How do you translate a phylogenetic tree or cladogram into another language without changing the evolutionary relationships it depicts? Read the branching structure before reading the labels. Identify the root, tips, internal nodes, clades, shared ancestors and any branch-length scale, then translate the explanatory language while preserving which lineages are related. Two tips printed side by side are not necessarily closer relatives, and a branch that looks longer is not automatically older unless the diagram defines length that way.
This guide explains how to translate phylogenetic trees, cladograms and evolutionary relationship diagrams accurately into written text or another language. It is designed for biology students, educators, science translators, accessibility writers and editors of research or museum materials. It covers common ancestors, sister taxa, clades, rooted and unrooted trees, evidence and uncertainty, support values, scientific names and the misleading everyday language that can turn a sound figure into an inaccurate story.
A dependable approach combines linguistic accuracy with topology: preserve which branches connect to which nodes. A good translation never says a living species evolved from another living species merely because their names appear at different points on the page. It distinguishes ancestry from resemblance, diagram order from time and supported relationships from unresolved ones. Those distinctions are the central meaning of the figure.
A phylogenetic tree is a model of relationships, not a family portrait
A family tree of individuals often records documented parent-child relationships. A phylogenetic tree describes hypothesised evolutionary relationships among taxa, populations, genes or other biological sequences. Its branches represent lineages and its internal nodes represent inferred common ancestry or branching events under a particular model. The exact meaning depends on how the tree was constructed and labelled.
A phylogeny is not a photograph of history. Scientists infer relationships from evidence such as DNA sequences, morphological traits or other data. A tree can be revised when evidence, sampling or methods change. If the original caption says inferred, estimated or proposed, the translation must not turn the result into an unquestionable historical record.
The OpenStax explanation of phylogenetic trees provides a useful foundation: rooted trees have a direction from a common ancestral lineage, branch points denote divergence and rotating a node does not change the represented relationships. These properties are more important than the appearance of a tidy left-to-right list.
Identify the object: rooted tree, unrooted tree or cladogram
Rooted trees depict an ancestral direction relative to a designated root. Unrooted trees show connections among groups without identifying one common starting point in the diagram. A cladogram generally emphasizes branching order rather than quantified elapsed time. Other phylogenetic representations include phylograms with branch lengths representing inferred evolutionary change and chronograms with a time scale.
Before translating, inspect the root designation, legend and axis labels. A graphic drawn from left to right is not automatically rooted on the left. Circular or radial layouts can also represent rooted or unrooted structures depending on the construction. If the source does not designate a root, do not assign an earliest ancestor simply because one branch is near the edge of the page.
The chosen English terms may overlap in everyday speech, but a scientifically careful translation follows the author’s definitions. If branch lengths represent substitutions per site, say so. If the figure is a schematic cladogram without a temporal scale, avoid calling it a timeline. A central task is preserving what the drawing measures and what it does not.
Read tips, nodes and edges before translating taxon names
A tip, also called a terminal node or leaf, labels an included taxon, specimen, sequence or other operational unit. An internal node connects branches and can represent a reconstructed common ancestor or the divergence of lineages, depending on the analytical model. Edges are the segments connecting nodes. These graphical roles should be identified before a translator reads the text as ordinary prose.
Imagine tips labelled A, B, C and D. If A and B connect to one internal node and C and D to another, the tree depicts two pairs of sister groups in that arrangement. Rewriting the names into a straight alphabetical list A, B, C, D loses those pairings. A good text alternative can state that A and B form one sister pair, C and D form another, and the two pairs join at a deeper node.
The task is not to make the map of relationships sound dramatic. It is to make the structure checkable. Ask whether a reader can reconstruct which groups share the nearest depicted common ancestor. If the translation cannot answer that question, it has probably preserved labels while discarding the biology.
Sister taxa are defined by common ancestry, not adjacency
Sister taxa are two descendant lineages that share an immediate common ancestral node relative to the rest of the sampled tree. Their visual position can change if the drawing is rotated around nodes. Two adjacent words near the right margin are not necessarily sister taxa; lines may connect them through distant branching points. Trace the paths from each tip back to the nearest shared node.
Never infer closeness from alphabetical sorting, line spacing or the number of centimetres separating printed names. A journal can rearrange a tree for page layout without changing its meaning. A translator who reorders labels for readability should preserve their branch attachments, not merely their original vertical positions.
Use phrases such as A and B are sister taxa, or A and B share a more recent common ancestor with each other than either does with C, when the tree supports those statements. Avoid phrases implying that A is the parent of B or that B directly evolved from A unless the original research establishes such a specific ancestral relationship.
Rotating branches does not change the tree’s topology
A common misunderstanding treats the topmost species as more advanced or the bottommost species as primitive. On a normal rooted phylogenetic tree, rotating two descendants around an internal node preserves their branching relationship. The order of names at the tips can therefore change while the evolutionary hypothesis stays exactly the same. This is a core reason translation must preserve connections rather than visual order.
Imagine the same tree printed once with A above B and once with B above A. If both connect to the same node, there is no new evolutionary claim. A translator who says the upper branch evolved later because it is drawn higher invents a temporal claim from layout alone. The orientation of a page is not a scale of evolutionary superiority.
A simple classroom exercise makes this intuitive: draw a Y-shaped split and rotate it on the paper. The two tips change positions but still meet at the same node. Then add another branch and ask which tips share the nearest common node. This reinforces that topology, not left-right aesthetics, carries the relationship.
Root position determines the interpretation of ancestral direction
A root provides an orientation for reasoning about common ancestry. In a rooted tree, tracing from tips toward the root moves toward older inferred ancestors; tracing from root toward tips follows descendant lineages. A tree may have its root at the bottom, left or centre. The visual direction must be read from the figure’s actual rooting, not assumed from a typical diagram.
Some publications include a designated outgroup to help locate the root. An outgroup is chosen outside the ingroup under study and can provide a comparative reference; it is not necessarily the oldest living organism in the entire figure. Translate outgroup and ingroup according to their analytical roles instead of using casual expressions such as outsider or original species.
If the root is unclear or the tree is explicitly unrooted, avoid claims about which lineage diverged first. You may still describe local groupings supported by the unrooted topology, but statements about chronology need a rooted interpretation. Respecting the limits of the diagram is part of an accurate translation.
Branch points represent divergence, not a ladder of improvement
An internal branch point indicates that the inferred lineage diverged into descendants or that a relationship is represented by a branching node. It does not imply that one surviving species transformed into another surviving species. Describing one modern tip as a halfway stage toward another invites the false impression of a linear evolutionary ladder.
Evolutionary trees are branching models. Multiple present-day lineages have been evolving from their shared ancestors, often for long periods, whether one looks morphologically simple or complex today. A diagram that places one taxon near the base is not declaring it inferior. The language of basal should be used only in its relational phylogenetic sense and with suitable context.
A faithful description names the branching event and descendant groups. If a particular lineage diverges earlier relative to the sampled group, state that comparative relation. Do not write that it stopped evolving, remained unchanged or failed to progress. Those claims require evidence independent of its location in the tree.
Explain clades without excluding descendants
A clade is a group consisting of a common ancestor and all of its descendants, within the relevant phylogenetic representation. If a diagram highlights a monophyletic group, the inclusion of every descendant lineage from the selected node matters. Translating it as a collection of similar-looking species can misrepresent both membership and evolutionary rationale.
OpenStax’s discussion of clades and evolutionary relationships emphasizes that clades include all descendants from a branch point. A diagram may shade a region to mark a clade. The translator should follow the bracket, shading or node annotation rather than guessing membership from adjacent text.
A useful accessible sentence says The highlighted clade begins at the common ancestral node of A, B and C and includes all three descendant lineages. If the graphic highlights only A and B but their shared ancestor also has C as a descendant, investigate whether the shading denotes a different concept or whether the grouping is non-monophyletic. Do not fix a scientific disagreement by silently redefining the figure.
Distinguish clade, taxon and grade
Taxon is a named or operational grouping at some level of classification. Clade is defined by ancestry and complete descent from a common ancestor. Grade can refer to organisms grouped by a level of organisation or characteristics rather than one exclusive clade. These words can overlap in classroom discussions but are not universal synonyms.
When an author writes clade, do not automatically replace it with species, family or type. A clade may contain organisms across several formal taxonomic ranks. Conversely, not every everyday category based on appearance is a clade. Translate the category as precisely as the source permits and explain it for learners rather than swapping scientific terms for familiar words.
If a figure marks a group by a bracket, verify whether it is a named taxon, a reconstructed clade, a trait grouping or simply a visual focus. Colour and proximity do not determine biological membership without a legend. The wording should preserve the author’s intended grouping and its evidential basis.
Polytomies communicate unresolved branching or representation choices
A polytomy is a node from which more than two branches emerge in the displayed tree. In many instructional settings it can indicate unresolved relationships among the descendant lineages. Some trees may also represent an intended multifurcation under a particular interpretation. The translator should read the caption before assigning the source’s reason.
A major mistake is to force a three-way branch into a specific sequence because ordinary language prefers one thing happening before another. If A, B and C emerge from an unresolved node, writing first A diverged, then B, then C invents a resolution. A careful sentence says the diagram does not resolve the branching order among A, B and C, if that is what the source means.
Uncertainty is information. Learners need to understand that scientists sometimes know several taxa belong together without being able to determine the order of their divergences confidently. Preserving this uncertainty is more educational than producing a neat but unsupported story.
Branch lengths may represent change, time or nothing quantitative
Some trees scale branch lengths to estimated evolutionary substitutions or other changes. Time-calibrated trees may scale them to millions of years. Cladograms can use arbitrary graphic lengths. Without reading the legend, a long drawn branch cannot be translated as a long period or a large amount of evolutionary change.
A tree with a scale bar labelled substitutions per site is measuring inferred genetic change along branches under a model. It is not directly a calendar timeline. A chronogram with a million-years axis permits a different interpretation. If the diagram lacks a scale, describe branch order rather than branch duration.
A translator should also keep numerical labels attached to the correct edges. Moving a branch-length number to a neighbouring branch changes the estimated history. In a dense tree, automated image extraction may read a number but not know which segment it belongs to. A field-by-field visual check is necessary.
Support values do not automatically mean probability of truth
Phylogenetic trees often annotate nodes with bootstrap support, posterior probabilities or other statistical quantities. These measures have specific interpretations under the relevant method. A bootstrap percentage is not simply the chance that an evolutionary event happened, and a posterior probability depends on the statistical model and assumptions. Translate the named support measure exactly.
Suppose an internal node is labelled 85 with a caption saying bootstrap support in percent. The target text should preserve that node’s 85 percent bootstrap support label. It should not claim an 85 percent probability that the species share an ancestor unless the statistical interpretation supports that phrasing. Similarly, a posterior probability of 0.95 should retain its metric rather than be relabelled a 95 percent match.
Some figures omit low support values or collapse poorly supported branches. Blank space then does not necessarily mean zero support. If uncertainty annotations are central to the publication, make their conventions clear in the translated legend and text alternative. The translation should protect readers from overconfidence as well as from omitted data.
The difference between a gene tree and a species tree matters
A gene tree depicts evolutionary relationships among sampled gene copies or sequences; a species tree depicts relationships among species or lineages. The two can differ for reasons such as gene duplication, incomplete lineage sorting, introgression or horizontal gene transfer. A translator who calls every tip a species risks misrepresenting a figure built from genes.
Read the labels carefully. A tip might contain a species name followed by a gene identifier or accession code, and one species may appear multiple times because multiple sequences were sampled. Those duplicate labels are not necessarily editorial mistakes. The figure may compare paralogous gene copies or multiple samples for a particular biological purpose.
When describing the figure, identify the units at its tips: species, individuals, sequences, haplotypes or gene copies. If the manuscript discusses discordance between a gene tree and a species tree, preserve the word discordance and its proposed causes, not a vague phrase meaning the diagrams disagree. The type of disagreement is a biological result that deserves accurate language.
Scientific names are identifiers, not ordinary vocabulary
Species and genus names follow biological naming conventions. Translating a Latinised binomial into everyday local vocabulary can destroy a stable identifier, even when a common name exists. For a species label such as Homo sapiens, preserve the scientific name with conventional formatting, including italicisation where feasible. A familiar common name may be supplied alongside it if appropriate.
Capitalisation and rank should also be checked. In binomial nomenclature, the genus name starts with a capital letter and the species epithet is lower-case. Author names, subspecies epithets and strain designations may add further identifiers. A text-extraction workflow can lose italics or punctuation; the translator should restore only what is supported by source and naming standards.
Do not assume that two similar common names refer to identical taxa across languages. Regional common names may overlap or denote different species. Use the scientific name as an anchor for cross-language identity and explain any target-language common name choice when confusion is plausible.
Traits on branches are not the same as names at tips
Some phylogenetic trees mark the inferred origin of characteristics such as a vertebral column, a particular molecular change or a shared derived trait. These labels refer to ancestral states or changes associated with branches, not to the names of the taxa at the tips. Translating them as if every trait belonged only to the nearest printed species can misstate the distribution.
If a trait label appears on the branch preceding several descendant groups, the diagram may imply that those descendants inherit that characteristic from the inferred ancestor, subject to later gains or losses and the model’s assumptions. Translate its position and scope. Ask which descendants occur beyond that marked branch, and whether the figure includes exceptions or reversals.
Trait evolution is also not necessarily a one-time process. Similar features can evolve independently in different lineages, and ancestral-state reconstructions involve uncertainty. If the original uses phrases such as independently evolved, convergent or inferred gain, retain them. Avoid saying the diagram proves a single origin when the source is more cautious.
Traits, similarity and ancestry are separate questions
Two species can look similar without being each other’s closest relatives. Adaptations to similar environments can lead to convergent features. Conversely, closely related taxa can appear strikingly different after evolving in different settings. A translator must distinguish morphological similarity from phylogenetic closeness, especially when teaching readers who instinctively group organisms by visible resemblance.
A tree may be built from DNA evidence even when its accompanying photographs encourage other impressions. Describing the pictured organisms as grouped together because they look alike may be wrong. Ask what evidence the source authors used: molecular sequences, anatomy, developmental traits or combined datasets. If the caption states the basis, keep it in the translated explanatory text.
This is a powerful cross-disciplinary lesson in reading visuals. Pictures attract attention, but the diagram’s formal claim lies in its structure and its method. A fluent description that tells an appealing visual story can be less accurate than a plain sentence naming a shared ancestor and a branching pattern.
Do not confuse taxonomic ranks with evolutionary branching events
A classification table may nest domain, kingdom, phylum, class, order, family, genus and species. A phylogenetic tree shows inferred ancestry, which does not always align neatly with the familiar ladder of ranked categories. Translating a branch point as the moment a formal taxonomic rank appeared can add an unsupported chronology.
Some figures include ranks for educational purposes, while others intentionally avoid them. If labels show family names or clade names, retain those as identifiers. Do not insert missing levels to make the figure appear complete. Taxonomy and phylogeny answer related but distinct questions: how groups are named and classified versus how their lineages are hypothesised to be related.
A useful translation can mention that a clade contains several named families if the original establishes that grouping. But the name of a family should not be treated as a single living ancestor from which all other labelled species descended. Keep the levels of representation separate.
Worked example: four tips in two sister pairs
Imagine a rooted instructional tree with tips A, B, C and D. A and B join at node X, C and D join at node Y, and X and Y connect at deeper node Z. A faithful text rendition says that A and B are sister taxa, C and D are sister taxa, and both pairs share a deeper common ancestor represented by Z.
Now rotate the branches around X so B is above A, then rotate those around Y so D is above C. None of the evolutionary relationships changes. The drawing’s tip order can now look different, but the same text remains correct. A translation that depends on the phrases first at the top or last in the list would be fragile.
The lesson is to write descriptions that survive legitimate layout changes. This matters in translated figures because text lengths and writing direction may force editors to move labels. Their visual placement can change; their branch attachments must not.
Worked example: an unresolved three-way split
Imagine a tree with one internal node from which A, B and C emerge, with a caption that says their relationships are unresolved. The correct explanation is that the figure does not establish which of the three lineages form a closer pair. You can still say the three belong to the displayed broader grouping when the diagram supports that conclusion.
An incorrect explanation would rank A as the oldest, B as intermediate and C as the most recent simply because their names are printed vertically. This adds both a temporal order and a false progression. Another error would arbitrarily call A and B sister taxa, replacing the polytomy with a bifurcating resolution absent from the original.
The translation should explicitly retain unresolved, uncertain or not yet determined where relevant. Students learning evolutionary reasoning gain an important lesson: recognizing what a model cannot distinguish is a sign of understanding, not a failure to explain it.
Worked example: branch lengths measured in substitutions
Consider a gene tree with a scale bar showing an inferred sequence divergence measure, and two tips connected by comparatively long branches. A careful description identifies the scale unit and the branches involved. It does not convert length to calendar years unless the source provides a calibration. The same horizontal distance on different figure types can mean different things.
If the original shows a numerical support value at the node and a scale bar beneath the tree, keep them separate. One is evidence or method-related support for a grouping; the other is a scale for branch lengths. Conflating the two can make a scientifically nonsensical sentence sound superficially technical.
A clear text alternative may say The branches are drawn to a genetic-change scale, with the scale bar as labelled, while internal node numbers show the stated support statistic. The exact units and values must come from the original figure. That cautious wording is more useful than invented precise divergence dates.
Worked example: the same taxon sampled more than once
Imagine a molecular tree with three sequence tips bearing the same species name but different accession identifiers. The correct reading may be that multiple genetic samples were included. A translator who deletes repeated names to avoid repetition removes data. A translator who calls the duplicates three separate species fabricates diversity.
Keep specimen or accession codes intact. Explain that these are distinct sampled sequences or records only if the caption supports that interpretation. If one taxon appears in different parts of the tree, the pattern could be biologically important or reflect data or identification issues. Do not silently rearrange tips to make groups visually tidy.
A sound translated caption notes the unit of analysis and the repeated labels. This allows readers to appreciate why the tree’s result may differ from a simple species classification list, without imposing conclusions about taxonomic errors or evolutionary mechanisms that the figure itself has not established.
How to translate tree captions and figure legends
Captions often contain the key to the drawing’s meaning: dataset, analysis method, rooting, branch-length scale, support statistics, colours, highlights and abbreviations. Translate the caption before finalising the body description, but verify its claims by inspecting the graphic. A visually obvious branch can still carry a technical meaning determined by an otherwise small legend.
Preserve qualifiers such as maximum-likelihood estimate, Bayesian consensus tree, bootstrap replicate, posterior probability, outgroup and time-calibrated phylogeny. These terms explain how readers should evaluate the picture. Replacing them with generic language such as scientific tree may make the result appear easier while removing its evidential basis.
Likewise, translate figure-panel references accurately. If panel A is a gene tree and panel B is a species tree, do not swap the panel labels. If colours correspond to habitats or sample locations rather than clades, describe the colour’s assigned variable. A label’s function is determined by the specific figure and legend.
Names, scripts and page layout can change without changing topology
Longer translated names may not fit next to a tree’s tips. Editors may rotate or reformat text, add leader lines or move labels to a table. These are reasonable design adaptations only if every translated label remains attached to the correct tip. A misplaced label changes the reader’s interpretation of ancestry even when the words are individually accurate.
Right-to-left scripts require particular care. Do not mirror the complete diagram automatically when translating its surrounding text, because mirror operations can move the root and reorder groups in ways that confuse comparison with the original. If a layout is redesigned, explicitly verify the position of the root, each node and each taxon label.
Keep stable accession IDs and scientific names legible as identifiers. Mixed-script text may need direction controls in the publishing software to prevent numbers or parentheses from being reordered visually. Test the published version in the actual reading interface, not only in an editor’s preview.
Accessible descriptions need a hierarchy of relationships
A short text alternative can identify the image as a rooted phylogenetic tree comparing the named taxa. A longer description should explain the root, major clades, sister relationships, notable unresolved nodes, scale and statistical annotations important to the argument. That approach follows W3C guidance for complex images, which emphasizes equivalent information rather than a decorative phrase.
Do not list all leaf names without specifying the branching structure. That produces a catalogue, not an accessible phylogeny. Conversely, do not describe only the decorative appearance of branches. Readers need to know which groups are inferred to share ancestors and what the figure’s scale and support numbers mean.
For a large tree with hundreds of tips, provide an overview of major groupings and a supplementary structured representation or dataset where practical. The goal is to preserve the scientific conclusions the figure supports, not to overwhelm users with every tiny visual mark. A textual tree in a standard notation can aid technical users, but it does not replace a readable explanation for non-specialists.
Keep uncertainty explicit in evolutionary storytelling
Phylogenetic diagrams invite narratives: one group split, another diversified, a trait appeared and an environmental change followed. Such stories can be useful when supported by the evidence, but a tree alone may establish only a relative branching hypothesis. It does not automatically identify the habitat, cause, exact date or mechanism of each divergence.
Translate statements such as likely, inferred, consistent with, unresolved and supported under this model faithfully. Do not remove the qualifiers to make the target text more confident. A branch labelled with a confidence or support measure does not prove a unique historical scenario. Scientific education benefits when uncertainty is treated as a meaningful feature of evidence.
A good translator asks which parts of the account come from the topology, which from the scale, which from an independent fossil or molecular calibration and which from the author’s interpretation. Keep these sources of information distinct. Readers can then appreciate both the explanatory power and the limits of the tree.
Avoid anthropocentric and progress-based phrasing
Terms such as higher, lower, advanced, primitive and missing link can suggest a ladder of perfection rather than branching evolution. Even when an older source uses such expressions, a translator should determine their intended technical meaning and avoid adding stronger value judgements. Where fidelity requires retaining a historical phrase, contextualise it rather than presenting it as a current scientific ranking.
Modern organisms are not unfinished versions of another modern organism. A diagram showing a branch near the root does not establish that its surviving descendants have stopped changing. Nor does a small number of sampled tips indicate that a group lacks ecological or evolutionary importance. Translating the figure should help readers think in lineages, not in winners and losers.
The distinction matters in school science and public-facing museum text. Accurate wording may be simple: these lineages share an older common ancestor; this branch diverged earlier relative to the others shown; or this group retains an ancestral feature. Each is more defensible than a dramatic sentence about one species becoming another present-day species.
Practical glossary for multilingual biology writing
A useful glossary distinguishes phylogeny, phylogenetic tree, cladogram, root, tip, leaf, node, branch, clade, monophyletic group, sister taxa, outgroup, ingroup, polytomy, branch length, divergence, common ancestor and support value. Add the context and the specific figure where each term occurs. Some target languages may have close everyday equivalents that imply the wrong direction or certainty.
Keep scientific names in a separate identifier column. Record any accepted common names and note when one vernacular name covers several possible taxa. Terms such as gene tree, species tree, bootstrap support and posterior probability belong in another column with their methodological definitions. This reduces the temptation to reuse a familiar word for a technically different concept.
Learners can deepen this skill through the eduKateSG guide to concept maps and mind maps, where connecting lines can encode named relationships. The underlying habit is similar, although a phylogenetic branch has specifically evolutionary meaning.
A verification process that tests biology and language
First inspect the original tree and record whether it is rooted. Second list the labelled tips and identifiers. Third trace every important internal node and identify sister groups and clades. Fourth note branch-length scales, support values, unresolved nodes and colouring conventions. Fifth translate the caption and description, retaining methodological qualifications.
Next perform a topology comparison: for each sister pair in the source, is the same relationship stated in the target? For each highlighted clade, are the descendants included correctly? For any reported divergence order, does the original root and branching order justify it? Check that numbers remain attached to the right nodes or branches and that missing values have not become zero.
Finally, have someone read the translation without the image and answer a few structured questions. Which taxa are sister groups? Is the tree rooted? Is branch length time-calibrated? Where is uncertainty shown? If the answers differ from the source, revisit the description. This is a stronger quality test than checking whether every species name is spelled correctly.
Questions readers often ask about translating phylogenetic diagrams
Can I translate scientific names? Usually preserve formal scientific names as identifiers and translate common names or explanations separately where appropriate. Check spelling, rank and formatting. Do not replace a scientific name with an ambiguous regional common name without context.
Does the top branch mean the most evolved species? No. Tip position is typically a layout choice, and rotating internal nodes does not change evolutionary relationships. A position on the page is not an evolutionary ranking.
Can a branch length tell me how old a species is? Only if the figure explicitly defines a time-calibrated scale supporting that interpretation, and even then the scientific meaning needs care. Many trees scale branches to genetic change or use arbitrary lengths. Check the legend.
What if I cannot tell whether the tree is rooted? Preserve relationships that can be read safely and state that the root or direction of ancestry is not identified in the available version. Do not invent an earliest ancestor to complete the narrative.
How do I describe a huge tree for accessibility? Begin with major groups and the central scientific conclusion, then provide deeper structured details or a linked representation where useful. Avoid a flat list of hundreds of labels that hides the branching relationships.
Further reading and connected knowledge
For biology foundations, read OpenStax on the organisation of the tree of life and OpenStax on determining evolutionary relationships. These references clarify rooted and unrooted diagrams, sister taxa, clades and the limits of branch-length interpretation. The original scientific figure and its caption remain the authority for the particular tree you are translating.
To compare different kinds of structured translation, see eduKateSG’s guide to family trees and kinship diagrams, which deals with documented human relationships rather than inferred evolutionary branching, and the guide to Venn diagrams, which teaches precision about group membership. Similar visual skills support different scientific claims.
The reliable way to translate phylogenetic trees
Recover the tree’s topology before translating its words. Identify whether it is rooted, trace the nodes, preserve sister groups and clades, read the scale and statistics correctly, and keep scientific names stable. Explain uncertainties rather than hiding them. A reader of the translated version should see the same relationships and the same limits of evidence that the original figure presents, even when the language and layout have changed.