Macroevolution

Intuition Beginner

Microevolution is change in allele frequencies within a single population -- a small-scale, short-term process. Macroevolution is evolution at or above the species level -- the grand patterns of biodiversity over millions of years. Macroevolution includes the origin of new species (speciation), the diversification of lineages (adaptive radiation), episodes of catastrophic species loss (mass extinctions), and the long-term evolutionary trends that shape the tree of life.

Speciation is the process by which one species splits into two or more species that are reproductively isolated from each other (they can no longer interbreed to produce fertile offspring). The most common mechanism is allopatric speciation: a population is physically divided by a geographic barrier (a mountain range, a river, an ocean), the two populations evolve independently in their different environments, and over time they accumulate enough genetic differences that they can no longer interbreed even if the barrier is removed.

Other modes of speciation include sympatric speciation (new species arise within the same geographic area, without physical separation -- often through polyploidy in plants or ecological specialization), parapatric speciation (speciation across a gradient of environmental conditions), and peripatric speciation (a small peripheral population becomes isolated and evolves rapidly due to founder effects and genetic drift).

Adaptive radiation occurs when a single ancestral species rapidly diversifies into many descendant species, each adapted to a different ecological niche. Classic examples include Darwin's finches in the Galapagos (13 species from a common ancestor, each with a beak shape specialized for a different food source), Hawaiian honeycreepers, and the radiation of mammals after the extinction of the dinosaurs 66 million years ago.

The history of life has been punctuated by mass extinctions -- events in which a large proportion of Earth's species are wiped out in a geologically short period. The "Big Five" mass extinctions include the end-Permian (252 Ma, the most severe, eliminating approximately 96% of marine species) and the end-Cretaceous (66 Ma, caused by an asteroid impact that killed the non-avian dinosaurs). Mass extinctions reset the evolutionary playing field, clearing ecological niches and creating opportunities for surviving lineages to radiate.

Visual Beginner

Modes of speciation:

Mode	Geographic context	Mechanism	Example
Allopatric	Physical separation	Independent evolution in isolation	Kaibab and Abert's squirrels (Grand Canyon)
Peripatric	Small peripheral population isolated	Founder effect + drift + selection	Island species
Parapatric	Adjacent populations across a gradient	Differential selection along a cline	Anthoxanthum grasses (mine contamination)
Sympatric	Same geographic area	Ecological specialization, polyploidy	Apple maggot fly (host shift); wheat polyploidy

The Big Five mass extinctions:

Event	Time (Ma)	Severity	Likely cause
End-Ordovician	~444	~86% species lost	Climate change (glaciation)
Late Devonian	~360	~75% species lost	Climate change, ocean anoxia
End-Permian	~252	~96% marine species lost	Massive volcanism (Siberian Traps)
End-Triassic	~201	~80% species lost	Volcanism (Central Atlantic Magmatic Province)
End-Cretaceous	~66	~76% species lost	Asteroid impact (Chicxulub)

Worked example Beginner

The classic case of allopatric speciation is the pair of squirrels inhabiting opposite rims of the Grand Canyon. The Kaibab squirrel (Sciurus kaibabensis) lives on the north rim; the Abert's squirrel (Sciurus aberti) lives on the south rim. They are descended from a common ancestor that was split when the Grand Canyon formed, creating an impassable geographic barrier.

Over approximately 5-10 million years of separation, the two populations accumulated genetic differences through mutation, natural selection (acting on different environmental conditions on each rim), and genetic drift. Today, they differ in coat color (the Kaibab squirrel has a dark body with a white tail; the Abert's squirrel has a grey body with a dark back) and other morphological traits. They are considered separate species because they are reproductively isolated: even if brought together, they would not interbreed due to behavioral differences accumulated during their long separation.

This example illustrates the three requirements for allopatric speciation: (1) a geographic barrier that prevents gene flow, (2) sufficient time for genetic divergence, and (3) the evolution of reproductive isolating mechanisms (pre-zygotic barriers like behavioral differences, or post-zygotic barriers like hybrid inviability).

Check your understanding Beginner

Formal definition Intermediate+

Macroevolution encompasses evolutionary patterns and processes at and above the species level, including speciation, extinction, adaptive radiation, and long-term trends in lineage diversity.

Speciation

The biological species concept (Mayr, 1942) defines a species as a group of interbreeding natural populations that are reproductively isolated from other such groups. Reproductive isolating mechanisms are classified as:

• Pre-zygotic barriers (prevent hybrid formation): habitat isolation, temporal isolation (different breeding seasons), behavioral isolation (different courtship rituals), mechanical isolation (incompatible reproductive structures), gametic isolation (sperm cannot fertilize egg).
• Post-zygotic barriers (reduce hybrid fitness): reduced hybrid viability (hybrids die early), reduced hybrid fertility (hybrids are sterile, e.g., mules), hybrid breakdown (first-generation hybrids are fertile but subsequent generations are inviable or sterile).

The biological species concept has limitations: it cannot be applied to asexual organisms, fossils, or allopatric populations that cannot be tested for interbreeding. Alternative concepts include the morphological species concept (based on structural features), phylogenetic species concept (the smallest monophyletic group on a phylogenetic tree), and ecological species concept (a group sharing a distinct ecological niche).

Rates of evolution

Evolutionary rates vary enormously. The darwin (d) is a unit of evolutionary rate defined as change by a factor of $e$ (approximately 2.718) per million years in a morphological measurement, or equivalently:

$$d=\frac{\ln (x_2)-\ln (x_1)}{\Delta t}\times 10^6$$

where $x_1$ and $x_2$ are measurements at times $t_1$ and $t_2$, and $\Delta t$ is in millions of years. Observed rates range from $10^{-3}$ darwins (long-term stasis in marine invertebrates) to $10^5$ darwins (rapid change in laboratory selection experiments and recent colonizations).

Key innovations

A key innovation is a novel trait that allows a lineage to exploit a previously inaccessible ecological niche, often triggering an adaptive radiation. Examples include:

• Feathers (originally for insulation/display, co-opted for flight in birds)
• The amniotic egg (allowed vertebrate reproduction on land, freeing tetrapods from aquatic environments)
• C4 photosynthesis (allowed plants to photosynthesize efficiently in hot, dry conditions; evolved independently >60 times)
• Nectar spurs in columbine flowers (associated with pollinator specialization and rapid speciation)

Evo-devo

Evolutionary developmental biology (evo-devo) examines how changes in developmental processes produce evolutionary novelties. Key principles:

1. Changes in gene regulation (not just protein-coding sequences) are the primary driver of morphological evolution. The Pax6 gene is nearly identical in flies and humans, yet the eyes it produces are radically different because the downstream regulatory network differs.

2. Heterochrony (changes in developmental timing) can produce major morphological changes. Neoteny (retention of juvenile features in the adult) in axolotls results from a change in thyroid hormone signaling.

3. Heterotopy (changes in spatial expression) can produce novel structures. The evolution of snake body form from a lizard ancestor involved expansion of Hox gene expression domains that specify thoracic identity along more of the body axis, producing the elongated, limbless body plan.

Key results Intermediate+

Result 1 (Speciation rate and adaptive radiation). Adaptive radiations are characterized by exceptionally high speciation rates early in the radiation, followed by a decline as ecological niches are filled and competition increases. The early burst model predicts that trait diversification rates decline over time, a pattern observed in many clades (e.g., Anolis lizards, cichlid fishes). This declining rate reflects density-dependent cladogenesis: as the number of species in a clade increases, the opportunity for new species to establish in unoccupied niches decreases.

Result 2 (Extinction selectivity). Mass extinctions are not random with respect to species traits. During the end-Cretaceous extinction, species with narrow geographic ranges and tropical distributions were disproportionately affected, while generalist species with broad ranges survived at higher rates. This selective extinction can reshape evolutionary trajectories: the mammals that survived the K-Pg extinction were small, generalist insectivores, and the post-extinction radiation produced the enormous diversity of body plans and ecological roles seen in modern mammals.

Exercise 1

Exercise 2

Advanced treatment Master

The relationship between microevolution and macroevolution has been a central debate in evolutionary biology. The extrapolationist view holds that macroevolution is simply microevolution extended over geological time: the same processes (mutation, selection, drift, gene flow) that change allele frequencies within populations also produce speciation and the patterns of the fossil record. The hierarchical view, articulated most forcefully by Gould (2002), argues that macroevolution involves emergent processes operating at the species level that cannot be reduced to population-level mechanisms.

Species selection is the proposed macroevolutionary analogue of natural selection operating at the species level. Just as individual organisms differ in their probability of survival and reproduction, species differ in their probability of persistence (resistance to extinction) and proliferation (speciation rate). If these species-level properties are heritable (phylogenetically conserved) and variable, then species selection can shape long-term evolutionary patterns. For example, lineages with broad geographic ranges may persist longer (lower extinction rate) but speciate less frequently (lower speciation rate) than lineages with narrow ranges, creating a macroevolutionary trade-off that structures clade diversity over geological time.

Punctuated equilibrium (Eldredge and Gould, 1972) makes two empirical claims: (1) the predominant pattern in the fossil record is one of stasis (species remain morphologically stable for millions of years), punctuated by brief intervals of rapid morphological change associated with speciation; and (2) most morphological change occurs during speciation events, not during the long intervals of species stasis. The mechanism proposed is peripatric speciation: small, geographically isolated populations at the periphery of a species' range evolve rapidly due to founder effects, strong selection in novel environments, and genetic drift. The ancestral population, being large and well-adapted to its stable environment, remains in stasis. The fossil record, which predominantly samples the large, stable central population, therefore shows abrupt appearances of new forms (the peripheral isolates that have evolved into new species and expanded) interspersed with long periods of no change.

The empirical support for punctuated equilibrium is mixed. Comprehensive studies by Hunt (2006, 2007) on bryozoans and by Gingerich on mammals have found that the relative frequency of punctuated versus gradual patterns varies among groups. A meta-analysis by Hannisdal (2007) found that approximately 55-60% of stratigraphic sequences show punctuated patterns and 40-45% show gradual patterns, suggesting that both modes are common and that the debate should focus on identifying the conditions that favor each.

Evo-devo and macroevolution. The deepest insight from evolutionary developmental biology is that the genetic toolkit for building animal bodies is both highly conserved and modular. The same Hox genes, the same signaling pathways (Wnt, BMP, Hedgehog, Notch), and the same transcription factor families (Pax, Fox, Sox, T-box) are deployed across the animal kingdom. Macroevolutionary change -- the origin of new body plans, new organs, new developmental programs -- does not typically require the evolution of new genes. Instead, it arises through changes in cis-regulatory elements (the DNA sequences that control when and where genes are expressed), changes in gene regulatory network architecture, and changes in developmental timing and spatial patterning.

The limb skeleton illustrates this principle. All tetrapod limbs develop from a bud of mesenchyme under the control of a conserved gene regulatory network: apical ectodermal ridge (A ridge producing FGFs), zone of polarizing activity (ZPA producing Shh), and Hox gene expression domains specifying proximal-distal identity. The enormous diversity of tetrapod limbs -- bat wings, horse legs, whale flippers, snake limblessness, human hands -- arises from modifications of this conserved network: changes in FGF and Shh signaling duration and intensity (affecting limb length), changes in Hox expression domains (affecting digit number and identity), and loss of the entire limb program (snakes, through loss of Shh expression).

The molecular clock and dating evolutionary events. The molecular clock hypothesis, proposed by Zuckerkandl and Pauling in 1962, states that amino acid or nucleotide substitutions accumulate at a roughly constant rate over time in selectively neutral regions of the genome. By calibrating the substitution rate with fossil dates at known phylogenetic branching points, the molecular clock can estimate the timing of divergence events that lack a fossil record. In practice, the clock is not perfectly constant: substitution rates vary among lineages (generation time effect, metabolic rate effect), among genes (different selective constraints), and over time (variable selective pressures). Relaxed molecular clock models (such as the uncorrelated lognormal model implemented in BEAST) allow substitution rates to vary along branches of the phylogeny while still providing probabilistic estimates of divergence times with credible intervals. Molecular clock analyses have resolved several major macroevolutionary questions, including dating the origin of placental mammals to approximately 100-85 Ma (before the K-Pg extinction, but with the major diversification occurring after), and confirming that the human-chimpanzee divergence occurred approximately 6-7 Ma.

Mass extinctions in detail. The end-Permian mass extinction (~252 Ma), the most severe in Earth's history, provides a case study in how environmental catastrophe drives macroevolutionary change. The eruption of the Siberian Traps released enormous volumes of CO2 and toxic gases over approximately 600,000 years. The consequences cascaded through Earth systems: extreme global warming (surface temperatures increased by approximately 8-10 degrees C), ocean acidification, widespread ocean anoxia (the expansion of oxygen-minimum zones eliminated most marine habitats), and ozone layer destruction (increasing UV radiation at the surface). Approximately 96% of marine species and 70% of terrestrial vertebrate species were lost. The recovery took 5-10 million years, far longer than after any other mass extinction. The "Lilliput effect" -- the preferential survival of small-bodied organisms -- characterized the post-extinction fauna. The eventual recovery saw the rise of new groups: archosaurs (the lineage including crocodiles and dinosaurs) became dominant terrestrial predators, and the ancestors of modern marine reptile groups radiated into the emptied marine niches.

The Cambrian explosion and its explanations. The Cambrian explosion (~541-485 Ma) saw the first appearance of most modern animal phyla in the fossil record within a geologically brief interval of approximately 40-50 million years. Multiple hypotheses have been proposed to explain this rapid diversification. The environmental triggers hypothesis emphasizes rising atmospheric oxygen (which crossed the threshold needed to support aerobic metabolism in large, active animals), the end of "Snowball Earth" glaciations (which had periodically frozen the planet in the preceding Cryogenian), and changes in ocean chemistry (increased calcium and phosphate concentrations facilitating skeletonization). The ecological triggers hypothesis emphasizes the evolution of predation (the appearance of the first large predators, such as Anomalocaris, created selective pressure for defensive armor, burrowing, and other anti-predator adaptations, launching an evolutionary arms race that drove diversification). The developmental genetics perspective, discussed above, notes that the genetic toolkit for bilaterian development (Hox genes, signaling pathways) evolved in the preceding Ediacaran period, providing the developmental capacity for rapid morphological innovation. Most researchers now favor a multi-causal explanation in which environmental, ecological, and developmental factors converged to produce the Cambrian radiation.

Exercise 3

Exercise 4

Trends in the fossil record. Cope's rule -- the observation that body size tends to increase within lineages over geological time -- illustrates how macroevolutionary trends can arise through multiple mechanisms. The trend toward larger body size in horses (from the dog-sized Hyracotherium 55 Ma to modern Equus) was long cited as a classic example of directional evolution driven by natural selection for larger size in open grassland habitats. However, re-analysis by Jablonski (1997) showed that the trend is better explained by species sorting: small-bodied and large-bodied species both arose, but large-bodied species had higher speciation rates and lower extinction rates, causing the clade to become dominated by large forms over time. The mechanism driving the trend was not selection for larger size within populations but differential proliferation of species that happened to be large.

A similar debate surrounds the trend toward increasing complexity over the history of life. The earliest organisms were prokaryotes with relatively simple genomes and body plans. Over 3.5 billion years, eukaryotes, multicellularity, complex developmental programs, nervous systems, and social behavior evolved. However, whether this represents a driven trend (selection favors complexity) or a passive trend (the only direction available from a minimal starting point is upward, and random walks produce increasing variance and therefore increasing maximum complexity) remains unresolved. Gould (1996) argued for the passive interpretation, noting that the mode of complexity has not changed (most organisms are still bacteria), while only the right tail of the complexity distribution has extended.

Key innovations and their macroevolutionary consequences. The evolution of flight has occurred independently four times: in insects (350 Ma), pterosaurs (228 Ma), birds (150 Ma), and bats (52 Ma). Each origin of flight was followed by a significant adaptive radiation, illustrating how a key innovation can open a vast new adaptive zone. The consequences extend beyond the flying lineage: the evolution of insect flight preceded and likely enabled the evolution of flight in vertebrate predators (pterosaurs, birds, bats), creating a cascade of coevolutionary macroevolutionary events. Similarly, the evolution of C4 photosynthesis (a biochemical innovation that concentrates CO2 around RuBisCO, avoiding photorespiration in hot, dry, high-light environments) has evolved independently more than 60 times in angiosperms. Each origin was followed by increased diversification rates, and C4 grasslands (which expanded dramatically during the late Miocene, approximately 7-5 Ma) reshaped terrestrial ecosystems globally, driving the evolution of grazing adaptations in horses, bovids, and other herbivores.

The role of contingency. The question of whether the history of life is largely contingent (dependent on specific historical events that could have turned out differently) or largely determined by the selective pressures acting on organisms has profound implications for understanding macroevolution. Gould's "replay the tape of life" thought experiment argued that contingency dominates: if the history of life were replayed from an early point, the outcome would likely be radically different. The evolutionary biologist Simon Conway Morris has argued the opposite: convergence (the independent evolution of similar traits in distantly related lineages, such as camera eyes in cephalopods and vertebrates, wings in birds and bats, and social behavior in insects and mammals) demonstrates that the number of viable solutions to environmental challenges is limited, and that replaying the tape would produce broadly similar outcomes.

The resolution may lie in recognizing that contingency and convergence operate at different levels. At the level of broad ecological roles (large marine predator, flying insectivore, photosynthesizer), convergence is common because physics and chemistry constrain the range of viable solutions. At the level of specific evolutionary outcomes (which lineage fills that role, what the detailed anatomy looks like, which genetic mechanisms are used), contingency dominates. The end-Cretaceous asteroid impact eliminated the non-avian dinosaurs but spared small mammals -- a contingent event that determined which lineage would radiate into the large terrestrial vertebrate niches, but did not change the fact that those niches would eventually be filled.

Connections Master

• Body plans 18.01.01. The macroevolution of animal body plans is governed by the conserved developmental genetic toolkit described in 18.01.01. Hox gene duplications and regulatory changes are the molecular substrate for the diversification of body plans during the Cambrian explosion and subsequent adaptive radiations. The concept of deep homology -- that morphologically distinct structures share conserved genetic regulatory circuits -- connects the macroevolutionary patterns discussed here to the developmental mechanisms described in the body plans unit.

• Ecosystem ecology 19.11.01. Mass extinctions and adaptive radiations fundamentally restructure ecosystems, altering energy flow, trophic structure, and nutrient cycling. The post-K-Pg mammalian radiation reshaped terrestrial ecosystems that had been dominated by dinosaurs for over 150 million years. The diversification of flowering plants (angiosperms) during the Cretaceous transformed terrestrial productivity and created new ecological niches for pollinating insects, illustrating how macroevolutionary innovations cascade through ecosystem-level processes.

• Conservation biology 19.14.01. The current extinction crisis is evaluated against the background extinction rates and mass extinction patterns established by paleontological macroevolution. Understanding extinction selectivity from past mass extinctions informs predictions about which taxa are most vulnerable to current threats. The sixth mass extinction, if it proceeds, will represent a macroevolutionary event comparable to the Big Five, with permanent consequences for the tree of life.

• Biogeography 19.12.01. The geographic context of speciation (allopatric, peripatric, sympatric) links macroevolution to biogeographic patterns. Continental breakup and collision create the geographic barriers and connections that drive speciation and extinction over geological time. Island biogeography theory, which predicts species richness from area and isolation, is a macroevolutionary framework applied at ecological timescales.

• Coevolution 19.13.01. Macroevolutionary patterns are frequently driven by coevolutionary interactions. The escape-and-radiate dynamics described by Ehrlich and Raven -- in which a lineage evolves a novel defense and radiates into unoccupied niche space, followed by counter-adaptation by its enemies -- is a coevolutionary macroevolutionary process. The correlated diversification of flowering plants and their insect pollinators during the Cretaceous is one of the most consequential coevolutionary events in the history of life.

• Microevolution 19.07.01. The relationship between micro- and macroevolution is a foundational question. Population genetics provides the mechanisms (mutation, selection, drift, gene flow) that drive allele frequency change within species; macroevolution asks whether these mechanisms alone are sufficient to explain the patterns observed in the fossil record, or whether species-level processes (species selection, differential extinction) provide additional explanatory power. The current consensus is that both levels are necessary for a complete understanding of evolutionary history.

Historical & philosophical context Master

Macroevolution was shaped by the tension between fossil-scale pattern and population-level mechanism. Darwin supplied common descent and natural selection; paleontology supplied extinction, radiation, stasis, and deep-time sequence; the modern synthesis connected large-scale evolutionary change to genetics and population processes. Later debates over punctuated equilibrium, species selection, constraint, and evolvability asked whether macroevolution is only microevolution accumulated over time or whether higher-level patterns require additional explanatory tools. The productive view keeps both sides in play.

The concept of macroevolution emerged from the recognition that the fossil record documents patterns -- mass extinctions, adaptive radiations, long-term trends, and abrupt transitions -- that cannot be fully explained by studying individual populations over short time scales. George Gaylord Simpson, one of the architects of the modern synthesis, argued in Tempo and Mode in Evolution (1944) that the rates and patterns of evolution observed in the fossil record were compatible with the population genetics of the modern synthesis, but he also recognized that the fossil record showed phenomena (rapid bursts of diversification, prolonged stasis, selective extinction) that required special attention.

The punctuated equilibrium debate of the 1970s and 1980s, initiated by Niles Eldredge and Stephen Jay Gould, challenged the gradualist assumptions of the modern synthesis. They argued that the fossil record predominantly shows stasis (species remaining morphologically constant for millions of years) punctuated by brief episodes of rapid change, and that this pattern reflects the speciation process rather than an imperfect fossil record. The debate was productive in forcing paleontologists to quantify stasis and change more rigorously, and in stimulating the development of statistical methods for analyzing stratigraphic sequences.

David Raup and Jack Sepkoski's 1982 statistical analysis of marine fossil families revealed that mass extinctions are not merely the tail of a continuous distribution of extinction intensities but are statistically distinct events, occurring at roughly 26-million-year intervals (a finding that inspired the now largely discredited hypothesis of a periodic extraterrestrial impactor). The confirmation that the end-Cretaceous extinction was caused by an asteroid impact (the Alvarez hypothesis, supported by the iridium anomaly, shocked quartz, and the Chicxulub crater) demonstrated that extrinsic catastrophes can override the normal processes of natural selection and competition, profoundly shaping the history of life.

The integration of molecular phylogenetics with paleontology, beginning in the 1990s and accelerating with genome-scale data, has transformed macroevolution into a more quantitative and predictive discipline. Molecular clocks can date divergences that lack a fossil record; phylogenetic comparative methods can reconstruct ancestral states and test hypotheses about the directionality and rate of trait evolution; and phylogenetic diversity metrics can quantify how much evolutionary history would be lost under different extinction scenarios. These tools have made macroevolution relevant not only to understanding the deep history of life but also to predicting its future trajectory under anthropogenic pressure.

The philosophical question of whether macroevolution is reducible to microevolution remains open. The extrapolationist position (that macroevolution is microevolution extended over geological time) is supported by the success of population genetic models in explaining many macroevolutionary patterns. However, emergent phenomena at the species level (species selection, the effect of species-level traits like geographic range on extinction probability) and the role of developmental constraint in limiting the range of accessible phenotypes suggest that macroevolution involves processes that are not straightforwardly predictable from population genetics alone. The most productive framework treats macroevolution as a multi-level process in which population-level mechanisms, species-level dynamics, and developmental constraints all contribute to the patterns observed in the history of life.

The practical relevance of macroevolutionary thinking extends beyond paleontology into conservation biology, agriculture, and medicine. Understanding how past mass extinctions selectively eliminated certain types of organisms allows conservation biologists to predict which taxa are most vulnerable to the current extinction crisis and to prioritize conservation efforts accordingly. The phylogenetic comparative methods developed by macroevolutionary biologists are used to identify evolutionary innovations that could be exploited for crop improvement or pharmaceutical development. The study of evolutionary rates and patterns of diversification informs our understanding of how pathogens evolve resistance to drugs and how pest species adapt to control measures. In these ways, macroevolution is not merely a historical science but a framework with direct applications to the challenges of the Anthropocene, providing the deep-time perspective needed to contextualize the rapid biological changes occurring in the present day.

Bibliography Master

1. Campbell, N. A. & Reece, J. B. Biology, 12th ed. (Pearson, 2020). Ch. 24-25.

2. Futuyma, D. J. & Kirkpatrick, M. Evolution, 4th ed. (Sinauer, 2017).

3. Gould, S. J. The Structure of Evolutionary Theory (Harvard University Press, 2002).

4. Mayr, E. Systematics and the Origin of Species (Columbia University Press, 1942).

5. Raup, D. M. & Sepkoski, J. J. "Mass extinctions in the marine fossil record." Science 215 (1982) 1501-1503.

6. Eldredge, N. & Gould, S. J. "Punctuated equilibria: an alternative to phyletic gradualism." In Schopf, T. J. M. (ed.), Models in Paleobiology (Freeman, Cooper, 1972) 82-115.

7. Carroll, S. B. Endless Forms Most Beautiful (W. W. Norton, 2005).

8. Raup, D. M. Extinction: Bad Genes or Bad Luck? (W. W. Norton, 1991).

9. Knoll, A. H. "Biomineralization and evolutionary history." Rev. Mineral. Geochem. 54 (2003) 329-356.

10. Losos, J. B. Improbable Destinies: Fate, Chance, and the Future of Evolution (Riverhead Books, 2017).

11. Jablonski, D. "Extinction: past and present." Nature 427 (2004) 589-595.

12. Erwin, D. H. Extinction: How Life on Earth Nearly Ended 250 Million Years Ago (Princeton University Press, 2006).

13. Stanley, S. M. "A theory of evolution above the species level." Proc. Natl. Acad. Sci. 72 (1975) 646-650.

14. Conway Morris, S. Life's Solution: Inevitable Humans in a Lonely Universe (Cambridge University Press, 2003).

15. Simpson, G. G. Tempo and Mode in Evolution (Columbia University Press, 1944).

16. Hunt, G. "The relative importance of directional change, random walks, and stasis in the evolution of fossil lineages." Proc. Natl. Acad. Sci. 103 (2006) 17309-17314.

17. Ehrlich, P. R. & Raven, P. H. "Butterflies and plants: a study in coevolution." Evolution 18 (1964) 586-608.

Exercise 5

Exercise 6

Exercise 7

Exercise 8