How Tree TSAR Delimits Genera

An Integrative Framework for Evolutionary Coherence, Diagnosability, and Stability

A phylogenetic tree contains far more branches than any practical classification can name as genera. The difficult part of genus delimitation is therefore not simply discovering where branches occur, but deciding which branches should be represented at genus rank.

Tree TSAR expects ordinary genera to correspond to coherent evolutionary lineages, yet genus rank remains a classificatory decision. There is no universal amount of genetic distance, elapsed time, morphological disparity, or species number that automatically defines a genus (Humphreys & Linder 2009).

The preferred treatment combines monophyly, strong phylogenetic support, diagnosability, morphology, ecology, biogeography, nomenclature, stability, and practical taxonomic value. The objective is not to maximize the number of genera. It is to turn evolutionary history into a durable classification.

Monophyly Is the Starting Point

For an ordinary non-reticulate lineage, Tree TSAR requires genus-level monophyly. A genus should contain a common evolutionary lineage rather than species scattered among unrelated branches.

Traditional classifications sometimes grouped species by convergent morphology. Bailey et al. (2006) showed how molecular evidence could expose this problem in Brassicaceae, where several historically important characters evolved repeatedly.

A demonstrated failure of monophyly requires a solution, but the evidence must first be robust. Sparse sampling, weak support, omitted type species, or major conflict among datasets can make an apparently simple revision premature.

Tree TSAR therefore treats non-monophyly as a problem to resolve, not as an instruction to accept the first proposed rearrangement.

Phylogenomics Reveals Relationships; It Does Not Assign Ranks

Genome-scale datasets can resolve numerous internal clades within a large genus. Those clades are biologically informative regardless of whether they are named as genera, subgenera, sections, alliances, or informal groups.

Muñoz-Rodríguez et al. (2023) frame this problem explicitly in Ipomoea: increasing molecular resolution has to be balanced against monophyly, diagnosability, completeness, and nomenclatural stability.

Tree TSAR therefore separates two questions. What relationships does the phylogeny support? comes first. Which circumscription represents those relationships most clearly and usefully? comes next.

Morphology and Diagnosability

A genus that can be recognized only by consulting a molecular tree has limited descriptive value. Tree TSAR therefore places substantial weight on morphological diagnosability.

Useful characters can involve reproductive structures, fruits, seeds, pollen, wood anatomy, leaves, architecture, developmental traits, or combinations of features. No single unique synapomorphy is required when a stable combination of characters makes the group recognizable.

Morphology must still be tested against phylogeny because convergence can imitate relationship. A character is most persuasive when it helps diagnose a lineage independently supported by molecular evidence.

Ecology and Biogeography

Ecological and geographic patterns can strengthen a genus treatment when they align with phylogenetic and morphological structure. A clade may represent a regional radiation, occupy a distinctive ecological regime, or share a characteristic reproductive strategy.

These factors are supporting evidence rather than mandatory criteria. Unrelated plants can converge ecologically, and a coherent genus can occupy many regions and habitats.

The important signal is congruence among evidence types.

Divergence Time Is Context, Not a Cutoff

Molecular dating can help compare candidate genus boundaries, particularly when a proposed split corresponds to a very deep divergence accompanied by substantial biological differentiation.

Tree TSAR does not apply a universal age threshold. Plant lineages evolve at different rates, and taxonomic ranks have never been assigned consistently by time alone.

Comparisons must also distinguish stem age from crown age. The former concerns separation from the closest surviving sister lineage; the latter concerns the common ancestor of the diversity contained within the clade. Treating the two as interchangeable can make rank comparisons misleading.

Large Genera Are Not Defective by Definition

Moonlight et al. (2024) documented 86 flowering-plant genera with more than 500 species. Large genera therefore represent a substantial component of angiosperm diversity rather than an exceptional taxonomic failure.

Large size can make identification and revision difficult, but species number alone is not a reason to split. If the genus is monophyletic, coherent, diagnosable, and capable of useful internal organization, Tree TSAR generally prefers to preserve the broader group.

Infrageneric Classification Preserves Internal Structure

Subgenera, sections, subsections, series, alliances, and informal clades allow important internal branches to be named without converting every branch into a genus.

Tree TSAR uses these levels when they improve understanding. Formal infrageneric names are retained when appropriate; strongly supported but unnamed clades can be described informally without pretending that a nomenclatural act has occurred.

The species Lineage element can then expose this internal structure to readers. A broad genus and a detailed phylogeny are not mutually exclusive.

When a Traditional Genus Is Not Monophyletic

A non-monophyletic genus may have several defensible solutions. Tree TSAR considers, among other possibilities, expanding the genus to include a nested lineage, dividing it into several genera, resurrecting historical genera, or retaining a temporary broader presentation while critical evidence remains unresolved.

The Ipomoea example is instructive. Muñoz-Rodríguez et al. (2023) found several traditional segregate genera nested within Ipomoea and favored an expanded circumscription because the segregates were neither satisfactorily monophyletic nor diagnosable. The value of the example is not that broad genera always win, but that rank should solve the phylogenetic problem without creating a worse classificatory problem.

Nomenclature Constrains the Available Solutions

Generic names are typified. When a genus is divided, the clade containing the nomenclatural type retains the name, while other clades must use available names or require newly published ones.

This can make two otherwise phylogenetically valid solutions very different in practice. One may preserve most familiar combinations; another may require hundreds of new combinations.

Tree TSAR does not allow convenience to preserve a demonstrably indefensible genus, but nomenclatural feasibility and stability legitimately help choose among several defensible classifications.

When the Biology Outruns the Nomenclature

Phylogenetic evidence can strongly support a species’ placement before the required new combination has been validly published. Tree TSAR can mark such a placement with [TSAR] when no suitable formally published name or combination is available.

[TSAR] is an editorial marker, not botanical authorship and not valid publication under the Madrid Code. It tells readers that the biological interpretation is provisional at the nomenclatural level.

If an appropriate published name already exists, Tree TSAR uses that name rather than a provisional marker.

Hybridization Does Not Define the Boundary Between Genera

Plant genera are not defined by absolute reproductive incompatibility. The Madrid Code explicitly provides for hybrid nomenclature involving different genera, and intergeneric hybrids are biologically well documented.

Extensive natural hybridization can still raise questions about a proposed boundary, particularly if the parental genera are otherwise weakly differentiated. Occasional hybridization does not automatically require merger.

Reticulate lineages require additional care. A lineage can have hybrid ancestry and later become stable, diagnosable, fertile, and persistent. Its evolutionary origin and the nomenclatural status of its name are separate questions; Tree TSAR does not remove an applicable hybrid marker merely because the lineage is biologically stable.

Monophyly and Reticulation Need Different Language

Strict monophyly describes a branching history. Some plant lineages have demonstrably reticulate origins involving more than one parental branch.

For ordinary branching lineages, monophyly remains the genus standard. For reticulate lineages, Tree TSAR describes the network history explicitly rather than forcing it into language that implies a single bifurcating ancestry.

Reticulate origin does not make a lineage unreal. It changes the evolutionary model needed to describe it.

Conflicting Genomes Are Evidence Too

Nuclear, plastid, and mitochondrial histories may disagree because of incomplete lineage sorting, introgression, hybridization, organellar capture, inadequate sampling, or analytical limitations.

Tree TSAR prefers relationships supported across independent evidence but does not simply average discordance away. A surprising plastid placement should not reorganize a genus automatically when broad nuclear and morphological evidence strongly conflicts with it; neither should the discordance be ignored.

The conflict itself can reveal evolutionary history.

Evidence Used in Genus Delimitation

Factor	Role in Genus Delimitation
Monophyly	Required for ordinary branching genera; demonstrated non-monophyly requires resolution
Phylogenetic Support	Relationships should be strongly supported with adequate sampling; conflicts require investigation
Diagnosability	Strongly favored so the genus has descriptive and predictive biological value
Morphological Coherence	Important, while recognizing that convergence can mislead
Ecology and Biogeography	Supporting evidence for coherent evolutionary history
Divergence Time	Comparative context only, never a universal cutoff
Species Number	Not an independent reason to split or lump
Reticulation and Hybridization	Special evolutionary evidence requiring explicit interpretation
Nomenclatural Feasibility	Determines which names can correctly follow from the taxonomic decision
Stability and Utility	Important when several phylogenetically defensible classifications are available

The Tree TSAR Decision Process

Establish the phylogenetic problem. Determine whether the genus is monophyletic and whether the evidence includes adequate species sampling and critical nomenclatural types.
Compare independent evidence. Examine nuclear, organellar, morphological, ecological, geographic, and other relevant datasets.
Identify all defensible solutions. Consider broader circumscription, splitting, historical genera, infrageneric classification, or temporary retention pending stronger evidence.
Test diagnosability and coherence. Ask whether the proposed genera can actually be recognized and whether their biological identities add information.
Audit nomenclatural consequences. Determine which clade contains the type, what names are available, and how disruptive each solution would be.
Choose and explain the most informative stable treatment. Prefer the circumscription that combines evolutionary integrity, diagnosability, nomenclatural feasibility, and long-term usefulness.

The workflow prevents “the phylogeny says so” from substituting for taxonomic reasoning.

Small Species, Big Genera

Tree TSAR often favors well-supported terminal species within broadly useful genera. The phrase small species, big genera summarizes that tendency without functioning as an absolute rule.

The rationale is scale. Species rank records comparatively terminal evolutionary independence; genus rank organizes those species into broader relationships. A strongly differentiated species does not automatically justify promoting its entire clade to genus rank.

Tree TSAR therefore asks whether a generic split adds information that cannot be represented clearly through infrageneric classification.

When Tree TSAR Disagrees

Reasonable taxonomists can examine the same phylogeny and adopt different generic circumscriptions. One may emphasize narrow diagnosability, another nomenclatural continuity, and another recognition of every major branch at genus rank.

A Tree TSAR treatment should identify what the alternatives share and where they differ. The most useful disagreement is one a reader can reconstruct from the evidence.

References and Further Reading

Bailey CD, Koch MA, Mayer M, Mummenhoff K, O’Kane SL Jr, Warwick SI, Windham MD, Al-Shehbaz IA (2006) Toward a global phylogeny of the Brassicaceae. Molecular Biology and Evolution 23(11): 2142-2160. https://doi.org/10.1093/molbev/msl087

Humphreys AM, Linder HP (2009) Concept versus data in delimitation of plant genera. Taxon 58(4): 1054-1074. https://doi.org/10.1002/tax.584002

Moonlight PW, Baldaszti L, Cardoso D, Elliott A, Särkinen T, Knapp S (2024) Twenty years of big plant genera. Proceedings of the Royal Society B: Biological Sciences 291: 20240702. https://doi.org/10.1098/rspb.2024.0702

Muñoz-Rodríguez P, Wood JRI, Wells T, Carruthers T, Sumadijaya A, Scotland RW (2023) The challenges of classifying big genera such as Ipomoea. Taxon 72(6): 1201-1215. https://doi.org/10.1002/tax.12887

Turland NJ, Wiersema JH, Barrie FR, Gandhi KG, Gravendyck J, Greuter W, Hawksworth DL, Herendeen PS, Klopper RR, Knapp S, et al. (eds.) (2025) International Code of Nomenclature for algae, fungi, and plants (Madrid Code). Regnum Vegetabile 162. University of Chicago Press. https://www.iaptglobal.org/madrid-code-online