Evolution II: The History and Diversity of Life, Coevolution, Molecular Evolution, Phylogenetics and Macroevolution

The Earth formed about 4.54 billion years ago. Miller and Urey’s 1953 experiment showed that amino acids form when electric sparks pass through a mixture of methane, ammonia, hydrogen and water, supporting the idea of prebiotic chemistry; the RNA world hypothesis proposes that RNA, which can both store information and catalyse reactions (ribozymes), preceded DNA and proteins. The oldest widely accepted evidence of life is about 3.5 billion years old, in stromatolites built by microbial mats. Oxygenic photosynthesis by cyanobacteria produced the Great Oxidation Event about 2.4 billion years ago; eukaryotes appear in the fossil record by about 1.6–1.8 billion years ago, arising through the endosymbiosis that gave mitochondria and, later, chloroplasts.

The Cambrian explosion, beginning about 539 million years ago, saw most animal phyla appear in the fossil record within a geologically short time, preceded by the enigmatic soft-bodied Ediacaran biota. Palaeobiological evidence comes from fossils (body fossils, trace fossils, chemical fossils), from radiometric dating — the fraction of a parent isotope remaining after t is (1/2)^t/t½, so a rock retaining 25 % of its potassium-40 (half-life about 1.25 billion years) is two half-lives, 2.5 billion years, old — and from transitional forms such as Tiktaalik (between fish and tetrapods), Archaeopteryx (between dinosaurs and birds) and the early whales with legs.

Woese’s analysis of ribosomal RNA (1977, formalised 1990) divided life into three domains: Bacteria, Archaea and Eukarya, with Archaea more closely related to Eukarya than to Bacteria. Classical (Linnaean) taxonomy names organisms binomially and ranks them in a hierarchy — domain, kingdom, phylum, class, order, family, genus, species — grouping them by overall similarity. Systematics reconstructs evolutionary relationships, and cladistics (Hennig) groups organisms only by shared derived characters (synapomorphies), so every named group should be a clade — an ancestor and all its descendants.

Coevolution is reciprocal evolutionary change in interacting species. In mutualisms it can produce tight matching: each fig species is pollinated by its own fig wasp, which breeds in the fig’s flowers, and yucca moths actively pollinate yuccas and lay eggs in the ovary. In antagonistic interactions it produces arms races — prey defences met by predator counter-adaptations, toxic newts (tetrodotoxin) and resistant garter snakes — described by Van Valen’s Red Queen hypothesis: species must keep evolving just to keep their fitness relative to their enemies. Co-speciation is the parallel splitting of host and parasite (or mutualist) lineages, detected as congruent phylogenies, as in pocket gophers and their chewing lice.

Batesian mimicry is under negative frequency-dependent selection: as mimics become common, predators meet more palatable individuals with the pattern and learn that it is safe, so the mimic’s advantage falls. In host–parasite interactions, parasites that evolve faster (short generation times) often lead; hosts evolve resistance, parasites counter-resistance, and allele frequencies can cycle. The Red Queen view of sex argues that recombination produces novel host genotypes that parasites adapted to common genotypes cannot yet exploit. Virulence evolves under a trade-off: a parasite that exploits its host harder transmits faster but kills it sooner, so intermediate virulence often maximises transmission; the decline in virulence of myxoma virus in Australian rabbits after 1950 is the classic case.

Kimura’s neutral theory (1968) holds that most substitutions fixed at the molecular level are selectively neutral, fixed by drift rather than selection. A striking consequence: in a diploid population of N, 2Nμ new neutral mutations arise each generation, each with fixation probability 1/(2N), so the substitution rate k = 2Nμ × 1/(2N) = μ — the rate of neutral evolution equals the mutation rate and is independent of population size. This underlies the molecular clock (Zuckerkandl and Pauling): if substitutions accumulate at a steady rate, the genetic distance between two lineages grows in proportion to the time since they split. Because both lineages accumulate changes, d = 2μt and t = d/(2μ): two species differing at 4 % of sites with μ = 1 × 10⁻⁹ substitutions per site per year diverged 0.04/(2 × 10⁻⁹) = 2 × 10⁷ years, or 20 million years, ago. Ohta’s nearly neutral theory adds slightly deleterious mutations whose fate depends on Nₑs.

Rates of evolution differ across the genome: synonymous sites and pseudogenes evolve fastest, close to the mutation rate, and functionally constrained sites (histones, ribosomal proteins) slowest. The ratio ω = dN/dS of non-synonymous to synonymous substitution rates diagnoses selection on a protein-coding gene: ω < 1 indicates purifying selection, ω ≈ 1 neutrality and ω > 1 positive (diversifying) selection, as in the antigen-binding sites of MHC genes. Genome evolution proceeds by gene duplication (the globin family), whole-genome duplication and polyploidy — autopolyploidy within a species and allopolyploidy after hybridisation between species, as in bread wheat (hexaploid) — by chromosomal fusions and inversions, and by introgression, the movement of genes between species through hybridisation and backcrossing, such as the Neanderthal segments in the genomes of many modern humans.

The coalescent looks backwards in time: trace two gene copies sampled from a diploid population of constant size Nₑ back until they meet in a common ancestor. The probability that they coalesce in any one generation is 1/(2Nₑ), so the expected coalescence time is 2Nₑ generations; for a sample of n copies the expected time to the most recent common ancestor is 4Nₑ(1 − 1/n) generations, approaching 4Nₑ for large samples. Neutral diversity is set by the population mutation rate θ = 4Nₑμ, which is also the expected number of differences per site between two sequences: with Nₑ = 10 000 and μ = 2.5 × 10⁻⁸ per site per generation, θ = 0.001. The coalescent is why present-day diversity carries a record of past population size, and why a bottleneck leaves a lasting signature.

Support for a clade is measured by the bootstrap (the percentage of trees from resampled data that contain it) or by its Bayesian posterior probability. Tree space grows explosively: there are (2n − 5)!! unrooted and (2n − 3)!! rooted binary trees for n taxa, where k!! = 1 × 3 × 5 × … × k. Four taxa give 3 unrooted trees and five give 1 × 3 × 5 = 15 (and 105 rooted), which is why exhaustive search is possible only for small data sets. Under parsimony, a character with states 0, 1, 0, 1 in taxa A, B, C and D needs two changes on the tree ((A, B), (C, D)) but only one on ((A, C), (B, D)), so it supports the second tree. Phylogeography (Avise) studies the geographical distribution of genealogical lineages within and among closely related species, often using mitochondrial DNA and the coalescent, and reveals glacial refugia, colonisation routes and cryptic species.

Species concepts differ in what they emphasise. Mayr’s biological species concept defines species as groups of interbreeding natural populations reproductively isolated from other such groups — it cannot be applied to asexual organisms or fossils. The morphological concept uses diagnosable differences in form; the phylogenetic concept the smallest diagnosable monophyletic group; the ecological concept a lineage occupying a distinct adaptive zone. Reproductive isolation is prezygotic (habitat, temporal, behavioural, mechanical and gametic isolation) or postzygotic (hybrid inviability, hybrid sterility — the mule — and hybrid breakdown). Haldane’s rule: when one sex of an F₁ hybrid is absent, rare or sterile, it is the heterogametic sex (XY males in mammals, ZW females in birds and butterflies).

• Allopatric speciation: a geographical barrier splits a population (vicariance) or a small group colonises a new area (peripatric speciation); isolation, drift and divergent selection build reproductive isolation. The commonest mode.
• Parapatric speciation: divergence along an environmental gradient with limited gene flow across a narrow contact zone, as in grasses on mine tailings.
• Sympatric speciation: without geographical separation — instantly by polyploidy, common in plants, or through host shifts, as in the apple maggot fly Rhagoletis pomonella, and possibly the crater-lake cichlids.
• Reinforcement: selection against unfit hybrids strengthens prezygotic isolation where incipient species meet.

Adaptive radiation is the rapid diversification of a lineage into many species occupying different niches, usually after colonising a new area or after an extinction frees ecological space: Darwin’s finches of the Galápagos, the cichlid fishes of the East African lakes, the Hawaiian silverswords and honeycreepers, and the mammals after the end-Cretaceous extinction. Convergent evolution produces similar traits in unrelated lineages under similar selection — the streamlined bodies of sharks, ichthyosaurs and dolphins; the placental and marsupial "wolves", "moles" and "mice"; the succulent stems of American cacti and African euphorbias. Extinction is continual at a background rate, punctuated by the mass extinctions; the fossil record shows that most species that ever lived are extinct. Tempo is debated: phyletic gradualism against punctuated equilibrium (Eldredge and Gould), long stasis interrupted by rapid change at speciation.

Biogeographic processes explain where lineages are: dispersal (crossing a barrier that already exists) and vicariance (a barrier arising within a once-continuous range). Wegener proposed continental drift in 1912; plate tectonics later gave it a mechanism. The supercontinent Pangaea split into Laurasia in the north and Gondwana in the south; Gondwana later broke into South America, Africa, Antarctica, Australia, Madagascar and India. The distribution of the seed fern Glossopteris and of the reptiles Mesosaurus and Lystrosaurus across these now-separate continents was early evidence. India drifted north and collided with Asia about 50 million years ago, raising the Himalaya; its Gondwanan heritage survives in groups such as caecilians and some frogs of the Western Ghats, while the collision let Asian and Malayan lineages flood in.