Concept

The evidence for common descent

T-020Home BU-104Threads evolution

Statement

Homology, fossils and molecules point to shared ancestry.

Why it matters

natural-selection explains the mechanism by which populations change over time; evidence-common-descent addresses a separate, historical question — whether all present-day organisms actually share ancestry — and shows that multiple, entirely independent lines of evidence converge on the same answer. This convergence is what elevates common descent from a hypothesis to one of the best-supported conclusions in biology, and it is the shared ancestral framework within which fitness-adaptation, speciation, and sexual-selection all operate.

Understanding how these evidence lines work together, rather than any one alone, is also what protects the result against a common objection: no single line of evidence (fossils, or anatomy, or molecules) is individually claimed to be proof; it is the independent agreement of several very different kinds of data that makes the inference robust.

Hypotheses

If two lineages share a common ancestor, they are expected to retain shared derived similarities (homologies) not fully explained by shared function alone.This is the logical basis for reading anatomical and molecular similarity as evidence of ancestry: a shared underlying structure inherited from a common ancestor, then modified differently in each descendant lineage, is a different kind of similarity from two unrelated lineages independently evolving a similar solution to a similar problem (analogy). Molecular sequence divergence between lineages accumulates, on average, with time since their common ancestor.This average relationship holds only approximately and varies by gene and lineage (some genes evolve faster than others, and rates can differ between lineages under different selective pressures); it supports the general nested-hierarchy pattern used here, but is not treated as delivering exact dates without additional calibration.

Proof

\[ \text{Homologous structures (e.g. the vertebrate pentadactyl limb) share the same underlying skeletal plan across species despite serving very different functions.} \]

A human hand, a whale flipper, a bat wing, and a horse leg all share the identical basic bone arrangement, modified for grasping, swimming, flying, and running respectively — consistent with descent from a common ancestor whose limb plan was subsequently modified along each separate lineage, and not what independent, unrelated origins for each limb would be expected to produce. A

\[ \text{The fossil record shows a temporally ordered succession of forms, including transitional fossils with intermediate character combinations.} \]

Fossils such as Tiktaalik (intermediate between fish and early tetrapods, combining fish-like scales and gills with limb-like fin bones and a mobile neck) occur in rock strata of intermediate geological age, exactly where a gradually branching lineage predicts an intermediate form should be found, both anatomically and temporally. A

\[ \text{Vestigial structures are reduced, non-functional or minimally functional homologs of structures that are fully functional in related species.} \]

Whale pelvic bones, reduced and disconnected from any hind-limb function, are exactly what a lineage descended from four-limbed, walking ancestors and only later returned to an aquatic, limbless-hindquarters lifestyle would be expected to retain in modified, reduced form — not a feature an independently, purpose-built aquatic body plan would be expected to include at all. A

\[ \text{Molecular sequence comparisons across species produce a nested hierarchy of similarity that matches the nested hierarchy independently inferred from morphology and the fossil record.} \]

The near-universality of the genetic code and of core biochemical machinery (ribosomes, ATP synthase, the central glycolytic pathway) across all three domains of life indicates single common ancestry for life as a whole, while finer-grained DNA and protein sequence comparisons among more closely related species reproduce, largely independently, the same branching pattern already inferred from bones and fossils. B

\[ \text{Biogeographic distribution: island species resemble their nearest mainland relatives more closely than they resemble unrelated species occupying a similar ecological niche elsewhere.} \]

This matches descent-with-modification from a colonising mainland ancestor, subsequently adapted locally on the island, rather than each island species having originated independently and simply converged on whatever form suits the local environment, which would instead predict similarity to other same-niche species regardless of geographic proximity. A

Result

\[ \text{Homology + fossil succession + vestigial structures + molecular nested hierarchy + biogeography} \Rightarrow \text{common descent} \]

Reading. Several structurally very different kinds of evidence — anatomical, palaeontological, molecular, and geographic — independently converge on the same branching pattern of relatedness among species, which is exactly the pattern predicted if all life shares common ancestry.

Scope. A historical, evidentiary case rather than a single derivable formula; its strength rests specifically on the independent convergence of multiple distinct data types onto one consistent tree, not on any single line of evidence alone (Discussion).

Corollaries & converses

speciation is the process that generates the branch points this result's nested hierarchy reflects; common descent is the historical record of speciation events accumulated over the whole of life's history.
fitness-adaptation and sexual-selection operate within and along the branches of the shared ancestral tree this result establishes, shaping how each lineage diverges from its common ancestor with others.
Because analogous (convergently evolved) structures do not share the same underlying homologous plan, finding two similar-function structures built on entirely different underlying architecture (e.g. an insect wing versus a bird wing) is itself evidence against a close common ancestry specifically for that structure, even between otherwise related lineages.

Fails without

If homologies actually reflected only convergent function rather than shared ancestry (Hypotheses): unrelated lineages solving the same functional problem should then be expected to converge on similar underlying structure as well as similar function; instead, structurally very different solutions to the identical function (e.g. the entirely different underlying architectures of an insect wing, a bird wing, and a bat wing, all used for flight) are routinely observed, which the ancestry-based, not purely functional, explanation of homology accounts for directly.
If molecular and morphological/fossil-based hierarchies disagreed substantially: that would undermine the case for a single consistent evolutionary tree; instead, sequence-based trees, built from data types entirely unavailable to and unanticipated by Darwin, independently reproduce the same branching pattern already inferred from anatomy and the fossil record, a strong corroboration this result depends on.

Common errors

Confusing homology (shared structure from common ancestry, Step 1) with analogy (shared function from convergent, independent evolution, e.g. insect and bird wings) — the two are frequently conflated but support opposite inferences about ancestry.
Assuming the theory of common descent requires an unbroken, gap-free fossil record; it does not — fossilisation is a rare event, and the theory predicts a generally nested pattern with intermediate forms occasionally preserved (Step 2), not complete preservation of every transitional individual.
Describing humans as having "evolved from" modern apes, rather than humans and modern apes sharing a common ancestor species that no longer exists, with both lineages having since diverged and changed from it.
Treating vestigial structures as necessarily entirely useless; many retain a reduced or secondary function (Step 3), and "vestigial" specifically means reduced or repurposed relative to the fully functional homologous structure in related species, not functionless by definition.

Discussion

Charles Darwin's 1859 On the Origin of Species assembled homology, biogeography, and the (then very incomplete) fossil record into the first systematic case for common descent, well before the mechanisms of heredity or the existence of DNA were known at all. The molecular evidence of Step 4 became available only much later, once protein and then DNA sequencing matured through the mid-to-late 20th century, and its independent agreement with the pre-existing anatomical and fossil-based tree is one of the strongest corroborations the theory has received since Darwin's original argument.

Because sequence divergence accumulates roughly with time but at rates that vary somewhat by gene and lineage, molecular trees are calibrated against independently dated fossils where possible, rather than assumed to give absolute dates from sequence divergence alone; the underlying nested-hierarchy pattern used as evidence here does not depend on this calibration being exact.

Common misconception: that evolutionary theory claims humans descended directly from present-day chimpanzees or other living apes. All the evidence instead points to humans and modern great apes sharing a common ancestral species, now extinct, from which both lineages have since diverged and independently accumulated their own changes.

Worked examples

\[ \text{Compare the forelimb skeleton of a human, bat, whale, and horse.} \]

All four share the identical underlying bone sequence (one proximal bone, two distal bones, then wrist/hand bones), despite being used respectively for manipulation, flight, swimming, and running — the same homologous plan, modified differently along each lineage since their shared ancestor, exactly as Step 1 of the Proof describes. A

\[ \text{Compare cytochrome c amino-acid sequence divergence between humans and several other species.} \]

Species judged more closely related on independent morphological/fossil grounds (e.g. other mammals) show fewer amino-acid differences in this protein from the human sequence than more distantly related species (e.g. yeast), reproducing, from molecular data alone, essentially the same relatedness ordering already established from morphology (Step 4). A

\[ \text{Anatomical tree} \approx \text{molecular tree} \Rightarrow \text{independent corroboration of common descent} \]

Reading. Two entirely different kinds of data — skeletal anatomy and protein sequence — independently produce essentially the same relatedness ordering among species, which is precisely the outcome common descent predicts and an independent-origins alternative would not.

Scope. The same logic (compare an independently derived tree against another data type) generalises to any pair of independent evidence lines among those listed in the Proof.

Further example. A researcher compares the amino-acid sequences of the protein cytochrome c in two species, Species A and Species B. The protein consists of 104 amino-acid residues. In the alignment, 12 positions show different amino acids between the two species. Assuming a molecular clock model where the rate of amino-acid substitution is constant at \( r = 2.0 \times 10^{-9} \) substitutions per site per year, estimate the time since the most recent common ancestor (TMRCA) of Species A and Species B. (Note: The observed difference represents the total divergence accumulated along both lineages since they split.)

\[ \text{Calculate the observed fraction of differing sites (p).} \]

The number of differences is 12 out of 104 total sites. The fraction of sites that differ, \( p \), is the ratio of differences to total sites. This is the raw data input for the molecular clock calculation. A

\[ p = \frac{12}{104} \approx 0.1154 \]

This value represents the total evolutionary distance between the two sequences. Since the sequences diverged from a common ancestor, this distance is the sum of the changes accumulated in Species A's lineage plus the changes accumulated in Species B's lineage. A

\[ \text{Relate total divergence to time (T) and rate (r).} \]

Under the molecular clock hypothesis, the rate of substitution \( r \) is defined per lineage per year. If \( T \) is the time since divergence, the total expected number of substitutions per site is \( 2rT \) (one lineage of length \( T \) plus the other lineage of length \( T \)). Thus, \( p \approx 2rT \). A

\[ T = \frac{p}{2r} \]

Rearranging the equation from Step 3 to solve for time \( T \). We divide the total observed divergence by the combined rate of accumulation for both lineages. A

\[ T = \frac{0.1154}{2 \times (2.0 \times 10^{-9})} \]

Substituting the values: \( p \approx 0.1154 \) and \( r = 2.0 \times 10^{-9} \). The denominator is \( 4.0 \times 10^{-9} \). A

\[ T \approx 2.885 \times 10^{7} \text{ years} \]

Performing the division: \( 0.1154 / 4.0 \times 10^{-9} = 0.02885 \times 10^{9} = 2.885 \times 10^{7} \). Rounding to three significant figures, the estimated time is 28.9 million years. A

\[ T \approx 2.89 \times 10^{7} \text{ years} \]

Reading. The molecular data suggests that Species A and Species B diverged approximately 28.9 million years ago. This quantitative estimate provides a specific timeline for their common descent, which can be cross-referenced with the fossil record (e.g., looking for transitional fossils from that era) to test the consistency of molecular and morphological evidence.

Scope. This calculation assumes a constant substitution rate (strict molecular clock) and that the observed differences are a direct proxy for the number of substitutions (valid for small divergence values where multiple hits at the same site are rare). For larger divergences, more complex models (like Jukes-Cantor) would be required to correct for multiple substitutions at the same site.

Problems

A student observes that both dolphin fins and shark fins are used for swimming and have a broadly similar external shape, and concludes they are homologous. Evaluate this claim using the Result.
Solution
External similarity in shape and function alone is not sufficient to establish homology; the relevant test (Hypotheses) is shared underlying structure inherited from a common ancestor. Dolphin fins contain the modified pentadactyl limb bones shared with other mammals (Step 1), while shark fins have an entirely different, cartilaginous internal skeletal structure with no such shared plan. The similarity is functional convergence (analogy, both adapted for aquatic locomotion), not homology; dolphins and sharks do not share a recent common ancestor with fins.
Explain why finding a fossil with a mix of reptilian and mammalian features, dated to a geological period between the known reptile and mammal fossil ranges, would count as strong supporting evidence for common descent (Step 2), rather than being explainable equally well by independent origins.
Solution
Common descent specifically predicts that transitional lineages should show a combination of ancestral and derived traits and should occur in strata of intermediate geological age between the two groups being connected — both the anatomical mixture and its temporal placement are specific, falsifiable predictions. Independent-origins models make no such prediction about either the traits or the timing; finding a fossil that satisfies both constraints simultaneously is therefore evidence specifically for the branching, common-ancestry explanation.
Two unrelated desert plant lineages, one a cactus and one a distantly related succulent from another continent, independently evolved thick, water-storing stems and reduced leaves. Explain why this observation does not weaken the case for common descent.
Solution
This is convergent evolution (analogy) — two lineages independently arriving at a similar functional solution (water storage) to a similar selective problem (arid environment), without sharing a recent common ancestor for that specific trait. Common descent predicts homology among genuinely related structures, not that every similar-function trait across all of life must be homologous; convergence is an expected, well-documented outcome under selection acting independently on unrelated lineages facing similar environmental pressures, and does not contradict the far larger body of homology-based evidence.
Humans and chimpanzees share thousands of endogenous retroviruses (ERVs) inserted at exactly the same positions in their genomes. Retroviruses integrate at essentially random sites. Explain why shared insertion sites are strong evidence of common descent, and estimate the chance that a single ERV would land at the same one of roughly \(10^{8}\) possible sites independently in both lineages.
Solution
An ERV is a retroviral genome that integrated into a germ-line cell and was then inherited like any other DNA. If two species carry the same ERV at the same site, the simplest explanation is a single integration in their common ancestor, inherited by both descendant lineages. For two independent integrations to hit the same site by chance, with about \(10^{8}\) possible sites, the probability is about \(1/10^{8} = 10^{-8}\) per insertion; for thousands of shared insertions it is effectively zero. Moreover, the shared ERVs are distributed among primates in a nested pattern (some shared by all apes, fewer by human and chimp alone) that matches the tree built from other data, exactly as descent with modification predicts.
Two species, X and Y, are compared for their DNA sequences in a gene of length 500 base pairs. They differ at 25 positions. If the mutation rate for this gene is known to be \( 1.0 \times 10^{-8} \) mutations per base pair per year, and assuming a molecular clock where the rate is constant in both lineages, calculate the time since their most recent common ancestor. Express your answer in millions of years (Ma).
Solution
1. Calculate the fraction of differing sites (\( p \)): \[ p = \frac{25}{500} = 0.05 \] 2. The total divergence is the sum of mutations in both lineages. Let \( T \) be the time since divergence. The total expected mutations per site is \( 2 \times r \times T \), where \( r \) is the rate per lineage per year. \[ p = 2rT \] 3. Solve for \( T \): \[ T = \frac{p}{2r} \] 4. Substitute the values: \[ T = \frac{0.05}{2 \times (1.0 \times 10^{-8})} \] \[ T = \frac{0.05}{2.0 \times 10^{-8}} \] \[ T = 0.025 \times 10^{8} \] \[ T = 2.5 \times 10^{6} \text{ years} \] 5. Convert to millions of years: \[ T = 2.5 \text{ Ma} \] The time since the most recent common ancestor is 2.5 million years.