Lesson 24 — Convergence vs drift vs shared ancestry

BIO 202, Spring 2026, draft v0. Two lineages with the same trait. Did they inherit it? Did they each evolve it? Or do they just superficially look similar?

Draft skeleton. Stage scenarios and anchor quotes are in place; the simulator and code panels are not yet wired.
It's very rare for an organism to evolve a totally new niche. Most of that invading-a-new-habitat stuff happened long ago. Instead, you get something that looks like what we call a butterfly fish now — but it's not a butterfly fish, it's going to go extinct. Then some other group evolves that shape. That group goes extinct. Then another one. Lots of repetition. We're going to see crocodile-like things from the Mississippian onward — but they aren't all crocodiles. In fact very few of them are. Many groups evolve a crocodile-like body. They go extinct. Something else evolves back into that role. — 440_lec07_04

A — Trait similarity across a tree under Brownian evolution

Simulate a continuous trait evolving up a tree. Tips with similar trait values can be similar because they share an ancestor (homology), because they each independently arrived at the same value (homoplasy / convergence), or because the trait is so neutral that anyone can have any value (drift).

These basic building blocks — the nerve tube, the gut tube, the notochord, the arches, the somites — they form in all vertebrates, and in different vertebrates we just fold them in a weird way. That gives us the different anatomies. We'll see this with various bones of the body, lots of muscles. They're just folded and shifted and warped in different ways. So a lot of the anatomy is looking at the same structure across different vertebrates and asking: why is it bent this way? — 440_lec02_01
TODO: trait-on-tree sim. Pull trait values on tips; rank tip pairs by similarity; ask whether similarity correlates with phylogenetic distance.

B — Penguins and hummingbirds

Two distantly related birds with similar forelimbs. Two closely related birds with very different forelimbs. Three classifications: homology, homoplasy, ancestral retention. Sort five trait/lineage pairs.

It doesn't make sense ecologically that the penguin — swimming around, not flying, eating shrimp and krill and fish — and the hummingbird — barely ever landing, never going in water, eating nectar — should have forelimbs that are the most similar to one another. It's weird that penguins and hummingbirds have forelimbs that are more similar than penguins and dolphins. There's no logical reason these traits should be linked, and there's not even an ecological reason. — 202_lec02_06
TODO: five-pair classification drill. Penguin/hummingbird, pangolin/armadillo, etc.

C — The test that distinguishes them

Mapping the trait onto the phylogeny and counting transitions. If the trait requires many independent gains, it's convergent. If it requires one gain and many retentions, it's homologous. Parsimony as a first cut; likelihood as the principled version.

Look at the common ancestor of cartilaginous and bony fishes. Look at all his descendants. Two out of five have it. Three don't. So your options are either three groups lost it, or two groups evolved it independently. It's more likely for two groups to evolve it than for three to lose it. So we decided these are independently evolved. More likely that bony fish and cartilaginous fish evolved mouth teeth independently than that the others all lost them. — 440_lec03_06
TODO: ancestral-state reconstruction tool. Count minimum transitions for a trait on a tree; compare to likelihood-based reconstruction.

D — Anolis ecomorphs across the Caribbean

Anolis lizards on different Caribbean islands have independently evolved the same set of body plans (ecomorphs) — trunk-ground, twig, grass-bush, etc. Use data/clean/anolis_svl.csv and data/clean/anolis_tree.nwk to count how many independent origins each ecomorph has.

A perfectly reasonable thing to think: if I live in North Africa and I look at the animals and plants there and say, these things fit their habitat beautifully, they fit it perfectly, they're so well structured and well built for the desert — anywhere I find desert, I should expect to see them. That's a perfectly rational thing to think in a state of ignorance. The pangolins are all in the old world; the armadillos are all in the new world. They're superficially similar — homoplastic — but not homologous. Not built the same way. — 202_lec02_08
TODO: load anolis tree + ecomorph annotations; count ecomorph origins; .R export. Non-trivial code mod: shuffle ecomorph assignments on the tips and compute how often the observed pattern arises by chance.