[Paper Review] Mathematical Models and Biological Meaning: Taking Trees Seriously

Jeremy L. Martin, E. O. Wiley|arXiv (Cornell University)|Aug 3, 2008

Evolution and Paleontology Studies15 references18 citations

TL;DR

This paper compares three mathematical tree models in phylogenetics—phylogenetic trees, Hennig trees, and Nelson cladograms—demonstrating that while phylogenetic and Hennig trees are isomorphic and carry identical evolutionary information, Nelson cladograms require interpretation of internal vertices as hypothetical ancestral species to achieve biological meaning. The key contribution is clarifying that without such interpretation, cladograms carry only phenetic, not phylogenetic, significance.

ABSTRACT

We compare three basic kinds of discrete mathematical models used to portray phylogenetic relationships among species and higher taxa: phylogenetic trees, Hennig trees and Nelson cladograms. All three models are trees, as that term is commonly used in mathematics; the difference between them lies in the biological interpretation of their vertices and edges. Phylogenetic trees and Hennig trees carry exactly the same information, and translation between these two kinds of trees can be accomplished by a simple algorithm. On the other hand, evolutionary concepts such as monophyly are represented as different mathematical substructures are represented differently in the two models. For each phylogenetic or Hennig tree, there is a Nelson cladogram carrying the same information, but the requirement that all taxa be represented by leaves necessarily makes the representation less efficient. Moreover, we claim that it is necessary to give some interpretation to the edges and internal vertices of a Nelson cladogram in order to make it useful as a biological model. One possibility is to interpret internal vertices as sets of characters and the edges as statements of inclusion; however, this interpretation carries little more than incomplete phenetic information. We assert that from the standpoint of phylogenetics, one is forced to regard each internal vertex of a Nelson cladogram as an actual (albeit unsampled) species simply to justify the use of synapomorphies rather than symplesiomorphies.

Motivation & Objective

• To resolve long-standing confusion in phylogenetics about the biological meaning of different tree models.
• To clarify the mathematical and biological distinctions between phylogenetic trees, Hennig trees, and Nelson cladograms.
• To demonstrate that translation between phylogenetic and Hennig trees is possible via a simple algorithm, preserving all evolutionary information.
• To argue that Nelson cladograms are biologically meaningful only if internal vertices are interpreted as hypothetical ancestral species.
• To show that requiring all taxa to be leaves in cladograms reduces representational efficiency and risks inaccuracy.

Proposed method

• Uses graph theory to formalize trees as mathematical structures with vertices and edges, defining key concepts like root, parent, ancestor, and subtree.
• Introduces two algorithms: one to convert phylogenetic trees to Hennig trees (and vice versa), ensuring isomorphism and information preservation.
• Analyzes the structural differences in how monophyletic groups are represented in each model, showing that node-based and stem-based circumscriptions yield different results.
• Applies Algorithm C to convert Hennig trees into quasi-cladograms and Algorithm D to reverse the process, demonstrating recoverability.
• Evaluates the biological interpretation of edges and internal vertices in Nelson cladograms, arguing that without assigning them to ancestral species, they carry only phenetic information.
• Uses illustrative figures (e.g., Figures 1–6) to compare tree representations and demonstrate misinterpretations arising from incorrect contraction of cladograms.

Experimental results

Research questions

• RQ1 How do phylogenetic trees, Hennig trees, and Nelson cladograms differ in their mathematical and biological interpretation?
• RQ2 Can a one-to-one correspondence be established between phylogenetic trees and Hennig trees through a formal algorithm?
• RQ3 Why do Nelson cladograms fail as phylogenetic models without interpreting internal vertices as hypothetical ancestral species?
• RQ4 What are the consequences of requiring all taxa to be represented as leaves in cladograms for representational efficiency and accuracy?
• RQ5 How do different circumscription methods (node-based vs. stem-based) affect the identification of monophyletic groups in these models?

Key findings

• Phylogenetic trees and Hennig trees are isomorphic: they carry identical evolutionary information, and translation between them is possible via a reversible algorithm.
• Monophyletic groups are represented as subtrees in Hennig trees but as edge sets in phylogenetic trees, requiring careful circumscription to avoid polyphyly.
• Nelson cladograms, when not interpreted as representing ancestral species, carry only incomplete phenetic information and lack phylogenetic significance.
• Internal vertices in a Nelson cladogram must be interpreted as hypothetical ancestral species to justify the use of synapomorphies over symplesiomorphies.
• The requirement that all taxa be leaves in a cladogram makes the model less efficient and potentially less accurate than Hennig or phylogenetic trees.
• Incorrect contraction of a quasi-cladogram (e.g., merging vertices) leads to indistinguishable taxa, demonstrating the need for precise interpretation of internal nodes.