A3.2 Classification and cladistics

About two million species have been named and described, and many more remain undiscovered: estimates of the total run to several million or more. Without a system for sorting this diversity, information about organisms would be almost impossible to organise or find. Classification is the process of placing organisms into groups.

Once organisms have been classified, a wide range of further study becomes possible:

• Storing and finding information. Knowledge about each species is filed under its group, so biologists can retrieve what is already known about it and its relatives.
• Identification. An unfamiliar organism can be assigned to a large group (for example, flowering plants) and then to smaller and smaller groups until its species is found, often using a dichotomous key. Each species has a two-part scientific name used worldwide (see A3.1 Diversity of organisms).
• Prediction. Members of a group share many characteristics, so the features of a poorly studied species can be predicted from better-known members of its group.
• Studying evolution. Groups reflect shared ancestry, so a classification suggests hypotheses about how and when species evolved.
• Conservation. Knowing how many distinct species and lineages live in an area, and how unusual they are, helps set priorities for protecting biodiversity.

A taxon (plural taxa) is any group used in classification, and taxonomy is the science of naming organisms and placing them in taxa. In the traditional system each taxon has a rank in a fixed hierarchy: species are grouped into genera, genera into families, and so on up to kingdoms. Moving up the hierarchy, each taxon contains more species that share fewer characteristics.

Why the hierarchy does not fit evolution well

• Divergence is gradual. Lineages split and then drift apart slowly, so there is no objective point at which a group becomes different enough to be a separate genus, family or order. Two taxonomists can agree on which species belong together yet disagree on the rank to give the group.
• Ranks are not comparable. A “family” of insects and a “family” of mammals may differ hugely in age and in the number of species, so ranks say little about how much divergence has happened.
• Species boundaries can be unclear, for example where populations are partly interbreeding or reproduce asexually.
• Some traditional taxa do not match the branching of evolution. The class Reptilia excludes birds, which are placed in their own class, Aves, even though crocodiles are more closely related to birds than to lizards. Grouping all “fish” together ignores the fact that a salmon is more closely related to humans than to a shark.

Biologists now broadly aim for a classification that mirrors evolutionary relationships, so that each group is a real branch of the tree of life. For that, a group must pass two tests:

1. All the species that evolved from the group’s common ancestor are included in it.
2. Only species that evolved from that common ancestor are included.

A group that meets both conditions is a clade (also called a monophyletic group). “Reptiles” in the traditional sense fail condition 1, because birds evolved from the same common ancestor but are left out.

The advantage: prediction

Because every member of a clade inherited characteristics from the same ancestor, the characteristics of one member can be predicted from the others with reasonable confidence:

• A newly discovered species of mammal can be expected to have hair, mammary glands, three middle-ear bones and a diaphragm, before anyone has studied it in detail.
• The anticancer drug paclitaxel (Taxol) was first isolated from the bark of the Pacific yew (Taxus brevifolia). Related yew species were then investigated, and a closely related compound found in the needles of the European yew (Taxus baccata) can be converted into the drug, which does not require trees to be felled. Searching relatives of a useful species is effective because related species tend to make similar chemicals.

Classifications that do not follow evolutionary relationships make such predictions unreliable, because the members of a group may resemble each other only superficially.

A clade is a group of organisms that have evolved from a common ancestor, including the ancestral species and all of its descendants, whether living or extinct. Because of their common ancestry, members of a clade share characteristics.

Clades are nested: every species belongs to many clades of different sizes. A human is in the mammal clade, which is nested inside the amniote clade, inside the tetrapods, inside the bony vertebrates, and so on. A clade can contain millions of species or very few: Ginkgo biloba is the only living member of a clade that was once diverse.

Evidence for placing organisms in a clade

The most objective evidence comes from base sequences of genes and amino acid sequences of proteins:

• Sequences provide a very large number of characters (every base or amino acid position) that can be compared directly and counted.
• Each character has discrete states (A, C, G or T), so comparisons depend much less on personal judgement than descriptions of shape do.
• Homologous genes, such as those for ribosomal RNA or cytochrome c, are found in very many organisms, so even distantly related groups can be compared.

Morphological traits (structure and form) can also be used to assign organisms to clades, and are the only option for most extinct species known from fossils. However, structures can look alike because of convergent evolution in unrelated lineages (the wings of bats, birds and insects, for example), so morphology is more likely to mislead. Only homologous characteristics, inherited from a common ancestor, indicate membership of a clade.

A cladogram is a branching diagram that shows the probable sequence in which a group of organisms diverged from common ancestors. To analyse one you need four terms:

• Root: the base of the cladogram, representing the common ancestor of every organism on it.
• Node: a branching point. It represents a hypothetical common ancestor, an ancestral species that split into two lineages (occasionally more). It is “hypothetical” because the ancestor is inferred, not known as a fossil.
• Terminal branch: a branch ending at the tips, leading to a species or to a group not subdivided on the diagram.
• Clade: any node together with everything that branches from it.

Deducing relationships

1. To find the common ancestor of two species, trace their branches back from the tips until they meet. That node is their most recent common ancestor.
2. Two species whose branches meet at a node further from the root shared a common ancestor more recently, so they are more closely related.
3. To find a clade, choose a node and include every terminal branch that comes off it. Clades can be nested inside larger clades.

The order of names along the tips carries no information. A node can be rotated, swapping its two branches like a hanging mobile, without changing any relationship. Two cladograms that look different may show exactly the same hypothesis.

Some cladograms add numbers to show sequence differences, and some are drawn to scale, with branch lengths proportional to the estimated time since each split. Unless a scale is given, branch length means nothing.

Cladograms are hypotheses

A cladogram is a hypothesis about evolutionary relationships, based on the evidence available. It can be tested and may be revised when new sequences or fossils are found. When cladograms constructed independently from several different genes show the same branching pattern, the hypothesis is much stronger, but it is still not proof.

After two lineages split from a common ancestor, mutations accumulate independently in each of them. Over long periods, the differences between their base sequences, and therefore between the amino acid sequences of their proteins, build up gradually. If differences accumulate at a roughly constant rate, the number of differences can be used to estimate the time since the two clades diverged. This is the molecular clock: the more differences, the longer ago the split.

Calibrating and using the clock

The rate has to be calibrated using a split whose date is known independently, usually from fossils that can be dated by radiometric methods.

• Calibration: two species known from fossils to have diverged 20 million years ago differ at 60 bases in a gene, so the rate is 60 ÷ 20 = 3 differences per million years.
• Estimate: two other species differ at 21 bases in the same gene, so they diverged about 21 ÷ 3 = 7 million years ago.

Both lineages accumulate changes after the split, so each lineage changes at half the rate measured between the two species. For example, human and chimpanzee genomes differ at roughly 1.2% of aligned bases; at a rate of about 1 × 10⁻⁹ changes per base per year in each lineage, this gives a split around 6 million years ago, which fits the fossil record. Slower rates measured directly in families would place it earlier, which shows how sensitive the estimates are to the rate assumed.

Why it gives only estimates

The rate of change is not truly constant. It is affected by:

• Generation time: species with short generations copy their DNA more times per year, so they tend to accumulate changes faster.
• Population size: in small populations, chance (genetic drift) plays a larger part in which new mutations spread.
• Intensity of selection: in genes whose function is essential, most changes are harmful and removed by natural selection (see D4.1 Natural selection), so these genes change slowly; other genes change faster.
• Other factors, such as the efficiency of DNA repair, and the same base changing more than once over very long periods.

Each gene therefore has its own clock rate. Slowly changing sequences such as rRNA are used to compare very distantly related groups; fast-changing sequences such as animal mitochondrial DNA are used for recent splits.

Base sequences of genes and amino acid sequences of proteins are the main data used to construct cladograms. With a small set of sample data you can do it by hand:

1. Align the sequences of the same gene so that homologous positions line up in columns.
2. Identify the positions that vary between species.
3. Count the differences between each pair of species and record them in a table.
4. Group the most similar pair first, joining them at a node.
5. Add the next most similar species or group at the next node down, and continue until all are included.
6. The most different species branches off nearest the root.

Parsimony

The same species could be linked by several different branching patterns. Parsimony analysis compares them and prefers the one that needs the fewest mutations to explain every difference in the data, because a shorter route is judged more likely than a longer one. In the sample data, pairing P with Q (and so R with S) needs 10 base changes, whereas grouping P with R, or P with S, needs 12, so P and Q are paired. Parsimony compares branching patterns only: it does not say where the root goes. The root is placed using the most different species, S, as an outgroup. With real data sets of many species and thousands of bases, computer software compares the possible trees.

Traditional classifications were mostly based on morphology. As sequence data have become available for more and more species, cladistics has been used to check whether these groups correspond to clades. There are three possible outcomes:

• The traditional group is a clade: the classification is confirmed, as it has been in many cases.
• The group contains species that do not all share one common ancestor exclusive to the group, so it is split and species are moved.
• Species that share a common ancestor have been placed in different groups, so groups are merged or species are moved together.

Case study: the figwort family (Scrophulariaceae)

The figwort family was once one of the largest families of flowering plants, with several thousand species, grouped mainly on the structure of their flowers: typically tubular, two-lipped and bilaterally symmetrical. From the 1990s, cladistic analysis of chloroplast gene sequences showed that the family was made up of several separate clades. As a result:

• snapdragons, foxgloves and speedwells were moved to the plantain family (Plantaginaceae), even though plantains have tiny wind-pollinated flowers that look nothing like theirs;
• partly parasitic species such as eyebrights and yellow rattle were moved to the broomrape family (Orobanchaceae), with fully parasitic relatives;
• monkeyflowers were moved to a different family, and some new families were created;
• a few genera, such as the butterfly bush (Buddleja), were moved into the figwort family.

Fewer than half of the species remain in the figwort family. You do not need to memorise these details.

Cladistics has also shown that the classes Reptilia and Aves do not match evolutionary relationships: birds are a clade nested within the reptiles, closest to the crocodilians.

Traditional classification divided organisms into two kinds of cell, prokaryotes and eukaryotes, and in the five-kingdom system all prokaryotes were placed in a single kingdom. In 1977 Carl Woese and George Fox compared the base sequences of ribosomal RNA (rRNA) from the small subunit of the ribosome in a wide range of organisms. They found that prokaryotes fall into two groups that are as different from each other as either is from eukaryotes. Methane-producing prokaryotes, now called archaea, were the first members of the second group to be recognised.

Woese and Fox called these groups “urkingdoms”. Their work overturned the old prokaryote–eukaryote split and, in 1990, Woese and colleagues proposed a new level above kingdoms, the domain: every organism belongs to one of three domains, Bacteria, Archaea and Eukarya.

rRNA was a good choice because it is present in all cellular organisms, has the same essential function in protein synthesis, and changes slowly, so even very distantly related organisms can be aligned and compared.

The domains also differ in other features, which supports the rRNA evidence:

Many archaea live in extreme environments such as hot springs or very salty water, but archaea are also abundant in soil, the oceans and animal intestines.