19.08.03 · eco-evo-bio / macroevolution

Trends in the fossil record: Cope's rule, the Red Queen hypothesis, and evolutionary stasis

stub3 tiersLean: nonepending prereqs

Anchor (Master): Gould, S. J. — Wonderful Life (1989)

Intuition Beginner

The fossil record does not just record individual species -- it reveals long-term patterns that play out over millions of years. Three of the most important patterns are:

Body size tends to increase over time. This observation, called Cope's rule (after the American paleontologist Edward Drinker Cope), notes that lineages often start small and evolve toward larger body sizes over geological time. Horses evolved from the dog-sized Hyracotherium to modern Equus. Dinosaurs, mammals, and many marine invertebrate groups all show trends toward larger size. But Cope's rule is not a law -- there are many exceptions. Some lineages get smaller, and in many groups the trend is better described as an increase in the range of body sizes rather than a consistent increase in average size.

Some lineages barely change for millions of years. The coelacanth, a lobe-finned fish, looks almost identical to fossils that are 400 million years old. Horseshoe crabs, ginkgo trees, and the brachiopod Lingula are other examples of living fossils -- lineages that have remained morphologically static for vast stretches of geological time. This pattern of long-term stability, called stasis, is punctuated by brief episodes of rapid change, often associated with the origin of new species. This is the punctuated equilibrium model proposed by Eldredge and Gould in 1972.

Species must constantly evolve just to maintain their current fitness. The Red Queen hypothesis, named after the Red Queen in Lewis Carroll's Through the Looking-Glass who says "it takes all the running you can do, to keep in the same place," proposes that organisms are locked in an evolutionary arms race with their competitors, predators, parasites, and pathogens. Every time a species improves, its rivals improve too, and no one gets ahead. The result is perpetual change without overall progress -- running in place.

Visual Beginner

Major trends and patterns in the fossil record:

Pattern	Description	Example
Cope's rule	Body size increases within lineages over time	Horses: Hyracotherium (dog-sized) to Equus
Punctuated equilibrium	Long stasis punctuated by brief rapid change	Trilobite species in the fossil record
Red Queen dynamics	Constant evolutionary change with no net fitness gain	Host-parasite coevolution
Directional trend	Consistent change in a trait over time	Brain size increase in hominins
Living fossils	Morphological stasis over tens of millions of years	Coelacanth, horseshoe crab, Lingula

Worked example Beginner

Cope's rule in horses is one of the most cited examples of a directional evolutionary trend. The horse lineage spans approximately 55 million years, from the early Eocene Hyracotherium (formerly called Eohippus) to the modern genus Equus. Over this interval, several traits changed directionally:

Body size increased from approximately 10 kg (Hyracotherium, about the size of a small dog) to approximately 500 kg (modern Equus).
Tooth crown height (hypsodonty) increased. Early horses had low-crowned teeth suited for browsing soft leaves; modern horses have tall, continuously growing teeth adapted for grinding abrasive grasses.
Toe number decreased. Hyracotherium had four toes on the front feet and three on the hind feet; modern horses stand on a single toe (the hoof), with the other toes reduced to vestigial splint bones.
Leg length increased, producing longer stride and greater running speed.

For decades, this was presented as a textbook case of directional natural selection driving a lineage toward larger size, higher-crowned teeth, and fewer toes in response to the spread of grasslands during the Miocene. However, re-analysis by Jablonski (1997) showed that the pattern is more complex: multiple horse species coexisted at most time intervals, some small and some large, and the apparent trend toward larger average size resulted from the differential survival and speciation of larger-bodied species (species sorting) rather than from within-lineage directional selection. The overall trend in the clade masks a more complicated pattern of branching, extinction, and differential proliferation.

Check your understanding Beginner

Exercise (easy, multiple choice).

What does Cope's rule describe?

A. Species must constantly evolve to maintain fitness relative to competitors B. Body size tends to increase within lineages over geological time C. Evolution proceeds through long periods of stasis punctuated by rapid change D. Complex traits cannot be regained once lost

Hint

Cope's rule is named after Edward Drinker Cope, a paleontologist who studied fossil vertebrates. He noticed a recurring pattern in the body sizes of organisms over time.

Answer

Option B. Cope's rule describes the tendency for body size to increase within evolutionary lineages over geological time. This pattern has been documented in dinosaurs, mammals, marine invertebrates, and many other groups. However, it is not universal: many lineages show no trend, and some show size decrease. The rule is best understood as a common but not inevitable pattern, and in many cases the apparent trend toward larger average size results from increases in the variance and maximum of body size rather than from consistent within-lineage increase.

Exercise (easy, multiple choice).

The Red Queen hypothesis proposes that:

A. The largest species in a clade always have the lowest extinction risk B. Species must constantly evolve just to maintain their fitness relative to evolving competitors, parasites, and changing environments C. Evolution always proceeds at a constant, gradual rate D. Body size always increases over time

Hint

The name comes from a character in Lewis Carroll's Through the Looking-Glass who must keep running just to stay in the same place.

Answer

Option B. The Red Queen hypothesis, proposed by Leigh Van Valen in 1973, proposes that species must continually evolve and adapt not to gain an advantage but simply to survive in the face of constantly evolving competitors, predators, parasites, and pathogens. Like the Red Queen in Through the Looking-Glass, species must run (evolve) as fast as they can just to stay in the same place (maintain their current fitness relative to the biotic environment). The hypothesis predicts that extinction probability is approximately constant within lineages, regardless of how long they have existed, because the biotic environment is constantly deteriorating as other species evolve.

Exercise (easy, true/false).

Living fossils like the coelacanth have stopped evolving at the genetic level.

Hint

Morphological stasis (no visible change in body form) does not necessarily mean genetic stasis (no change at the DNA level).

Answer

False. Living fossils have remained morphologically similar to their ancient ancestors for tens of millions of years, but they have not stopped evolving at the genetic level. Molecular studies show that coelacanths, horseshoe crabs, and other living fossils continue to accumulate DNA substitutions at rates comparable to other lineages. Morphological stasis likely results from stabilizing selection (the current body plan is well-adapted to a stable niche, and most mutations that alter morphology are deleterious) rather than from an absence of genetic change.

Formal definition Intermediate+

Cope's rule

Cope's rule is the empirical generalization that body size tends to increase within evolutionary lineages over geological time. The rule is named after Edward Drinker Cope, who noted the pattern in fossil mammals from North America. Modern analyses have tested the rule quantitatively in many groups:

Support. Body size increase over time has been documented in dinosaurs (Hone et al., 2005), Cenozoic mammals (Alroy, 1998), foraminifera, and some marine invertebrate clades. In each case, the mean or maximum body size within the clade increases over geological time.
Counterexamples. Many groups show no directional trend in body size, and some show decreases. Birds, which evolved from large theropod dinosaurs, show an overall trend toward smaller body size (the "inverse Cope's rule"). Within many clades, decreases in body size are as common as increases.
Mechanisms. When the trend is real, several mechanisms may explain it: (1) individual-level selection for larger size, which confers advantages in competition for mates and resources, thermoregulation (Bergmann's rule in endotherms), and predator avoidance; (2) lower extinction risk of large-bodied species, which may have broader geographic ranges, more generalized diets, or greater behavioral flexibility; (3) species sorting, in which large-bodied species within a clade have higher speciation rates or lower extinction rates, causing the clade to become dominated by large forms over time without requiring within-lineage increase; (4) passive drift from a small starting point, since body size has a lower bound (minimum viable size) but no hard upper bound, random walks in size space will tend to increase the maximum and sometimes the mean.

Directional trends

Beyond body size, other directional trends have been proposed:

Complexity. The maximum structural and organizational complexity of organisms has increased over the history of life, from prokaryotes to eukaryotes to multicellular organisms to animals with complex nervous systems. Whether this represents a driven trend (selection favoring complexity) or a passive trend (the only direction available from a minimal starting point is upward) is debated.
Brain size in hominins. Cranial capacity increased from approximately 400 cm $^{3}$ in Australopithecus afarensis (~3.5 Ma) to approximately 1,400 cm $^{3}$ in modern Homo sapiens. This is one of the best-documented directional trends in the fossil record, driven by selection for cognitive abilities related to social interaction, tool use, and environmental problem-solving.
Diversity. The total number of species (or genera, or families) in the global fossil record has generally increased from the Cambrian to the present, with major drops during mass extinctions 19.08.02 pending. This increase may reflect a genuine increase in global biodiversity, or it may be an artifact of better preservation and sampling in younger rocks (the "Pull of the Recent").

The Red Queen hypothesis

Leigh Van Valen (1973) proposed the Red Queen hypothesis based on his analysis of extinction rates in the fossil record. He found that extinction probability within a lineage is approximately constant regardless of how long the lineage has existed: old lineages go extinct at the same rate as young ones. This is unexpected if species gradually improve over time (which would cause extinction probability to decrease with age).

Van Valen's explanation was that the effective environment of a species is constantly deteriorating because other species in its community are evolving. A predator's prey evolves better defenses; a parasite's host evolves better immunity; a competitor evolves more efficient resource use. Each species must evolve just to keep up with the evolving biotic environment. The result is a coevolutionary arms race in which all participants are running as fast as they can just to maintain their relative fitness.

Formally, if $w_{i} (t)$ is the absolute fitness of species $i$ at time $t$ , the Red Queen predicts that $d w_{i} / d t \approx 0$ even though the underlying genetic and phenotypic composition of the population is constantly changing. The species is not standing still -- it is evolving -- but its fitness relative to the biotic environment stays roughly constant.

Punctuated equilibrium

Punctuated equilibrium (Eldredge and Gould, 1972) makes two central claims:

Stasis is the predominant pattern. Most species in the fossil record remain morphologically stable for the duration of their existence, often millions of years. Morphological change is concentrated in brief episodes (thousands to tens of thousands of years) associated with speciation events.
Speciation is the engine of change. Most morphological evolution occurs during speciation, not during the long intervals of species stasis. The mechanism proposed is allopatric speciation: small peripheral populations, isolated from the large central population, evolve rapidly due to founder effects, genetic drift, and strong selection in novel environments. The large, stable central population remains in stasis, and the fossil record, which preferentially samples this central population, shows long periods of no change.

The alternative is phyletic gradualism: the classical view that morphological change accumulates gradually and continuously within lineages, with species slowly transforming into new forms over time. Under phyletic gradualism, the apparent stasis in the fossil record is an artifact of incomplete preservation, and the "sudden" appearances of new species represent gaps in the record rather than rapid evolutionary events.

Stasis mechanisms

Several mechanisms can maintain morphological stasis over millions of years:

Stabilizing selection. When the optimal phenotype remains constant (as in a stable environment), selection removes extreme variants, maintaining the population near the optimum. This is the simplest explanation for stasis.
Habitat tracking. As environmental conditions change (e.g., temperature, sea level), species may shift their geographic range to track their preferred habitat rather than evolving new adaptations. The species stays in the same environment and therefore does not change morphologically, even though its geographic location shifts.
Developmental constraints. The genetic and developmental architecture of a species may limit the range of phenotypes that can be produced. Even if selection favors a change, the developmental system may not be able to produce it, resulting in stasis.
Gene flow. Large, geographically widespread species experience gene flow among populations, which homogenizes allele frequencies and prevents local populations from diverging. This counteracts the divergent selection that would drive morphological change.

Key theorem with proof Intermediate+

Theorem (Van Valen's Red Queen -- constant extinction probability). If the fitness of each species depends on the state of other species in the community (biotic environment), and all species are evolving, then the probability of extinction per unit time for any given species is approximately constant and independent of the species' age.

Proof sketch. Consider a community of $N$ interacting species. Let the fitness of species $i$ depend on a vector of traits $x_{i}$ and on the states of other species: $w_{i} = f (x_{i}, x_{- i})$ , where $x_{- i}$ denotes the traits of all species other than $i$ . Species $i$ evolves to maximize $w_{i}$ , but as $x_{i}$ changes, it alters the fitness landscape of other species. Each species is simultaneously adapting to a moving target.

Model the trait evolution of each species as a random walk in a high-dimensional trait space, where the "optimal" trait vector for species $i$ is constantly shifting because other species are also evolving. The probability that species $i$ goes extinct in a small time interval $Δ t$ depends on how far its current trait vector is from the shifting optimum. Because the optimum is shifting at a rate determined by the evolution of other species (which is itself an ongoing process), the probability of falling far enough below the optimum to go extinct is approximately constant over time.

More formally, let the fitness of species $i$ relative to its community be modeled as:

w_{i} (t) = w_{0} - d (x_{i} (t), x_{i}^{*} (t)),

where $x_{i}^{*} (t)$ is the time-varying optimum (determined by the traits of other species) and $d$ is a distance function. Species $i$ goes extinct when $w_{i}$ falls below a critical threshold. If both $x_{i}$ and $x_{i}^{*}$ are performing random walks (the species is evolving, but the optimum is also moving), then the distance $d$ follows a distribution that reaches a stationary state. In this stationary state, the probability that $d$ exceeds the extinction threshold in any interval $Δ t$ is constant and independent of how long the species has existed. $□$

The key empirical prediction is that a plot of the proportion of lineages surviving as a function of their age should show an approximately exponential decay -- indicating a constant per-unit-time extinction hazard. Van Valen (1973) showed that survivorship curves for many groups of fossil taxa approximate this pattern, consistent with the Red Queen prediction.

Exercise 1

Exercise 1 (medium, short answer).

Explain the difference between a driven trend and a passive trend in body size evolution. How can the two be distinguished using data from a fossil lineage?

Hint

A driven trend means there is a force pushing the lineage in one direction. A passive trend means the lineage wanders randomly but hits a boundary at one end. Consider what happens to the minimum and maximum values in each case.

Answer

A driven trend occurs when there is a consistent directional force (such as natural selection favoring larger size) pushing the lineage in one direction. In a driven trend, both the minimum and maximum body size in the clade increase over time, and the ancestor of each new species tends to be larger than the ancestor of the previous species.

A passive trend occurs when evolution is random (no directional bias) but there is a boundary at one end of the trait distribution. For body size, there is a minimum viable size (set by physiological constraints) but no hard maximum. Random increases and decreases in body size are equally likely, but decreases cannot go below the minimum. Over time, the variance increases, and the maximum body size drifts upward even though there is no selective force favoring larger size. The minimum stays near the boundary.

To distinguish the two: examine whether the minimum body size in the clade increases over time. If both minimum and maximum increase, the trend is driven. If only the maximum increases while the minimum stays roughly constant (or the distribution becomes increasingly right-skewed), the trend is passive. McShea (1994) formalized this test and applied it to several fossil groups, finding that many apparent Cope's rule trends are better explained as passive increases in variance rather than driven increases.

Key results Intermediate+

Result 1 (Punctuated equilibrium is common but not universal). Meta-analyses of stratigraphic sequences by Hunt (2006, 2007) and others have found that approximately 55-60% of well-sampled fossil sequences show punctuated patterns (stasis interrupted by rapid change), while 40-45% show gradual patterns. Both modes are real and common. The relative frequency varies among groups: marine invertebrates with good skeletal preservation (bryozoans, trilobites, foraminifera) tend to show more punctuated patterns, while planktonic microfossils and some mammal lineages show more gradual change. Punctuated equilibrium is not an alternative to gradual microevolutionary change -- both patterns are produced by the same microevolutionary processes (selection, drift, gene flow) operating under different ecological and demographic conditions.

Result 2 (Stasis is empirically robust). The prevalence of stasis in the fossil record is one of its most robust and underappreciated patterns. Species in the fossil record commonly persist for 1-10 million years with little or no directional morphological change, despite the expectation from population genetics that even weak directional selection should produce observable change over such intervals. This "stasis paradox" suggests that stabilizing selection, habitat tracking, gene flow, and developmental constraints are more effective at maintaining morphological constancy than population genetic models predict. The prevalence of stasis is a strong argument that the default state of a species in a stable environment is morphological constancy, not gradual change.

Result 3 (The Red Queen is supported with qualifications). Van Valen's finding that extinction probability is roughly constant within lineages has been broadly supported by subsequent analyses, though with important qualifications. Constant extinction probability is observed primarily in periods of background extinction (not during mass extinctions 19.08.02 pending, which override the Red Queen dynamic). The mechanism is best supported for tightly coupled coevolutionary systems (host-parasite, predator-prey) where arms races are well-documented. For species whose fitness is primarily determined by the abiotic environment (temperature, sea level, chemistry), the Red Queen is less directly applicable, though climate change itself may be driven or amplified by biotic processes.

Result 4 (Cope's rule is often a passive trend). Analyses by McShea (1994), Jablonski (1997), and others have shown that many apparent instances of Cope's rule are better explained as passive increases in the variance and maximum of body size distributions rather than as driven trends. The critical test is whether the minimum body size in the clade increases over time. In most cases where this test has been applied, the minimum stays constant or decreases, while only the maximum increases -- the signature of a passive trend from a small ancestor near a lower boundary, not of directional selection for larger size. This does not mean Cope's rule is never driven, but it means that the null hypothesis of a passive trend must be rejected before invoking driven increase.

Exercise 2

Exercise 2 (medium, short answer).

Describe the punctuated equilibrium model and explain how it differs from phyletic gradualism. What evidence would you look for in the fossil record to distinguish between the two?

Hint

Both models attempt to explain the fossil record. Gradualism predicts smooth, continuous change; punctuated equilibrium predicts long periods of no change interrupted by abrupt transitions. What would a stratigraphic sequence of fossils look like under each model?

Answer

Punctuated equilibrium (Eldredge and Gould, 1972) proposes that species typically exist in morphological stasis for most of their duration (often millions of years) and that significant morphological change is concentrated in brief episodes (thousands to tens of thousands of years) associated with speciation events, particularly allopatric speciation in small peripheral populations. The fossil record of the large, stable central population shows no change (stasis), and new species appear abruptly when peripheral isolates that have evolved into distinct forms expand their range into the area being sampled.

Phyletic gradualism proposes that morphological change accumulates gradually and continuously within lineages over time, without requiring speciation events. Under this model, species slowly transform into new forms, and the apparent stasis and abrupt transitions in the fossil record are artifacts of incomplete preservation (gaps in the record that make gradual change appear sudden).

To distinguish between the two in the fossil record: (a) look for stratigraphic sequences with fine-scale temporal resolution where gradual change could be detected if it occurred; (b) test whether morphological change is concentrated at speciation events (appearance of new species) or occurs continuously within species; (c) use statistical tests such as the Bookstein test (which compares the magnitude of change at speciation events to the magnitude of within-species variation) to assess whether change is punctuated or gradual. The best evidence comes from continuously deposited marine sedimentary sequences where populations are sampled at fine (centimeter-scale) intervals over millions of years.

Exercise 3

Exercise 3 (medium, short answer).

Explain Van Valen's evidence for the Red Queen hypothesis. What pattern in the fossil record did he observe, and how does the Red Queen hypothesis explain it?

Hint

Van Valen plotted the proportion of genera surviving as a function of their age. What shape did this curve have, and what does that shape imply about extinction probability?

Answer

Van Valen (1973) compiled data on the durations of fossil genera across many groups of organisms (mammals, marine invertebrates, protists) and plotted survivorship curves -- the proportion of genera surviving as a function of their age (time since origin). He found that these curves were approximately exponential, indicating a constant per-unit-time probability of extinction regardless of how long a genus had existed. Old genera were no more likely to go extinct (per unit time) than young genera.

This is puzzling under a simple adaptive model: if species and genera gradually improve through natural selection, then older, more adapted lineages should have lower extinction risk. The observation that they do not led Van Valen to propose the Red Queen hypothesis: the effective environment of a species is not static but is constantly deteriorating because other species (competitors, predators, parasites) are evolving. Each species must continually adapt just to maintain its fitness relative to this shifting biotic environment. Because all species are running this race simultaneously, no species gains a lasting advantage, and extinction probability remains constant regardless of age.

Advanced treatment Master

Statistical tests for punctuated equilibrium

The Bookstein test (Bookstein, 1987; used by Gould and Eldredge in their empirical analyses) provides a statistical framework for distinguishing punctuated from gradual change in stratigraphic sequences. The test examines whether the morphological distance between ancestral and descendant species is significantly greater than the range of within-species variation observed over the duration of each species. If change is punctuated, the between-species distance should exceed the within-species range; if change is gradual, the transition should show intermediate forms and the between-species distance should fall within the range of within-species variation.

The test operates on a stratophenetic series -- a sequence of morphological measurements taken from fossil populations sampled at successive stratigraphic horizons. For each species in the series, the variance (or range) of the measured trait is estimated across all horizons where the species is present. The morphological distance between consecutive species is then compared to this within-species variance. If the between-species step is significantly larger than expected from within-species fluctuation (assessed by an F-test or permutation test), the pattern supports punctuated equilibrium.

Gingerich's rate measures provide a complementary approach. Gingerich (1983, 1993) compiled evolutionary rates (measured in darwins or haldanes) across a wide range of timescales, from laboratory selection experiments (years) to the fossil record (millions of years). He found an inverse relationship between the rate of evolution and the timescale over which it is measured: rates measured over short intervals are fast, while rates measured over long intervals are slow. This pattern, called the rate-timescale relationship, is consistent with punctuated equilibrium (brief rapid change averaged over long intervals of stasis produces apparently slow rates) but is also consistent with other models (directional change that reverses direction, or a random walk where the net displacement grows more slowly than the total distance traveled).

Passive vs driven trends

McShea (1994) provided the definitive framework for distinguishing passive from driven trends in body size (or any other trait). The key insight is that in a passive trend, the lineage begins near a lower bound and diffuses away from it: the minimum stays near the boundary, the maximum increases, and the mean increases because the distribution becomes increasingly right-skewed. In a driven trend, a directional force pushes the entire distribution: the minimum, mean, and maximum all shift in the same direction.

The three-fold test:

Minimum test. Does the minimum body size in the clade increase over time? If yes, the trend is driven. If the minimum stays constant while only the maximum increases, the trend is passive.
Ancestor-descendant test. Are ancestor-descendant pairs consistently larger in the descendant? If most pairs show increase, the trend is driven. If increases and decreases are equally common but the clade's maximum drifts upward, the trend is passive.
Skewness test. Does the distribution of body sizes become increasingly right-skewed over time? A passive trend from a boundary produces increasing right skew. A driven trend produces symmetric or decreasing skew.

Application of these tests to fossil data has shown that many classic examples of Cope's rule -- including the horse lineage -- are better explained as passive trends. The increase in maximum body size in horses, dinosaurs, and mammals reflects the diffusion of body size away from a small starting point, not directional selection for larger size within every lineage.

Contingency vs convergence

Stephen Jay Gould's replay thesis, articulated in Wonderful Life (1989), argues that the history of life is profoundly contingent: if the tape of life were replayed from any early point, the outcome would be radically different. The Burgess Shale fauna, with its bizarre and disparate body plans, represents a richness of anatomical possibility that was pruned by the contingency of survival rather than by the deterministic sieve of natural selection. Gould argued that the organisms that survived the Cambrian were not objectively "fitter" but were the beneficiaries of historical accident.

Convergence provides the counter-argument. Convergent evolution -- the independent evolution of similar traits in distantly related lineages -- demonstrates that the number of viable solutions to environmental challenges is limited. The classic case is the marsupial-placental convergence: marsupial mammals in Australia evolved forms strikingly similar to placental mammals elsewhere: the marsupial wolf (Thylacinus) converged on the wolf, the marsupial mole (Notoryctes) on the mole, the marsupial sugar glider (Petaurus) on the flying squirrel, and the marsupial numbat on the anteater. These convergences suggest that natural selection reliably produces similar outcomes when similar ecological niches are available, supporting a degree of predictability in evolution.

The resolution is that contingency and convergence operate at different levels. At the level of broad ecological roles (large terrestrial herbivore, fast-swimming marine predator, flying insectivore), convergence dominates because physics, ecology, and functional morphology constrain the range of viable designs. At the level of specific evolutionary outcomes (which lineage fills a role, what its genome looks like, which developmental pathway produces the convergent trait), contingency dominates. The marsupial wolf looked like a wolf and acted like a wolf, but it was not a wolf -- it was a marsupial with a fundamentally different reproductive biology, cranial anatomy, and evolutionary history.

Dollo's law and the zero-force evolutionary law

Dollo's law (Dollo, 1893) states that complex traits, once lost, are unlikely to be regained in exactly the same form. The rationale is that the loss of a complex trait (e.g., limbs in snakes, teeth in birds, eyes in cave-dwelling organisms) involves the degradation or loss of many genes and developmental pathways. Once these are lost through mutation and drift, the probability that the exact same trait will re-evolve by the exact same genetic mechanism is vanishingly small.

Dollo's law is a probabilistic statement, not an absolute prohibition. Complex traits can be regained, but the pathway is different and the result is usually not identical to the original. The evolution of new limb-like structures in some lizards that have lost limbs (e.g., skinks with reduced limbs that have re-elongated) does not violate Dollo's law because the re-evolved structures use different developmental mechanisms than the original limbs.

The zero-force evolutionary law (ZFEL), proposed by McShea and Brandon (2010), states that in the absence of selection and constraint, diversity and complexity will tend to increase spontaneously. The argument is that in a system of heritable variation, random processes (mutation, drift) will generate variation in all directions. Without selection to maintain uniformity or constraint to limit the range of variants, the variance of traits within and among lineages will increase over time. This is the null model against which directional trends should be assessed: before invoking selection for larger size (Cope's rule) or greater complexity, one must show that the observed trend exceeds the passive increase expected from the ZFEL.

Key innovations and adaptive radiations

A key innovation is a novel trait that enables a lineage to exploit a previously inaccessible ecological niche, often triggering an adaptive radiation -- a rapid increase in species number and morphological diversity. Key innovations and the radiations they trigger are a major source of directional trends in the fossil record.

Examples:

C4 photosynthesis evolved independently more than 60 times in angiosperms, and each origin was followed by increased diversification rates. C4 grasslands expanded dramatically during the late Miocene (~7-5 Ma), reshaping terrestrial ecosystems and driving the evolution of grazing adaptations in horses, bovids, and other herbivores.
Nectar spurs in columbine flowers (Aquilegia) are associated with pollinator specialization and rapid speciation. Lineages with nectar spurs have higher net diversification rates than their close relatives without spurs.
Pharyngeal jaws in cichlid fishes (a second set of jaws in the throat that processes food independently of the oral jaws) freed the oral jaws from their dual role in food capture and processing, allowing rapid diversification of feeding morphologies and contributing to the explosive radiation of cichlids in African rift lakes.
Feathers in theropod dinosaurs, originally evolved for insulation and display, were co-opted (exapted) for flight, opening the aerial niche and triggering the radiation of birds into over 10,000 species.

The macroevolutionary consequence of key innovations is that they can produce asymmetries in the tree of life: a few lineages become enormously species-rich and morphologically diverse (e.g., beetles, with over 400,000 species; orchids, with over 28,000 species), while their close relatives remain species-poor. These asymmetries are a major pattern in the history of life that requires explanation at the macroevolutionary level.

Exercise 4

Exercise 4 (hard, analysis).

Evaluate the claim that the increase in organismal complexity over the history of life represents a driven evolutionary trend. What evidence supports a driven interpretation? What evidence supports a passive interpretation? How would you test between them?

Hint

The earliest organisms were prokaryotes with relatively simple body plans. Over 3.5 billion years, eukaryotes, multicellularity, complex development, and nervous systems evolved. But most organisms today are still bacteria. Is the trend in the mean or only in the maximum? Consider McShea's passive vs driven framework and the ZFEL.

Answer

Evidence for a driven trend. The maximum complexity of organisms has increased substantially over the history of life: prokaryotes appeared ~3.5 Ga, eukaryotes ~2 Ga, multicellular organisms ~1 Ga, animals with complex nervous systems ~540 Ma, and socially complex organisms (eusocial insects, humans) within the last ~100 Ma. Several mechanisms could drive this trend: (a) selection favoring complexity because complex organisms can exploit more diverse resources and occupy more niches; (b) the " ratchet" of gene duplication and divergence, which irreversibly increases genomic complexity; (c) coevolutionary arms races that drive increasing sophistication of offense and defense.

Evidence for a passive trend. Gould (1996) argued that the complexity trend is passive, not driven. His argument has three parts: (a) life began at the minimum possible complexity (a single cell with minimal genome), so the only direction available for change is upward; (b) random walks in complexity space will tend to increase the variance and therefore the maximum, even without any directional bias; (c) the mode of complexity has not changed -- most organisms today are still bacteria, and the "increase" in complexity reflects expansion of the right tail of the distribution, not a shift in the distribution as a whole.

How to test. Apply McShea's three-fold test: (1) Does the minimum complexity increase over time? If the simplest organisms have become more complex over geological time, the trend is driven. If the minimum stays near the bacterial baseline while only the maximum increases, the trend is passive. (2) Are ancestor-descendant pairs consistently more complex in the descendant? (3) Does the distribution of complexity become increasingly right-skewed over time? The available evidence suggests that the minimum has not increased (the simplest extant bacteria are no more complex than ancient ones), the distribution is strongly right-skewed, and the mode has remained at the bacterial level -- all consistent with a passive trend. However, within specific lineages (e.g., the vertebrate lineage leading to mammals), the trend may be driven by selection for increased cognitive and behavioral complexity. The global trend is likely passive; lineage-specific trends within the right tail of the distribution may be driven.

Exercise 5

Exercise 5 (hard, analysis).

Explain Dollo's law and discuss its relationship to contingency in evolution. Does the observation that complex traits are rarely regained in exactly the same form support Gould's contingency thesis, or does convergent evolution undermine it?

Hint

Dollo's law says lost complex traits are unlikely to be regained. Contingency says evolutionary outcomes depend on historical accident. Convergence says similar traits evolve independently. Are these compatible or contradictory? Consider whether convergence produces the same trait or only a functionally similar one.

Answer

Dollo's law states that complex traits, once lost, are unlikely to be regained in exactly the same form because the genetic and developmental machinery underlying the trait degrades after the trait is no longer under selection. Genes that were once essential for building the trait accumulate loss-of-function mutations, regulatory elements decay, and the developmental pathways that produced the trait are disrupted. Reversing this process -- reconstituting the exact genetic and developmental basis of the lost trait -- is vanishingly improbable because it would require simultaneous reversion of many independent mutations.

Dollo's law directly supports the contingency thesis: it demonstrates that evolutionary history matters because the loss of a trait constrains future evolutionary possibilities. Snakes lost their limbs through degradation of limb-development genes and regulatory elements; the probability that a snake will re-evolve exactly the same limbs with exactly the same genetic basis is essentially zero. This is contingency at the mechanistic level.

However, convergence at the functional level is entirely compatible with Dollo's law and does not undermine contingency. Organisms that have lost a trait may re-evolve a functionally similar trait through different genetic and developmental mechanisms. The lens of the vertebrate eye and the lens of the cephalopod eye are convergent (both focus light using a lens), but they develop by entirely different mechanisms (the vertebrate lens forms from surface ectoderm; the cephalopod lens forms from ectoderm in a completely different embryological context). The functional outcome is similar, but the mechanism is different -- contingency at the mechanistic level coexists with convergence at the functional level.

The synthesis is that Dollo's law constrains the mechanistic pathway of evolution (you cannot go back to exactly where you were), while convergence demonstrates that the functional landscape of evolution has a limited number of peaks (organisms independently find similar solutions to similar problems). Gould's contingency thesis is strongest when applied to the specific mechanistic and historical details of evolution; the convergence counter-argument is strongest when applied to the broad functional and ecological level. Both are correct at their respective levels of analysis.

Connections Master

Macroevolution 19.08.01. This unit extends the macroevolutionary framework introduced in 19.08.01, focusing specifically on the directional patterns and temporal dynamics observable in the fossil record. The concepts of punctuated equilibrium, species selection, and adaptive radiation introduced in 19.08.01 are developed here in the context of specific fossil trend data and quantitative tests.
Mass extinctions 19.08.02 pending. Mass extinctions can abruptly terminate long-term evolutionary trends and reset the starting conditions for subsequent diversification. The end-Cretaceous extinction terminated the trend toward larger body size in dinosaurs and created the opportunity for the mammalian Cope's rule trend. The end-Permian extinction nearly eliminated the trend toward increasing marine diversity that had characterized the Paleozoic. Mass extinctions and evolutionary trends are complementary perspectives on the fossil record.
Natural selection 19.03.01. The mechanisms underlying fossil trends -- directional selection (driven trends), stabilizing selection (stasis), and the Red Queen (coevolutionary arms races) -- are extensions of the microevolutionary processes described in 19.03.01. The question of whether macroevolutionary trends can be fully explained by microevolutionary selection or require species-level processes is a central theme connecting this unit to the foundations of evolutionary theory.
Genetic drift 19.04.01. Passive trends, which may account for many apparent instances of Cope's rule, arise through random drift in trait space away from a boundary. The ZFEL (zero-force evolutionary law) describes the expected increase in variance under drift in the absence of selection. Understanding the contribution of drift to macroevolutionary patterns is essential for distinguishing driven from passive trends.
Coevolution 19.13.01. The Red Queen hypothesis is fundamentally a coevolutionary concept: species evolve not in isolation but in response to the evolution of other species. Host-parasite arms races, predator-prey dynamics, and competitive interactions are the engines of Red Queen evolution. The coevolutionary processes described in 19.13.01 provide the mechanistic basis for the macroevolutionary pattern of constant extinction probability.
Phylogenetics 19.07.01. Phylogenetic comparative methods are essential for testing macroevolutionary trend hypotheses. Ancestral state reconstruction, phylogenetic independent contrasts, and models of trait evolution (Brownian motion, Ornstein-Uhlenbeck, directional trend models) allow quantitative tests of whether traits show directional change, stasis, or random walk dynamics on phylogenetic trees calibrated with fossil dates.

Historical & philosophical context Master

The study of evolutionary trends has been shaped by a tension between two views of the fossil record: as a document of directional progress, or as a record of contingent, non-directional change.

Progress and the scala naturae. The pre-evolutionary concept of the Great Chain of Being (scala naturae) arranged organisms from simplest to most complex, implying a natural directionality to life's organization. Early evolutionary thought inherited this directional bias: Darwin's Origin of Species described evolution as "descent with modification" producing "endless forms most beautiful," and many early evolutionary paleontologists interpreted the fossil record as documenting progressive improvement. Cope's rule, named in the late 19th century, was initially understood in this progressive framework: larger body size was seen as "better," and the trend toward larger size was seen as evidence of evolutionary progress.

Gould's critique of progress. Stephen Jay Gould mounted the most influential challenge to the notion of evolutionary progress. In Full House (1996, also published as Life's Grandeur), he argued that the apparent trend toward increasing complexity is a statistical artifact of life's origin at a minimal level of complexity. The "wall" of minimal complexity (below which life cannot exist) constrains the left tail of the complexity distribution, while the right tail expands freely by random drift. The resulting increase in maximum complexity is not evidence of directionality but of the inevitable expansion of variance from a bounded starting point. Gould's favorite example was the disappearance of .400 hitting in baseball: the best hitters do not get worse, but the entire distribution of batting averages narrows as the worst players are eliminated and the overall quality of play improves. The right tail shrinks not because excellence declines but because the entire distribution contracts. By analogy, the left tail of complexity is constrained, and the right tail expands -- not because of a driving force but because of passive diffusion.

The Red Queen and Van Valen. Leigh Van Valen's 1973 paper "A New Evolutionary Law" was rejected by multiple journals before being published in his own journal Evolutionary Theory. The paper's central observation -- constant extinction probability -- was based on meticulous compilation of fossil duration data, but its proposed explanation (the Red Queen hypothesis) was initially controversial because it implied that adaptation does not reduce extinction risk. The hypothesis has since become a cornerstone of evolutionary ecology, inspiring the development of coevolutionary theory (the "Red Queen effect" in host-parasite dynamics, where sexual reproduction is favored because it generates novel genotypes that escape coevolving parasites), community ecology (the "zero-sum" game of ecological replacement), and macroevolutionary theory (the relationship between biotic interactions and extinction risk).

Punctuated equilibrium and the tempo of evolution. The punctuated equilibrium debate of the 1970s-1990s was one of the most heated in evolutionary biology. Eldredge and Gould's 1972 paper challenged the gradualist assumption of the modern synthesis by arguing that the fossil record predominantly shows stasis, not gradual change. Critics (Dawkins, 1986; Gingerich, 1983) argued that punctuated equilibrium was compatible with gradual microevolutionary change (the "punctuations" are rapid in geological terms but gradual in generational terms) and that the debate was partly semantic. The controversy was productive in forcing paleontologists to develop rigorous statistical methods for quantifying stasis and change, and in demonstrating that the fossil record is not merely an imperfect archive but a source of unique evolutionary data that cannot be obtained from the study of living organisms alone.

Contingency and Wonderful Life. Gould's Wonderful Life (1989) used the Burgess Shale fauna to argue for the primacy of contingency in evolutionary history. The Burgess Shale, a Middle Cambrian deposit preserving soft-bodied organisms in extraordinary detail, documents a richness of anatomical disparity that far exceeds the disparity of modern marine animals. Gould argued that many Burgess Shale organisms represented extinct body plans that were eliminated not by competitive inferiority but by the luck of the draw during the Cambrian-Ordovician transition. If the tape were replayed, different body plans would survive, and the history of life would take a radically different course. Conway Morris (1998) challenged this interpretation, arguing that the Burgess Shale organisms are less disparate than Gould claimed and that most can be accommodated within modern phyla. The debate remains unresolved in its details but has been enormously productive in stimulating research into the nature of disparity, the causes of the Cambrian radiation, and the role of contingency in macroevolution.

Dollo's law and irreversibility. Louis Dollo (1893) formulated his principle of irreversibility based on the study of fossil vertebrates, arguing that evolution cannot retrace its steps. The law has been refined over the subsequent century: it is now understood as a probabilistic statement about the improbability of exact reversal, rather than an absolute prohibition. Molecular studies have identified apparent exceptions (the re-evolution of lost traits in some lizard lineages, the regain of complex life cycles in some parasites), but these exceptions typically involve re-evolution through different genetic and developmental mechanisms, supporting the spirit of Dollo's law even when the letter is violated. The law's significance for macroevolutionary theory is that it demonstrates the path-dependence of evolution: where a lineage can go depends on where it has been, and lost possibilities are not easily regained.

Bibliography Master

Alroy, J. "Cope's rule and the dynamics of body mass evolution in North American fossil mammals." Science 280 (1998) 731-734.
Bookstein, F. L. "Random walk and the existence of evolutionary rates." Paleobiology 13 (1987) 446-464.
Dollo, L. "Les lois de l'evolution." Bull. Soc. Belge Geol. Paleontol. Hydrol. 7 (1893) 164-166.
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.
Gingerich, P. D. "Rates of evolution: effects of time and temporal scaling." Science 222 (1983) 159-161.
Gould, S. J. Wonderful Life: The Burgess Shale and the Nature of History (W. W. Norton, 1989).
Gould, S. J. Full House: The Spread of Excellence from Plato to Darwin (Harmony Books, 1996).
Hone, D. W. E., Dyke, G. J., Haden, M. & Benton, M. J. "Body size evolution in dinosaurs: not so simple." J. Evol. Biol. 18 (2005) 1317-1320.
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.
Jablonski, D. "Body-size evolution in Cretaceous molluscs and the status of Cope's rule." Nature 385 (1997) 250-252.
McShea, D. W. "Mechanisms of large-scale evolutionary trends." Evolution 48 (1994) 1747-1763.
McShea, D. W. & Brandon, R. N. Biology's First Law: The Tendency for Diversity and Complexity to Increase in Evolutionary Systems (University of Chicago Press, 2010).
Van Valen, L. "A new evolutionary law." Evol. Theory 1 (1973) 1-30.
Conway Morris, S. The Crucible of Creation: The Burgess Shale and the Rise of Animals (Oxford University Press, 1998).