Module 7: Major Evolutionary Transitions

John Maynard Smith and Eörs Szathmáry's 1995 synthesis identified eight major transitions in evolution: unique, highly improbable events where entities that previously could replicate independently came together to form larger units that could only replicate jointly. Each transition involves novel mechanisms of information transmission and levels-of-selection change. This module covers eukaryogenesis (the most consequential transition), multicellularity, the Cambrian Explosion, and game-theoretic models of cooperation emergence.

7.1 The Eight Major Transitions

Maynard Smith & Szathmáry proposed that biological complexity increased through a small number of unique transitions, each meeting three criteria: (1) entities capable of independent replication before the transition now replicate only as part of a larger unit; (2) novel information transmission mechanisms evolve; (3) division of labour emerges. The eight transitions:

Replicating molecules → populations of molecules in compartments
Independent replicators → chromosomes
RNA as gene and enzyme → DNA genes + protein enzymes
Prokaryotes → eukaryotes (endosymbiosis)
Asexual clones → sexual populations
Single-celled protists → multicellular organisms (animals, plants, fungi)
Solitary individuals → eusocial colonies (ants, bees, termites, naked mole rats)
Primate societies → human societies with language

Szathmáry & Maynard Smith emphasise that each transition posed a levels-of-selection problem: why would selfish lower-level units cooperate to form a higher-level unit, given the lower unit's temptation to defect? Each transition required mechanisms (kinship, compartmentalisation, enforcement) to align the interests of the lower and higher levels.

7.2 Eukaryogenesis

The origin of the eukaryotic cell, roughly 2.1-1.6 Gya, was arguably the most singular event in evolution after the origin of life itself. Lynn Margulis's endosymbiotic theory (1970) proposes that the ancestral eukaryotic cell arose when a primitive host archaeon engulfed (or was invaded by) an alphaproteobacterium, which persisted as the ancestor of mitochondria. The chloroplast emerged later from a cyanobacterial endosymbiont in the ancestor of Archaeplastida.

Evidence for endosymbiosis

Mitochondrial (and chloroplast) genomes are circular and bacterial-like
Mitochondrial ribosomes are 70S (bacterial) not 80S (eukaryotic)
Mitochondrial lipid composition (cardiolipin) is bacterial
Mitochondrial genes cluster with alphaproteobacteria phylogenetically
Mitochondrial division uses FtsZ-like machinery related to bacterial cell division

Asgard archaea

The identity of the host has become dramatically clearer with the discovery of Asgard archaea. Spang et al. (2015) reported “Lokiarchaeota” from Arctic sediments near the Loki's Castle hydrothermal system, whose genome contains genes for actin, profilin, GTPases, and ESCRT machinery previously thought eukaryote-specific. Zaremba-Niedzwiedzka et al. (2017) added Thor-, Odin-, and Heimdall-archaeota. In 2020, Imachi et al. cultured the first Asgard archaeon (MK-D1): it forms long, tendril-like cellular extensions, reminiscent of a host-symbiont embrace.

The hydrogen hypothesis

Martin & Müller (1998) proposed the hydrogen hypothesis: the original symbiosis was metabolic, not phagocytic. A hydrogen-dependent methanogenic archaeon lived syntrophically with a facultative alphaproteobacterium that supplied H₂ as a byproduct of fermentation. Genetic transfer eventually made the bacterium obligately endosymbiotic, and H₂ -> O₂-based respiration as the environment oxygenated. This model elegantly unifies the timing (after the GOE) with the metabolic dependency.

7.3 Origins of Multicellularity

Multicellularity has evolved independently at least 25 times, producing the three major lineages of complex multicellularity (animals, plants, fungi) plus simpler forms in brown algae, red algae, slime moulds, and at least a dozen bacterial lineages including cyanobacteria, myxobacteria, and actinomycetes.

Snowflake yeast experiments

Ratcliff et al. (2012) evolved multicellularity in real time in the laboratory. Saccharomyces cerevisiae subjected to gravity-based selection (larger cells sink faster in daily centrifugation) evolved snowflake-like clusters in 60 days. Daughter cells failed to separate after division, forming mother-daughter chains. In further generations, selection favoured larger and more differentiated snowflakes with specialised apoptotic cells at junction points allowing controlled fragmentation.

The experiment captures key features of early multicellularity: group-level selection, genetic bottlenecks at the single-celled bottleneck (propagules), and the emergence of cell-cell adhesion “glue” as the key invention.

The bottleneck principle

Why do almost all multicellular organisms start from a single-celled stage (egg, spore, propagule)? The bottleneck ensures genetic uniformity within the organism: all cells share the same genotype, eliminating within-body selection and aligning the interests of individual cells with the organism as a whole (Buss 1987, Grosberg & Strathmann 2007). Without the bottleneck, cancer-like selfish cell lineages can outcompete cooperating cells.

7.4 The Cambrian Explosion

Between 541 and 485 Mya, most major animal phyla appeared in the fossil record in a burst of evolutionary creativity compressed into ~25 Myr. Preceded by the Ediacaran fauna (635-541 Mya) of soft-bodied organisms and the trace-fossil-defined “Cambrian substrate revolution”, the Cambrian explosion produced:

Bilateria with left-right symmetry, centralised nervous systems, through-guts
First hard parts: mineralised exoskeletons, shells, teeth
First eyes (both compound and camera-type)
All modern animal phyla except Bryozoa (evolved in Ordovician)

Proposed causes

Oxygen rise: second oxygenation enabled energetic body plans (high aerobic metabolism)
Hox gene innovation: duplication of Hox clusters enabled modular body plan development
Ecological arms race: predation (Anomalocaris, ~1 m apex predator) drove shells, armour, eyes
Calcium biomineralisation: changes in ocean chemistry permitted calcium carbonate and phosphate skeletons
Developmental genetic toolkit: pre-existing in the Ediacaran, finally unleashed by ecological opportunity

Diversification math

The diversity dynamics follow a birth-death process:

\[ \frac{dS}{dt} = (\lambda - \mu) S \quad \Rightarrow \quad S(t) = S_0 \, e^{(\lambda - \mu) t} \]

where \(\lambda\) is speciation rate and \(\mu\) is extinction rate. Cambrian estimates: \(\lambda \approx 0.02\)/Myr, \(\mu \approx 0.005\)/Myr, giving a doubling time of ~46 Myr and producing hundreds of genera from tens within the explosion window. Modern exponential fits over the Phanerozoic (539 Mya-present) show long-term \(r \approx 0.003\)/Myr interrupted by the five major mass extinctions.

7.5 Cooperation & Evolutionary Game Theory

Evolutionary game theory provides a framework for understanding levels-of-selection puzzles. The hawk-dove game captures a conflict between two strategies in a population:

Payoff to	vs Hawk	vs Dove
Hawk	(V-C)/2	V
Dove	0	V/2

The evolutionarily stable strategy (ESS, Maynard Smith & Price 1973) is a mixed strategy with a fraction of Hawks:

\[ p^\star = \frac{V}{C} \quad \text{(when } V < C\text{)} \]

When the value of winning is less than the cost of losing, neither pure hawks nor pure doves are stable: the population settles at \(V/C\) Hawks. This provides an evolutionary route to cooperation without appealing to group selection: individually selfish strategies produce group-level coexistence.

Replicator dynamics

The replicator equation captures the dynamics:

\[ \dot{x}_i = x_i \left( f_i(x) - \bar f(x) \right) \]

where \(f_i\) is the fitness of type \(i\) and \(\bar f\) is the mean fitness. Strategies below the mean decline; above-mean strategies grow. This equation applies equally to genes competing within a genome, cells competing within an organism, and organisms competing within a population.

7.6 The Eighth Transition: Language

Symbolic language is the defining feature of the human lineage. Unlike animal signalling systems (which are semantically fixed), human language is:

Recursive: unlimited generation of meanings from a finite vocabulary via syntactic nesting
Displaced: can refer to absent, past, future, or hypothetical entities
Cultural: inheritance through social learning, producing cumulative culture — a second non-genetic information channel

Culture is a genuine new information-transmission channel on top of DNA, similar in scope to the appearance of DNA after RNA. It enables cumulative technology, science, and large-scale cooperation among genetically unrelated individuals — the principal driver of human ecological dominance.

7.7 The Evolution of Sex

Sexual reproduction imposes a twofold cost — “the cost of males” — yet is nearly universal among eukaryotes. Why? The Muller's ratchet argument: in asexual lineages, deleterious mutations accumulate irreversibly, fitness declines over generations. Sex via recombination regenerates mutation-free genotypes from partially-loaded parents:

\[ \bar W_\text{asex}(t) \sim \bar W_0 \exp\!\left(-\frac{U t}{N_0}\right) \]

where \(U\) is the deleterious mutation rate per genome per generation and \(N_0\) is the effective population size. Sex short-circuits this decline. The Red Queen hypothesis (Van Valen 1973) provides a second argument: sexual recombination generates novel genotypes that stay ahead of co-evolving parasites.

Python: Snowflake Yeast Multicellularity & Cambrian Diversification

Panel 1 simulates a Ratcliff-style selection for larger cell clusters under three adhesion probabilities; high adhesion coupled with size-dependent survival drives exponential growth in cluster size over 40 generations. Panel 2 solves the birth-death diversification model across the Phanerozoic with rate parameters capturing the Cambrian explosion, post-Cambrian slowdown, and the five major mass extinctions.

Python

script.py105 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# Snowflake-yeast style multicellularity evolution
# Population of cells; mutation can make cells adhere after division
# Selection pressure: e.g. predation favours larger clusters

np.random.seed(23)

fig, axes = plt.subplots(1, 2, figsize=(16, 7), facecolor='#03120a')
fig.subplots_adjust(wspace=0.35)

# ---- Panel 1: Cluster size evolution ----
ax1 = axes[0]
ax1.set_facecolor('#061a0d')

def simulate(adhesion_prob, n_gen=40, n_clusters=100, pred_k=2.0):
    # Each cluster has some number of cells; sinking speed ~ size^(2/3) (Stokes-like)
    clusters = np.ones(n_clusters)
    history = []
    for g in range(n_gen):
        # Growth: each cluster doubles in cell count
        clusters = clusters * 2
        # Fragmentation depending on adhesion
        new_clusters = []
        for c in clusters:
            # Fragment into pieces of mean size  ~ 1/(1-adh)
            pieces = max(1, np.random.poisson((1 - adhesion_prob) * c))
            for _ in range(pieces):
                new_clusters.append(max(1, c / pieces + np.random.randn() * 0.1))
        new_clusters = np.array(new_clusters)
        # Selection: clusters that sink fast (large) survive predation
        settling = new_clusters ** (2/3)
        probs = settling / settling.sum()
        idx = np.random.choice(len(new_clusters), size=n_clusters, p=probs)
        clusters = new_clusters[idx]
        history.append(clusters.copy())
    return history

for adh, col, lab in [(0.05, '#60a5fa', 'Low adhesion'),
                      (0.35, '#fbbf24', 'Medium adhesion'),
                      (0.75, '#34d399', 'High adhesion')]:
    hist = simulate(adhesion_prob=adh)
    means = [np.mean(h) for h in hist]
    maxes = [np.max(h) for h in hist]
    ax1.plot(np.arange(len(means)), means, color=col, lw=2.5, label=f'{lab} mean')
    ax1.plot(np.arange(len(maxes)), maxes, color=col, lw=1.5, ls='--', label=f'{lab} max')

ax1.set_yscale('log')
ax1.set_xlabel('Generation', color='#6ee7b7', fontsize=11)
ax1.set_ylabel('Cluster size (cells)', color='#6ee7b7', fontsize=11)
ax1.set_title('Snowflake Yeast Style Multicellularity\n(adhesion vs fragmentation under size selection)',
              color='#a7f3d0', fontsize=12, fontweight='bold')
ax1.legend(fontsize=8, labelcolor='#a7f3d0', facecolor='#061a0d', edgecolor='#10b981')
ax1.tick_params(colors='#6ee7b7')
ax1.grid(True, which='both', alpha=0.2, color='#10b981')
for sp in ax1.spines.values(): sp.set_edgecolor('#10b981')

# ---- Panel 2: Cambrian diversification (birth-death model) ----
ax2 = axes[1]
ax2.set_facecolor('#061a0d')

# dS/dt = (lambda - mu) * S
# Cambrian: high lambda, low mu; Ordovician: moderate; Mass extinctions: burst mu
t = np.linspace(540, 0, 1000)   # Mya
lam = 0.018 - 0.005 * np.tanh((t - 460) / 20)
mu  = 0.011 + 0.003 * np.exp(-((t - 252)**2) / 20) + 0.002 * np.exp(-((t - 66)**2) / 10)
r = lam - mu

S = np.zeros_like(t)
S[0] = 50
dt = t[0] - t[1]
for i in range(1, len(t)):
    S[i] = S[i-1] + r[i] * S[i-1] * dt

ax2.plot(t, S, color='#34d399', lw=2.8)
ax2.fill_between(t, 0, S, alpha=0.25, color='#34d399')

# Annotate mass extinctions
extinctions = [(252, 'End-Permian\n(-95%)'), (201, 'End-Triassic'), (66, 'K-Pg')]
for tex, lab in extinctions:
    idx = np.argmin(np.abs(t - tex))
    ax2.axvline(tex, color='#f87171', ls=':', lw=1.2, alpha=0.75)
    ax2.text(tex + 8, S[idx] * 0.3, lab, color='#fca5a5', fontsize=8)

ax2.axvspan(541, 485, alpha=0.12, color='#f472b6', label='Cambrian Explosion (541-485 Mya)')
ax2.invert_xaxis()
ax2.set_xlabel('Time (Mya)', color='#6ee7b7', fontsize=11)
ax2.set_ylabel('Number of genera (log-exponential model)', color='#6ee7b7', fontsize=11)
ax2.set_title("Cambrian Explosion & Phanerozoic Diversification\n(birth-death model)",
              color='#a7f3d0', fontsize=12, fontweight='bold')
ax2.legend(fontsize=9, labelcolor='#a7f3d0', facecolor='#061a0d', edgecolor='#10b981')
ax2.tick_params(colors='#6ee7b7')
ax2.grid(True, alpha=0.2, color='#10b981')
ax2.set_yscale('log')
for sp in ax2.spines.values(): sp.set_edgecolor('#10b981')

fig.suptitle('Multicellularity & Phanerozoic Diversification',
             color='#a7f3d0', fontsize=14, fontweight='bold')
plt.savefig('output.png', bbox_inches='tight', facecolor=fig.get_facecolor(), dpi=120)
plt.close()
print('Transitions simulation complete.')

Click Run to execute the Python code

Code will be executed with Python 3 on the server

Python: Hawk-Dove ESS & Cooperation Emergence

Panel 1 integrates the replicator equation for the hawk-dove game from several initial fractions of hawks, converging to the ESS fraction \(V/C\) regardless of initial condition. Panel 2 plots the ESS as a function of the payoff ratio, revealing the transition at \(V = C\) where the population switches from a mixed equilibrium to pure Hawks.

Python

script.py84 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# Hawk-Dove game: emergence of cooperation + evolutionary stability

np.random.seed(31)

# Payoffs: V (value), C (cost of injury)
V, C = 10, 40
A = np.array([
    [(V - C) / 2, V],          # Hawk vs Hawk, Hawk vs Dove
    [0,            V / 2]      # Dove vs Hawk, Dove vs Dove
])

def replicator(x0, T=300, dt=0.05):
    x = np.array(x0, dtype=float)
    hist = [x.copy()]
    for _ in range(T):
        fit = A @ x
        avg = x @ fit
        x = x + dt * x * (fit - avg)
        x = np.clip(x, 0, 1)
        x = x / x.sum()
        hist.append(x.copy())
    return np.array(hist)

fig, axes = plt.subplots(1, 2, figsize=(16, 7), facecolor='#03120a')
fig.subplots_adjust(wspace=0.35)

# ---- Panel 1: Replicator trajectories ----
ax1 = axes[0]
ax1.set_facecolor('#061a0d')

for x0_hawk in [0.05, 0.15, 0.35, 0.55, 0.85]:
    traj = replicator([x0_hawk, 1 - x0_hawk])
    ax1.plot(traj[:, 0], color=plt.cm.viridis(x0_hawk), lw=2.0,
             label=f'x_hawk0 = {x0_hawk:.2f}')

# ESS line
ess = V / C
ax1.axhline(ess, color='#fbbf24', ls='--', lw=2, label=f'ESS x_hawk = V/C = {ess:.2f}')

ax1.set_xlabel('Generation', color='#6ee7b7', fontsize=11)
ax1.set_ylabel('Fraction Hawks', color='#6ee7b7', fontsize=11)
ax1.set_title('Hawk-Dove Replicator Dynamics\nEvolutionarily Stable Strategy',
              color='#a7f3d0', fontsize=12, fontweight='bold')
ax1.legend(fontsize=9, labelcolor='#a7f3d0', facecolor='#061a0d', edgecolor='#10b981')
ax1.tick_params(colors='#6ee7b7')
ax1.grid(True, alpha=0.2, color='#10b981')
for sp in ax1.spines.values(): sp.set_edgecolor('#10b981')

# ---- Panel 2: ESS as function of V/C ----
ax2 = axes[1]
ax2.set_facecolor('#061a0d')

Vs = np.linspace(1, 50, 100)
C_fixed = 40
ess_curve = np.clip(Vs / C_fixed, 0, 1)

ax2.plot(Vs / C_fixed, ess_curve, color='#34d399', lw=3)
ax2.fill_between(Vs / C_fixed, 0, ess_curve, alpha=0.25, color='#34d399', label='Hawks favoured')
ax2.fill_between(Vs / C_fixed, ess_curve, 1, alpha=0.25, color='#60a5fa', label='Doves favoured')

ax2.plot([1, 1], [0, 1], color='#f87171', ls='--', lw=1.5, label='V = C threshold')

ax2.set_xlabel('V / C (value / injury cost)', color='#6ee7b7', fontsize=11)
ax2.set_ylabel('ESS fraction of Hawks', color='#6ee7b7', fontsize=11)
ax2.set_title('ESS vs Payoff Ratio\nWhen does cooperation emerge?',
              color='#a7f3d0', fontsize=12, fontweight='bold')
ax2.legend(fontsize=9, labelcolor='#a7f3d0', facecolor='#061a0d', edgecolor='#10b981')
ax2.tick_params(colors='#6ee7b7')
ax2.grid(True, alpha=0.2, color='#10b981')
ax2.set_xlim(0, 1.3)
ax2.set_ylim(0, 1.05)
for sp in ax2.spines.values(): sp.set_edgecolor('#10b981')

fig.suptitle('Cooperation & Evolutionary Game Theory',
             color='#a7f3d0', fontsize=14, fontweight='bold')
plt.savefig('output.png', bbox_inches='tight', facecolor=fig.get_facecolor(), dpi=120)
plt.close()
print('Hawk-Dove simulation complete.')

Click Run to execute the Python code

Code will be executed with Python 3 on the server

References

Maynard Smith, J. & Szathmáry, E. (1995). The Major Transitions in Evolution. W.H. Freeman.
Margulis, L. (1970). Origin of Eukaryotic Cells. Yale University Press.
Martin, W. & Müller, M. (1998). The hydrogen hypothesis for the first eukaryote. Nature, 392, 37-41.
Spang, A. et al. (2015). Complex archaea that bridge the gap between prokaryotes and eukaryotes. Nature, 521, 173-179.
Imachi, H. et al. (2020). Isolation of an archaeon at the prokaryote-eukaryote interface. Nature, 577, 519-525.
Ratcliff, W.C. et al. (2012). Experimental evolution of multicellularity. PNAS, 109, 1595-1600.
Grosberg, R.K. & Strathmann, R.R. (2007). The evolution of multicellularity: a minor major transition? Annual Review of Ecology, 38, 621-654.
Erwin, D.H. et al. (2011). The Cambrian conundrum: early divergence and later ecological success in the early history of animals. Science, 334, 1091-1097.
Maynard Smith, J. & Price, G.R. (1973). The logic of animal conflict. Nature, 246, 15-18.
Nowak, M.A. (2006). Five rules for the evolution of cooperation. Science, 314, 1560-1563.
Buss, L.W. (1987). The Evolution of Individuality. Princeton University Press.
Lane, N. (2015). The Vital Question. W.W. Norton.

←Previous: Great Oxygenation Next: Astrobiology Frontiers→