IB BIOLOGY · THEME A · UNITY AND DIVERSITY

Sorting life by its family tree

Biologists sort millions of species into groups. Here is why the old ranks stopped fitting, how DNA lets you build and read the tree of life, and how it split the prokaryotes into two domains. This guide is higher level only: standard level students can skip it.

IB Biology A3.2HL only3 hours HL9 chaptersdry lab

Start learningRun the dry labTeach it

SharkPerchFrogMouseLizardBirdA cladeA molecular clockA nodea common ancestor
A cladogram in one loop: the tree grows from the root, a clade is circled, a node is marked as a common ancestor, and a molecular clock ticks along a branch. Build one yourself in the guide.

01Learn

Nine chapters, all for higher level. Read one, play its figure through or tap its parts, and make each prediction before you open the answer.

  1. 1Why sort life at all? HL
  2. 2When the boxes do not fit HL
  3. 3Groups that are family trees HL
  4. 4A clade is an ancestor and all its descendants HL
  5. 5Differences that count the years HL
  6. 6The tree with the fewest changes HL
  7. 7Reading a cladogram HL
  8. 8Testing the boxes HL
  9. 9A new level above kingdom HL

A3.2.1 Why we classify

Why sort life at all?

There are more kinds of living thing than anyone can hold in their head. Biologists have named well over a million species, and many more are still waiting to be found. Nobody can learn them one by one. So we sort them. Classification turns the diversity of life from a heap into a map.

A good map does two jobs. It helps you name and find an organism: once you know which group a beetle or a fern belongs to, you know where to look it up and what it is probably like. And it makes further study possible. When a new species is found, its group tells a biologist what to ask about it, such as what its relatives eat, how they breed and how best to protect them, instead of starting from nothing. Medicine, farming and conservation all lean on groups in this way.

Try it first Before you read on, sort the six organisms in the figure into two groups, using any rule you like. Then compare your rule with a partner's. You will probably both have made sensible groups, and they will not be the same. A rule is worth keeping if the groups it makes tell you something. The next chapters ask which rule tells you the most.
  • Classification is needed because there are too many species to study one at a time.
  • Once organisms are classified, a broad range of further study becomes possible.
  • Different rules give different groups, so the rule matters.
Watch out Classification is not just naming. A name tells you nothing about a species. A group tells you what to expect from it.

Which is the best reason for classifying organisms?

fliesdoes not flyhas a backboneno backbonesix organisms, in no orderBatFishMushroomBirdFrogButterfly
  1. Six organisms, to sort any way you like.
  2. Rule one: flies. The bat, the bird and the butterfly go together.
  3. Rule two: has a backbone. Now the bat sits with the fish and the frog, and the butterfly has left it. Same six, different groups.

A3.2.2 The traditional hierarchy of taxa

When the boxes do not fit

Naturalists have long sorted life into a hierarchy of taxa. A taxon is a named group. Each species sits in a genus, each genus in a family, and so on up through order, class and phylum to kingdom. The system dates from Linnaeus, in the 1700s, and it grouped organisms mostly by features you can see.

It is tidy, and it has a problem. Evolution does not make boxes. It makes branches, and the branches split at different times. The ranks do not always follow that pattern.

Groups made from looks can gather unrelated lines: birds, bats and insects all fly, but each evolved its wings separately. Groups can also leave out relatives. In the old system crocodiles and lizards are in the class Reptilia, and birds have a class of their own, Aves. Yet the figure shows that crocodiles and birds share an ancestor that lizards do not.

Ranks are also arbitrary. Variation is gradual, and nothing in nature says how much difference makes a genus or a family. A family of beetles and a family of mammals need not be alike in age or in how much they differ, so a rank measures nothing fixed. Cladistics offers another way: unranked groups that follow the branching.

Remember it The ranks, from the top

  • KingKingdomthe biggest group
  • PhilipPhyluma major body plan
  • CameClassmammals, birds, insects
  • OverOrderprimates, carnivores
  • ForFamilycats, dogs, apes
  • GoodGenusclose species together
  • SoupSpeciesone kind of organism
A paradigm shift A paradigm shift is a fundamental change in how a science works. Classification has had one. Everyone used ranks, cases that did not fit piled up, and groups made from branching replaced them. Step through the second figure.

A taxonomist finds that a group in the old system leaves out one species that descends from the same common ancestor as the rest. What does that show?

ranks, for a humanKingdomAnimaliaPhylumChordataClassMammaliaOrderPrimatesFamilyHominidaeGenusHomoSpeciesHomo sapienshow they are relatedBirdCrocodileLizardMouseold classAvesold classReptiliasharedancestor
  1. The traditional ranks: each one nests inside the next, from kingdom down to species.
  2. Now the branching tree of descent for four vertebrates, built from DNA (simplified).
  3. The old classes. Reptilia holds the crocodile and the lizard, and Aves holds the bird on its own.
  4. But the crocodile and the bird share an ancestor that the lizard does not. The old class cuts that family in two.
Normal scienceEveryone sorts life intoranks, and it mostly worksThings that do not fitGroups hide real familyties, and the cases pile upA new ideaGroup by branching: onlywhole clades countThe new normalCladograms and DNAguide the groupsa paradigm shift
  1. Normal science: a framework everyone shares. For classification, the ranks.
  2. Cases that do not fit build up, such as groups that split a family in two.
  3. A new idea replaces the rule for grouping: use the branching tree, and group only whole clades.
  4. The new idea becomes the new normal. A change in the basic ideas of a field is a paradigm shift.

A3.2.3 Classification that follows relationships

Groups that are family trees

Think of your own family. Cousins tend to resemble each other more than strangers do, because they share grandparents. Living things are the same, only the family is far bigger and the shared ancestors are further back. The ideal classification follows that family tree, so that every member of a group has evolved from one common ancestor, and none of that ancestor's descendants are left out.

The great strength of such groups is prediction. A feature that appeared in the common ancestor is passed to all its descendants, so it is shared across the group. If you know that one member of the group has hair and feeds its young on milk, you can predict the same of a newly found member without testing it. The figure follows one feature through a tree.

The predictions reach beyond looks, to biochemistry, life cycles, behaviour and disease. Plants in one group often make similar chemicals, so a chemical that proves useful in one plant is first looked for in its relatives. A group made from looks alone cannot promise this, because look-alikes with different ancestors need not share anything underneath.

  • The ideal classification follows evolutionary relationships: all members of a group descend from one common ancestor.
  • Features of one member can be predicted for the rest, because they are shared within the group.
Watch out A prediction is a good bet, not a promise. A feature can be lost along a branch: snakes belong to a group of four-limbed animals, but their ancestors lost the legs. That is why a prediction is then tested.

A newly found frog falls inside a clade whose other members all make a toxin in their skin. Which statement is best?

FrogLizardKangarooMouseHumanhair and milknew speciesso expecthair and milk
  1. Hair and milk appeared once, on the branch leading to the mammals. Every descendant of that ancestor inherits them, unless a line loses them.
  2. A newly found animal. Its DNA says it is the mouse's closest relative: it joins the tree there (a made-up animal).
  3. It sits inside the mammal clade, so you can predict its features before you have studied it.

A3.2.4 Clades

A clade is an ancestor and all its descendants

A clade is a common ancestor together with every organism that descends from it. Test any group with two questions. Do all its members share one common ancestor? And are all of that ancestor's descendants inside the group? Both answers must be yes. Leave out one descendant, or add one that is not a descendant, and it is not a clade.

Clades nest like boxes inside boxes: birds are a clade, and so is the larger clade that holds birds and crocodiles. Tap the lettered nodes on the tree to ring the clade that starts at each one, then tap the old group called reptiles.

How do you decide which organisms share an ancestor? The most objective evidence comes from the base sequences of genes or the amino acid sequences of proteins. A sequence is a long string that can be compared position by position, and anyone counting the differences gets the same answer. Morphological traits can also place organisms in clades, but they need a judgement about which features count, and they can mislead when unrelated lines evolve similar structures.

  • A clade is one common ancestor plus all of its descendants.
  • Sequence evidence is the most objective, and morphological traits can also be used.
Watch out A clade is not decided by how organisms look, how big the group is or what rank it once had. Only descent counts. A group of look-alikes that leaves out a relative is not a clade.

A tree shows an ancestor X with four descendants, P, Q, R and S. Someone proposes the group P, Q and R. Is it a clade?

SharkFrogMouseLizardCrocodileBirdEDCBAold "reptiles"

Tap a letter to ring the clade that starts at that node. Then tap the old group called reptiles.

A3.2.5 The molecular clock

Differences that count the years

Mutations change a gene's sequence at random. When one species splits in two, each line collects its own mutations from then on. So the longer ago two species split, the more differences you find between the same gene in each (you met this with sequences in A4.1). If changes arrive at a roughly steady rate, the differences work like a clock.

To turn a count into a date you need a calibration. Take pairs of species whose split is dated from fossils, and plot their differences against time. A straight line through the points is the molecular clock. Now a pair with no fossil date can be placed: read its number of differences across to the line and down to the time.

The result is an estimate, never a date. The rate at which changes build up is not the same everywhere, because it is affected by the length of the generation time, the size of the population, the intensity of selection and other factors. In the figure's last step, a line that collected changes faster gives a quite different answer from the same count.

Remember it What shifts the clock

  • GerbilsGeneration timeshort generations copy DNA more often each year
  • PlayPopulation sizechance spreads changes differently
  • SportsSelection pressureselection removes some changes, keeps others
Watch out A molecular clock is not a watch. It gives an average rate, calibrated against fossils, and different genes and different lines can tick at different rates. That is why scientists quote a range and not a single date.

A clock is calibrated on mammals that breed every few years. It is then used to date a split between two lines of insects that breed several times a year. What will the estimate probably be?

01020304005101520time since the two lines split (millions of years)differences in 30 basesfossil-dated pairscalibrated clockundated pair: about 12faster clocktwice as fast: about 6
  1. Two pairs of species whose split is dated from fossils: 5 differences at 10 million years and 15 at 30 million (made-up numbers).
  2. The differences grow steadily with time, so a straight line through the points is the molecular clock.
  3. A pair with no fossil date differs at 6 places. Read across to the clock, then down: about 12 million years.
  4. If this pair's lines collected changes twice as fast, the same 6 differences would mean about 6 million years. The clock gives an estimate, not a date.
Change the mutation rate and the generation time, and see how the same differences give a different date.

A3.2.6 Building a cladogram

The tree with the fewest changes

A cladogram is a hypothesis about the order of splits. To build one from sequences, line up the same gene from each species and compare it position by position. Many trees could join the same species, so you need a rule to choose between them.

The rule used here is parsimony: choose the tree that accounts for the sequence differences with the fewest changes. Mutations are rare, so the explanation that needs fewest of them is the most probable. Add an outgroup, a species that split off earlier, to show which base is the older one.

The figure uses made-up data, small enough to count by hand. Each site is a column. A site where one species stands alone costs one change on every tree, so it cannot help. The sites where two species share a change are the ones that decide. Then try the tool: join the beetles yourself, watch the count, and run it again with a DNA column.

A cladogram is not a fact, and the method matters. Different criteria for judgement can lead to different hypotheses. A different gene, a different rule or traits of the body instead of sequences could give another tree, and new data can overturn one.

Watch out The most parsimonious tree is the most probable, not the certain one. Two trees can tie on the count, and then more data, more positions or another gene, is needed to choose between them.

Three possible trees need 14, 11 and 11 changes. What does parsimony tell you?

site123456OrangutanAATCGAHumanGTTAAAChimpanzeeGTCCAGGorillaAACCGAdiffers fromthe orangutanno helpno helpTree 1OrangutanHumanChimpanzeeGorilla7 changesTree 2OrangutanHumanGorillaChimpanzee10 changesTree 3OrangutanChimpanzeeGorillaHuman9 changes
  1. A made-up stretch of six sites in one gene. Pink marks where a species differs from the orangutan, the outgroup.
  2. Three trees could join the human, chimpanzee and gorilla. Sites 4 and 6 differ in only one species, so they cost one change on every tree and cannot choose between them.
  3. Count the changes each tree needs to explain all six sites. Tree 1 needs 7, tree 2 needs 10 and tree 3 needs 9.
  4. Parsimony picks the tree that explains the data with the fewest changes: tree 1, with the human and chimpanzee as closest relatives.
Traits12345MossBramblePebbleEmberLanternMade-up beetles and data, for teachingMossoutgroupBramblePebbleEmberLantern2324351MossoutgroupBrambleEmberPebbleLanternThis tree: 7 changes
  1. The table gives each beetle's traits (made up for teaching). Moss is the outgroup: its states are taken as the ancestral ones.
  2. One tree, with a bar on each branch where a trait changes: 7 changes. Can you find a tree that needs fewer?
Join the beetles into a tree, mark where each trait changes, and watch the count.
Traits and DNA sites12345abcdefMossACGTACBrambleGCGTCCPebbleGTATCCCinderGTATCCEmberACGCACLanternACGCATMade-up beetles and data, for teachingMossoutgroupBramblePebbleCinderEmberLanternbc5f2abced2ae341MossoutgroupBramblePebbleCinderEmberLanternThis tree: 16 changes
  1. The table gives each beetle's traits (made up for teaching). Moss is the outgroup: its states are taken as the ancestral ones.
  2. One tree, with a bar on each branch where a trait changes: 16 changes. Can you find a tree that needs fewer?
The same job with a DNA column added. Two of the traits mislead: does the tree with the fewest changes change?

A3.2.7 Analysing cladograms

Reading a cladogram

A cladogram is a map of splits, so you read it from the base up to the tips. The base is the root. Each fork is a node, which stands for a hypothetical common ancestor: it must have existed to explain the split, but it need not be a known fossil. Each line out to a tip is a terminal branch. Tap the parts.

To deduce a relationship, follow two tips back until their lines meet. That node is their most recent common ancestor. The nearer it is to the tips, the closer the relatives. Two groups that branch from the same node are sister groups. In the figure the human and chimpanzee meet at a node that the gorilla has not yet joined, so a human is closer to a chimpanzee than to a gorilla.

Any node and everything beyond it is a clade, so you can read clades straight off the tree. Then use the clade to deduce what its members share.

Remember it The parts of a cladogram

  • RabbitsRootthe base: ancestor of everything
  • NibbleNodea fork: a hypothetical ancestor
  • TastyTerminal branchthe line out to a tip
  • CarrotsCladean ancestor and all its descendants
  • SlowlySister groupstwo groups from one node
Watch out The order of the tips means nothing, because every node can be spun like a mobile. Tips at the top are not more advanced, and a tip is never the ancestor of another tip: humans did not evolve from chimpanzees, they share an ancestor. In this course, only the branching order carries meaning.

In a cladogram, species X and Y branch from node 1, and node 1 and species Z branch from node 2. Which statement is correct?

cladeGibbonOrangutanGorillaHumanChimpanzeerootnodeterminal branchsister groupsChimpanzeeHumanGorillaOrangutanGibbonhuman andchimpanzee arestill sisters

Tap a part of the cladogram, or spin it.

SharkPerchFrogMouseLizardBird
Answer questions about a cladogram: common ancestors, clades and which pairs are closest relatives.

A3.2.8 Testing a classification with cladistics

Testing the boxes

Cladistics can test an old classification. If a group is real, its members should form a clade. A case study is the figwort family, a large family of flowering plants that was built on the structure of its members, mostly their flowers. You do not need to learn the names, only the story.

Researchers then compared the base sequences of a gene from the chloroplasts of these plants and built a cladogram. The figworts were not one clade. Many of the plants belonged with plants that were not figworts, and the family was broken up, its members moved to other families, leaving a much smaller figwort family.

Why had structure misled? Plants in different lines had evolved flowers that look alike, by convergent evolution. The likeness came from similar pressures, not from shared ancestry, so a grouping built on it was false.

This is science working, not failing. The old classification was a hypothesis, and it made a testable prediction: its members form a clade. New evidence tested it and it was falsified. Scientific knowledge claims may eventually be falsified, which is why they are always open to new tests.

In an exam You can use either the common name or the scientific name of an organism. You are not asked to memorise the details of this case study, but you should be able to explain what it shows.

Chloroplast DNA places two plants with almost identical flowers in different clades. What is the best explanation?

grouped by flower structurefoxglovesnapdragonmonkeyfloweryellow rattlemulleinfigwortplantainbroomrapemintgrouped by chloroplast DNAfoxglovesnapdragonplantainmonkeyfloweryellow rattlebroomrapemulleinfigwortmintonce in the figwort familythe figwort family today
  1. Grouped by flower structure: six plants with similar flowers sit together as the figwort family (a simplified sketch).
  2. Grouped by chloroplast DNA instead. The same plants now fall into several different clades, with other plants between them.
  3. Only a small clade is still the figwort family. The flower resemblance was convergence, so the old family was not a real clade.

A3.2.9 The three domains

A new level above kingdom

For a long time life was sorted into a few kingdoms: animals, plants, fungi, and two kinds of single-celled life, the ones with nuclei and the ones without. The cells without nuclei, the prokaryotes, were treated as one group. Then, in 1977, Carl Woese and George Fox compared rRNA base sequences from many organisms, and the picture changed.

Why rRNA? Every living cell has ribosomes, so every organism has rRNA genes to compare. They can be lined up across the whole of life, from a bacterium to a whale. The comparison showed that the prokaryotes were not one group. Some, such as the methane-makers, had rRNA as different from ordinary bacteria's as bacteria's is from ours.

Life falls into three great branches, now called domains: Bacteria, Archaea and Eukarya (organisms whose cells have nuclei). Archaea look like bacteria under a microscope, but they share a more recent common ancestor with Eukarya. This was a revolutionary reclassification, with an extra level above the kingdoms.

Where the kingdoms went Animals, plants and fungi are all in the domain Eukarya. The kingdoms did not vanish: they sit one level below the domains.
Watch out "Prokaryote" does not name a clade. Bacteria and Archaea are both prokaryotes, but neither is the other's closest relative. Many archaea live in hot springs and salty lakes, but not all do: some live in the guts of animals.

Why was rRNA a good choice for comparing all of life?

Escherichia coliCyanobacteriamethane-makerssalt-loversanimalsfungiplantsprotistsBacteriaArchaeaEukaryashared byArchaea andEukaryadomains: a new levelabove kingdom
  1. The tree from rRNA sequences: three main branches, the domains Bacteria, Archaea and Eukarya. Relationships inside each domain are not shown.
  2. Archaea look like bacteria under a microscope, but their rRNA shows they share a more recent ancestor with Eukarya.
  3. A new top level of classification, above kingdom, proposed in 1977.
A tree built from rRNA sequences: three domains, with Archaea nearer to Eukarya than to Bacteria.

02Dry lab: Grow a tree

Six organisms, a table of features and a stretch of DNA. Choose the data, let the tool build the candidate trees, and see which tree needs the fewest changes.

Your questionDoes the tree you get depend on the data you choose? When you add more features, or switch from features to sequences, does the winning tree change, and does it ever tie?
What you can changeThe data (a table of features or a DNA sequence), the features included, and the number of sites used. The tool counts the changes each candidate tree needs.
What to recordFor each run: the data you used, the tree that needed the fewest changes, its score, and the score of the runner-up. Change one thing at a time. The tool gives the same answer for the same data every time, so the experiment is in your choice of data.
Six made-up fish, for teaching. River fish is the outgroup.traits counted at first12345River fishNorth lake fishNorth cave fishSouth lake fishSouth cave fishMarsh fish1 tiny eyes2 no pigment3 taste buds on head4 barbels5 fin spinetrees scoredchanges needed
The bench: six made-up fish and their traits. Grow scores every one of the 105 possible trees by the changes it needs, and the graph shows how many trees need each number. Which tree wins, and does it change when you add data or change how changes count?

Find out

  • Does adding a feature change the tree? Start with three features and record the winning tree and its score. Add the features one at a time, and record the winner and its score after each. When, if ever, does the winner change?
  • Features or sequence? Run the six organisms on the table of features, then switch to the DNA sequence. Do both give the same tree? Which gives the clearer winner, and why might that be?
  • Can two trees tie? Look for a set of features or sites where two trees need the same number of changes. What would you do next, as a scientist, to break the tie?
  • A feature that fits no tree well. Add a feature that two distant organisms share, as wings are shared by a bat and a bird. What does it do to the scores, and can enough of them change the winner?

Record

  • For every run: the data (features or sequence), how many features or sites you used, the tree that needed the fewest changes, its score, and the score of the runner-up.
  • Change one thing at a time and keep the rest fixed.
  • The tool counts changes by a fixed rule, so the same data always gives the same answer. There is no scatter to average: the experiment is in the data you choose, so repeat a run only to check you recorded it correctly.
  • Keep a table of your runs. Compare it with a partner's: did you reach the same winning tree from the same data?

You add a feature that only one of the six organisms has. What does it do to the scores?

You add several features that two distant organisms share, as a bat and a bird share wings. What is the likely effect on the winning tree?

The tool gives exactly the same answer every time you give it the same data. Why is that not a weakness of the model?

03Check

A fresh paper from the exam bank on this guide's topics each time: past-paper style questions, marked as you go, and saved to your Exam practice progress when you are signed in. It opens on biologybybradford.com, so it needs the internet.

Exam practice

Exam questions on A3.2, marked

Start 15 marks of exam questionsYour exam practice

04Downloads

This file is the guide, saved to run without the internet.

Saved from biologybybradford.com/guides/ib-biology-a3-2-classification-and-cladistics, where the guide is kept up to date with the notes, the revision guide and, for the teacher, the lesson plans and slides.

05Teacher resources

The guide as four one-hour lessons. Each has a plan to print (the timings, what to project and where to stop, the activities with what they need, what students get wrong) and slides to teach it from, in the same order, with the teacher's moves in the notes.

LessonCoversFocusChaptersLesson planSlides
1Why sort life, and when the boxes fail HL A3.2.1 to A3.2.360 min Why classification is needed, why the traditional ranks do not always follow evolution, and what a classification by ancestry lets you predict 1, 2, 3 Lesson planPDF 676 KB SlidesPPTX 703 KB
2Clades, and a clock made of DNA HL A3.2.4, A3.2.560 min What a clade is, why sequences are the most objective evidence, and how the molecular clock estimates divergence times 4, 5 Lesson planPDF 668 KB SlidesPPTX 1.2 MB
3Building and reading cladograms HL A3.2.6, A3.2.760 min Choosing a tree by parsimony, reading nodes and clades, and what changes when you change the data 6, 7, dry lab Lesson planPDF 662 KB SlidesPPTX 970 KB
4Figworts, three domains and putting it together HL A3.2.8, A3.2.960 min Cladistics testing an old classification, rRNA and the three domains, and an exam question on the whole subtopic 8, 9 Lesson planPDF 682 KB SlidesPPTX 1.4 MB
PDFEvery lesson planAll four plans in one PDF, with this overview at the front. 1.2 MB KEYNote template, filled inThe class’s note template with every gap filled. 1.1 MB

The note template, the revision guide and the offline copy are in Downloads, for you and the class.

PRACTISE AND REVISE

Make it for your class, or set it

Build for your class

Already on the site for A3.2

Card sets on splycr

Open a set to host it as a class game, or make a copy to edit.

How this was made. The course structure, the activities and the teaching choices come from my own classroom, my presentations and my years of teaching IB and AP Biology. I use Claude, an AI, to help write and build the interactive guides, lesson plans, slides and practice questions, and to draw the diagrams and code the interactives. I direct the work and review it against the IB and AP specifications before it goes live.