30 Building a phylogenetic tree

30.1 Introduction

This activity builds intuition for phylogenetic trees using discrete morphological characters. It corresponds to the appendix chapter Building a phylogenetic tree in the main book and aligns with the macroevolution “build a tree” exercise guide (supplemental/Macroevolution_Building_A_Tree/FylTree_ENG_2015.pdf in the main-book materials).

Students score a handful of organisms for several discrete characters, compute a pairwise difference table, and group the most similar organisms first — recovering the logic of tree-building by hand before (optionally) checking against an R clustering result.

30.2 Key Concepts

  • Character table: organisms (rows) scored for discrete characters (columns), e.g. presence/absence.
  • Pairwise difference (distance) table: count of characters that differ between each pair.
  • Grouping rule: join the closest pair first, then add progressively deeper splits; internal nodes represent inferred most-recent common ancestors.

30.3 Learning Outcomes

By the end of this exercise, students will be able to: - Build a character table for a set of organisms. - Compute pairwise differences and translate them into a tree. - Relate a morphology-based tree to the idea of evolutionary relatedness.

30.4 Activity Overview

Suggested Timings: - 5 minutes: Introduce characters, character tables, and the grouping idea. - 15 minutes: Students score six organisms for seven characters and build the pairwise difference table. - 10 minutes: Draw the tree by grouping closest relatives first. - 5 minutes: Optional R check with dist() + hclust(); discussion.

30.5 Instructions for Facilitating

  1. Choose six organisms (use character cards if available, or invent six with clear, contrasting traits).

  2. Agree on seven discrete characters that vary among them (presence/absence, colour class, tail-length category, etc.).

  3. Fill in the character table, then compute the pairwise difference (Manhattan/Hamming) table.

  4. Draw the tree by joining the closest pair first and working outward.

  5. Optionally compare with R:

    characters <- data.frame(
  row.names = c("A", "B", "C", "D", "E", "F"),
  trait1 = c(1, 1, 0, 0, 0, 1),
  trait2 = c(0, 1, 1, 0, 0, 1),
  trait3 = c(1, 1, 1, 0, 0, 0)
)
dist_mat <- dist(characters, method = "manhattan")
tree <- hclust(dist_mat, method = "average")
plot(tree, main = "Tree from morphological characters")

30.6 Questions & Model Answers

  1. How do you decide which organisms are most closely related?
    • The pair with the fewest character differences (smallest distance) is grouped first; these share the most character states and are inferred to be most closely related.
  2. What do the internal nodes represent?
    • Inferred most-recent common ancestors of the organisms in that grouping, based on shared characters.
  3. Why might a morphology tree differ from a molecular tree?
    • Convergent evolution and shared ancestral (rather than shared derived) characters can mislead morphology-based grouping; molecular data provide many more, more independent characters. A good prompt for discussing homoplasy.

30.7 Teaching Tips

  • By hand first: build the tree manually before running hclust() so students own the grouping logic rather than treating R as a black box.
  • Choose contrasting organisms: characters that clearly vary make the groupings unambiguous and the exercise faster.
  • Discuss ties and conflicts: when two groupings are equally good, or characters conflict, use it to introduce homoplasy and parsimony.

30.8 Common Pitfalls

  • Characters that don’t vary: a character shared by all (or by none) carries no grouping information — check for this before computing distances.
  • Confusing similarity with relatedness: shared ancestral traits do not indicate close relationship; only shared derived traits do.
  • Reading the dendrogram height as time: branch lengths from hclust() reflect distance, not calibrated evolutionary time.