11.1 Classification and Cladistics

The principal ranks are domain, kingdom, phylum, class, order, family, genus and species. Moving downward narrows membership: every organism in one genus belongs to the same family, but a family normally contains several genera. A taxon is any named group at one of these ranks. The hierarchy is useful for storing and retrieving knowledge, but its rank labels are human conventions. Evolution produces branching lineages, not eight naturally spaced shelves.

The biological species concept cannot be applied directly to fossils, asexual organisms or populations that never meet in nature. Taxonomists then combine morphology, ecology, biochemical traits and sequence evidence, choosing criteria appropriate to the organisms and evidence available.

Natural classification groups organisms according to descent from common ancestors. It is predictive because related organisms inherit many features from shared ancestors: once an unfamiliar organism is placed reliably in a group, some unobserved characteristics can be predicted. Homologous features are especially useful because their similarity arose from shared ancestry. Superficial resemblance caused by similar selection pressures can instead produce an unnatural grouping.

The three-domain system separates Bacteria, Archaea and Eukarya. It was developed from comparisons including ribosomal RNA sequences, membrane chemistry and responses to antibiotics. Ribosomal RNA is present in all cellular life, performs a conserved role in translation and contains regions that can be compared across very distant lineages. The evidence showed that prokaryotic cells do not form one uniform evolutionary group: Archaea and Bacteria represent deeply separated lineages, with Archaea sharing a more recent common ancestry with Eukarya than traditional appearance-based schemes suggested.

Classification changes when evidence changes. This is a strength, not a failure. A group based on a convenient visible character may exclude descendants of its own ancestor or combine unrelated lineages. Molecular data have repeatedly divided traditional plant and animal groups into separate clades. Morphology remains valuable, especially for fossils and field identification, but DNA and protein sequences provide many independently comparable characters and can reveal homology where appearance has diverged.

A dichotomous key identifies an unknown specimen through a sequence of paired alternatives. Each couplet should use observable, mutually exclusive features and send the user either to another couplet or to one identification. Good choices divide the remaining specimens efficiently and avoid relative words such as large or small unless a measurable threshold is supplied. The same specimen must not satisfy both statements in a pair.

A key is not necessarily a phylogeny. Its branching order is chosen for reliable identification, so the first split might use wings because wings are conspicuous, even if winged organisms do not form one clade. When constructing a key from images, use features actually visible in every specimen. When testing it, follow each route independently and check that every endpoint is unique.

A clade contains an ancestor and all of its descendants. A cladogram represents hypotheses about the order in which lineages diverged. The root represents the ancestral lineage for all taxa shown; each node is a common ancestor where one lineage split; terminal branches are the sampled taxa. Two taxa that share the most recent node are sister taxa. Relatedness depends on recency of common ancestry, not on how close labels appear on the page.

Branches can be rotated around a node without changing relationships. A ladder-shaped tree and a radial tree can therefore encode the same topology. Unless a scale is provided, horizontal branch length and the vertical order of terminal names do not represent time or amount of change. Nor should a living species at one branch tip be called the ancestor of another living tip: both descended from an ancestral population represented by an internal node.

DNA base sequences or amino-acid sequences from homologous molecules can be aligned across species. Fewer differences usually indicate more recent divergence, because mutations accumulate after lineages separate. Cytochrome c and ribosomal molecules are useful across broad groups because their functions are widespread. Rapidly changing non-coding regions may discriminate recent divergence, while strongly conserved sequences can preserve a readable signal across ancient splits.

A molecular clock assumes that substitutions accumulate at an approximately predictable rate. If the rate has been calibrated using fossils or dated geological separations, sequence distance can estimate divergence time. The estimate has uncertainty: mutation rates differ among genes, lineages and genomic regions; natural selection removes many changes from functional sequences; and multiple substitutions can occur at the same site. A correlation between distance and time is therefore a model, not an exact universal clock.

Protein comparisons can conceal some DNA change because the genetic code is degenerate: different codons may specify the same amino acid. A synonymous substitution changes a base without changing the polypeptide. Conversely, protein structure may be more directly informative about functional conservation. Robust phylogenies compare multiple sequences and, where possible, combine molecular evidence with morphology, development, biogeography and fossils.