=2,3,4 have negative or zero real parts, while n≥5 has mixed signs.
- The same equilibrium point can change its stability type solely due to dimension. Always compute the Jacobian at the specific n you are studying; do not assume that stability carries over from lower dimensions.
- For n=4, the center implies neutral cycling; perturbing the system leads to a different closed orbit. For n≥5, the saddle spiral instability often gives rise to a stable limit cycle, sustaining cooperative growth of all members.

Page Objective

Understand how the central fixed point of the hypercycle changes dynamical character as the cycle dimension n increases. The key tool is normal-mode analysis: linearize around the central fixed point, form the Jacobian, and read the local behavior from its eigenvalues.

Teaching Points

At the central fixed point x̄_0 = (c_0/n)(1,...,1)^T, the linearized Jacobian determines the normal modes. The eigenvalues classify the local dynamics: the real part controls stability, while the imaginary part indicates oscillation or spiral behavior. Negative real part means stable, positive means unstable, and zero means neutral. The same equilibrium point can change stability type solely because the eigenvalues depend on n, so always compute the Jacobian for the specific dimension under study.

eigenvalues are real and negative, so the fixed point is a focus, a stable node-like attraction.

eigenvalues are complex with negative real part, so the fixed point is a spiral sink; nearby trajectories spiral inward.

eigenvalues are purely imaginary, so the fixed point is a center; it is neutrally stable with closed orbits. Perturbing the system gives a different closed orbit.

the fixed point becomes a saddle with spiral components. Some directions attract while others repel, because the eigenvalues have mixed signs. This prevents global asymptotic stability of the central fixed point and often gives rise to a stable limit cycle, sustaining cooperative growth of all members.

Lyapunov functions prove asymptotic stability of the central fixed point for n = 2, 3, and 4. For n ≥ 5, no such global Lyapunov function exists for the central fixed point because it is a saddle;

instead, numerical integration reveals a stable limit cycle.

Visual Cues

Use the phase portraits to compare the four cases: n = 2 shows inward arrows toward a focus;n = 3 shows an inward spiral to a sink;n = 4 shows closed dashed orbits around a center;

n ≥ 5 shows mixed attracting and repelling directions around a saddle with spiral components. Ask learners to connect each picture to the sign pattern of the eigenvalues.

Practice Guidance

1. Stability type for n = 3? Suggested answer: spiral sink. Correct because the eigenvalues have negative real part and nonzero imaginary part. If a learner misses it, return to the n = 3 bullet and the inward-spiral picture.

2. Dimension at which the fixed point becomes a center? Suggested answer: n = 4. Correct because the real part of the eigenvalues is zero and the eigenvalues are purely imaginary. If missed, review neutral stability and closed orbits.

3. What mathematical object has eigenvalues that determine normal-mode character?

the Jacobian matrix of the linearized system. Correct because normal modes come from linearization around the fixed point. If missed, contrast this with the exponent matrix P or the Lyapunov function.

4. Nature of the central fixed point for n ≥ 5? Suggested answer: saddle with spiral components. Correct because the eigenvalues include both positive and negative real parts. If missed, contrast this with a stable focus or spiral sink.

5. Why does the character change with n?Suggested answer: the eigenvalues of the Jacobian depend on n, so their real and imaginary parts change with dimension, altering stability. The equilibrium remains the central point x̄_0 = (c_0/n)(1,...,1)^T;

the direct explanation is not simply that the growth rate changes, that the number of species increases, or that the equilibrium position shifts.

Class Activity

Present the five-member hypercycle case. For n = 5, the central fixed point is a saddle with spiral components: unstable in some directions and stable in others. This differs from n = 4, where the fixed point is a center with neutral closed orbits. The shift from neutral to saddle shows that higher dimensions introduce instability, which can lead to limit cycle behavior. Ask learners for which dimensions a Lyapunov function could prove asymptotic stability. Suggested answer: n = 2, 3, and 4. For n ≥ 5, the saddle character blocks such a global Lyapunov function for the central fixed point, and the system may instead exhibit a stable limit cycle.

Check for Understanding

Have learners state the rule for real and imaginary parts of eigenvalues, then match n = 2, n = 3, n = 4, and n ≥ 5 to focus, spiral sink, center, and saddle spiral. Ask them to explain why the same equilibrium point can change type without changing its central role: the Jacobian eigenvalues depend on n. A quick sketch or verbal description of n = 4 versus n = 5 is a useful check.

Transition

Once learners can explain the eigenvalue-to-stability mapping and the dimensional shift, move on to applications or examples that use this classification, especially the contrast between neutral cycling at n = 4 and limit-cycle behavior for n ≥ 5.

<a id="lecture-page-10"></a>

### Lyapunov Stability and Broken Hypercycles

*Page 10*

> **Page Objective:** Learners should understand how Lyapunov functions prove stability for n=2,3,4 and how invariant subsystems describe broken hypercycles.

**Teaching Points**

- A Lyapunov function must satisfy V(x̂) = 0 and V(x) > 0 for x ≠ x̂. The stability proof then needs only one sign check: V must fall along trajectories (dV/dt < 0). You never have to solve the kinetic equations explicitly.
- The Lyapunov proof of asymptotic stability of the central fixed point holds for n = 2, 3 and 4. Do not extend it to n ≥ 5, where the cooperative behaviour takes a different, wave-like form.
- When a link breaks, look for the globally invariant subsystems instead of a closed cycle: the 2A edge, the 2B edge and the 3A face in the three-membered case. Analysing these reduced pieces reconstructs the behaviour of the broken hypercycle.

Page Objective

Understand how a Lyapunov function proves asymptotic stability of the central fixed point for small hypercycles, and how a broken hypercycle decomposes into globally invariant subsystems that can be analysed separately.

Teaching Points

1. Lyapunov criterion. A function V(x) is a Lyapunov function for the central fixed point x̂ when V(x̂) = 0 and V(x) > 0 for every x ≠ x̂. If V also decreases along trajectories, dV/dt < 0, then nearby orbits are drawn toward x̂, so x̂ is asymptotically stable. The argument is qualitative: you check the sign of the change, not a closed-form solution of the kinetic equations.

2. Stability for n = 2, 3 and 4. For these dimensions an explicit Lyapunov function can be constructed and it proves that the central fixed point x̄0 = (c0/n)(1,..., 1)^T is asymptotically stable. This complements the normal-mode picture, where the equilibrium changes character with dimension: a focus at n = 2, a spiral sink at n = 3, and a center at n = 4. The Lyapunov result upgrades the local description into a global stability statement. Do not extend this proof to n ≥ 5;

there the cooperative behaviour takes a different, wave-like form.

3. Broken hypercycles. If one catalytic link is removed, the network no longer closes into a single self-reinforcing cycle. Instead it partitions into globally invariant subsystems: sets of species whose dynamics never leak into the rest of the system. Because orbits cannot leave them, each subsystem can be studied on its own. In the three-membered case, the key pieces are the 2A edge, a flowing edge that transmits catalytic influence;

the 2B edge, a fixed-point edge that settles to rest;

and the 3A face, a two-dimensional subsystem spanning the remaining members. Analysing these lower-dimensional objects reconstructs the stability and selection behaviour of the broken cycle piece by piece.

Visual Cues

Use the side-by-side diagram to make the contrast concrete. On the left, the intact three-membered hypercycle is a closed cycle with species 1, 2 and 3, and the central fixed point x̂ is marked asymptotically stable. On the right, one link has been removed, the closed cycle is broken, and the remaining architecture is labelled by the invariant subsystems 2A, 2B and 3A. Point out that the same species can have two qualitatively different fates: one cooperating unit with a certified central fixed point, or a family of independent invariant subsystems.

Practice Guidance

The correct choice is V(x̂) = 0 and V(x) > 0 for every x ≠ x̂, with V decreasing along trajectories. Emphasize that a Lyapunov function must vanish at the equilibrium, be strictly positive elsewhere, and fall as the system evolves;

it must not rise or stay constant.

The correct choice is n = 2, 3 and 4. The constructive Lyapunov proof covers the small hypercycles; higher dimensions behave differently.

Breaking one catalytic link does not simply collapse the system onto a corner. It partitions the network into globally invariant subsystems that can be analysed separately.

Invariant subsystems are useful because orbits cannot leave them. That trapping of the flow allows stability and selection to be studied in a reduced, lower-dimensional setting; it does not eliminate fixed points or freeze the dynamics.

In the three-membered example with one link removed, the residual catalytic architecture is described by the 2A edge, the 2B edge and the 3A face, not by a single closed cycle.

For the intact three-membered hypercycle, justify asymptotic stability of the central fixed point by constructing V(x) with V(x̂) = 0, V(x) > 0 for x ≠ x̂, and dV/dt < 0 along trajectories. For n = 3 such a function exists, so x̄0 = (c0/3)(1, 1, 1)^T is asymptotically stable and nearby states move toward equal concentrations of all three species. After the 3 to 1 link is removed, the single closed cycle is destroyed. The network decomposes into globally invariant subsystems, namely the 2A edge, the 2B edge and the 3A face. Since orbits cannot leave these subsystems, each can be examined independently at lower dimension, recovering the stability and selection behaviour of the broken cycle piece by piece instead of as one cooperating whole.

Check for Understanding

Ask students to state the two conditions on V and the one sign condition on its time derivative. Then ask them to name the dimension cutoff for the Lyapunov proof and why it matters. Finally, have them identify the invariant subsystems of a broken three-membered hypercycle and explain why invariance makes each piece analytically tractable.

Transition

Once students can distinguish the certified central fixed point of an intact small hypercycle from the invariant decomposition of a broken one, they are ready to apply the same reasoning to higher-dimensional or selection-focused examples, where the cooperative behaviour changes form and the stability argument must be adapted.

Source Basis

Course pages 44–54.

<a id="lecture-page-11"></a>

### Limit Cycles and Coherent Growth

*Page 11*

> **Page Objective:** Learners should understand how numerical integration reveals stable limit cycles for n≥5 and why higher-dimensional hypercycles sustain cooperative coherent growth.

**Teaching Points**

- For n≥5, numerical integration provides evidence for a stable limit cycle.
- The limit cycle corresponds to wave-like coherent growth of all members.
- Higher-dimensional hypercycles do not settle to a pure state or a single corner.

Page Objective

Explain why hypercycles with n ≥ 5 no longer settle at the central cooperative fixed point and instead persist as a stable limit cycle, producing wave-like coherent growth of all members. Learners should be able to contrast this behavior with lower-dimensional hypercycles and with broken hypercycles.

Teaching Points

The central cooperative fixed point x̄0 = (c0/n)(1, …, 1)ᵀ is no longer the stable attractor for n ≥ 5. Normal-mode analysis traces how the fixed point changes character with dimension: it is a focus for n = 2, a spiral sink for n = 3, a center for n = 4, and a saddle with spiral components for n ≥ 5. Lyapunov arguments prove asymptotic stability of the central fixed point for n = 2, 3, and 4. For n ≥ 5, numerical integration provides evidence for a stable limit cycle in the concentration simplex. In this cycle, all members persist with phase-shifted oscillations: replication is temporally ordered, so the whole cycle is sustained as an integrated unit. No single member wins by competitive exclusion, and the system does not settle at a pure state or a single corner. This contrasts with broken hypercycles, whose dynamics can be confined to invariant edge or face subsystems, producing partial competition or once-for-ever selection.

Visual Cues

Use the left-hand time-series plot for n = 5 to show five concentrations, x1(t) through x5(t), oscillating with phase shifts. After transients die out, the oscillations remain sustained rather than decaying to a stationary value. Use the right-hand simplex diagram to distinguish the central fixed point from the closed orbit: corner fixed points are marked, the central point x̄0 is marked, and the closed orbit wraps around it. Reinforce that for n ≤ 4 the trajectory approaches x̄0, while for n ≥ 5 it winds onto the closed orbit.

Practice Guidance

The practice questions ask learners to explain their reasoning in complete sentences and cite one key piece of evidence or example. A strong answer should connect n ≥ 5 to numerical evidence for a limit cycle, mention phase-shifted coherent growth, or cite normal-mode analysis showing that the central fixed point is no longer attracting. A weak answer only states a conclusion without explaining the basis. If a learner struggles, have them mark the key phrase in the question, identify whether it concerns n ≥ 5, coherence, or the contrast with broken hypercycles, and then support the answer with one specific detail from the page.

Check for Understanding

What changes between n = 4 and n = 5?Why is the limit cycle described as coherent growth?How does this differ from a broken hypercycle?Expected responses: the central fixed point loses asymptotic stability at n = 5;

the limit cycle keeps all members present through phase-shifted oscillations;broken hypercycles can collapse into partial competition or once-for-ever selection.

Transition

After confirming that learners can distinguish the fixed point from the limit cycle, move to the limitations of hypercyclic organization and the question of why functional coupling may eventually be replaced by more complete structural integration of genes.

Source Basis

pages 44–54.