In hid book Wonderful Life, Stephen Gould presents the bewildering array of Cambrian fauna, and tries to make a sense of it. Along the way, he hits on a favorite theme: the power illustrations have to influence our thinking.
Gould was working towards the time a creature, Hallucigenia, was illustrated upside down in a monograph.* This interfered with the ability of researchers, though highly skilled in spacial visualization, to recognize the creature’s upside from downside and frontside from back.
He first hit on the misperceptions created by the popular march of progress view of human origins. The view is intended to help us make sense of the complexities of evolution. But it also powerfully creates the misperception that evolution is simple and orderly and progressive.
Life is not a linear progression, steadily marching towards a highly-advanced form. It is a copiously branching bush, Gould says.
Life is a copiously branching bush, continually pruned by the grim reaper of extinction, not a ladder of predictable progress.
Wonderful Life, Stephen J. Gould
I will be questioning the idea of life as a tree or a bush here. The idea that there is one true tree of life.
For there is, after all, one true tree of life, the unique pattern of evolutionary branchings that actually happened. It exists. It is in principle knowable. We don't know it all yet.
A Devils’ Chaplain, Richard Dawkins
[T]he history of life cannot properly be represented as a tree.
Phylogenetic classification and the universal tree, Ford Doolittle
Life is a march of progress. We can arrange selected instances from our ancestry in a straight line.
Life is a tree. It has branches joined at nodes. Life is like a tree, but actually a more copiously branching bush. It is a tree, but messier or more complex.
Life is a bird, landing on an island, and radiating into ecological niches. The rapidity of the branching makes more of a nest than a tree.
Life is the taxonomic families of bees, with shared characteristics placed in a matrix. It is hard to make a family tree for them.
Life is a layering of many trees, one on top of the other. We can peer through the layers, looking for a consensus of where the branches are.
Life is a braided stream. A complex network of splittings and joinings.
Life is a cloud of probabilities.
Phylogenic Trees
According to the facts of life, in a sexually reproducing species, one individual will have two parents. We can draw a pedigree chart like this:
In a branching phylogeny, by contrast, one taxonomic unit will have one parent. This is the monophyletic assumption. One parent, branching, is a clade.
Below is a proposed tree of life including Hallucigenia. The tree is a consensus of trees. Multiple trees are lying underneath the one tree shown. From numerous characteristics, different methods, and different parameters, different trees can be drawn. From a set of rival trees, different methods can be used to reduce them to one.
We can consider the spines of Hallucigenia, and the facts of life, whether the spines were an obstacle. But the fact is, no individual Hallucigenia ever had a single parent*. To graph a phylogeny, we must show the creature upside down, or backwards. This requires abstract imagination.
Parent individual Child taxon
\ /
Child individual Parent taxon
/ \
Parent individual Child taxon
A Braided Phylogenic Stream
Imagine a glacier at the top of a valley, Lee Berger asks. This is in the Nova program about the discovery of Homo naledi. He is braiding his way down to the lake where we are collected.
“So imagine in your mind a glacier at the top of a valley, and what happens is as it melts, it creates many many rivulets. And some of them are large and some of them are small. And they all move off down the valley. And almost inevitably, at the end of that valley is going to be a lake. Of which some, maybe the majority but not all, are contributing to. I think we need to look at these species we’re finding as almost individual channels in a braided stream. It’s clear they have something to do with the end population, that’s us, billions of human beings alive today, but it’s hard to tell which ones are the most responsible for us being here.”
Dawn of Humanity, PBS
Berger, his son actually, had discovered Australopithecus sediba. This species has a mix or mosaic of what are called primitive and derived traits. It had an unusual way of walking.
The combination of primitive and derived traits in Australopithecus sediba shows part of the transition from a form adapted to partial arboreality to one primarily adapted to bipedalwalking. but the legs and feet point to a previously unknown way of walking upright. With each step, Australopithecus sediba turned its foot inward with its weight focused on the outer edge of the foot. This odd way of striding may mean that upright walking evolved on more than one path during human evolution.
What does it mean to be human: Australopithecus sediba, Smithsonian
Upright walking having evolved on more than one path during human evolution does not make sense in the march of progress view.
No other known hominin walked like this, hinting that the way humans walk isn't the outcome of an ever-improving evolutionary trajectory, but one result out of several possible alternatives that evolved among our ancient relatives.
Is Australopithecus sediba the Most Important Human Ancestor Discovery Ever?, Scientific American
No one ever really believed this view. But it can still be strong enough in our minds to take advantage of an opportunity to challenge it.
And “more than one path during human evolution” is worth considering against the tree view of evolution as well.
Hybridization as Braided Stream
Imagine a population of ancestral humans migrating out of Africa, into Europe and Asia. They evolve into Neaderthal, they evolve into Denisovans.
Imagine later, a population of humans again migrating out of Africa, into Europe and Asia. There is some small amount of interbreeding with the existing population. A few introgression events.
This is a braided stream. A branching off, and a small trickle back in. A more than one path during human evolution. Something more complex than a clade.
Deep Coalescence
Imagine a river in a valley below a glacier. The point now is branching not braiding. The point is what branches when.
There are streambeds where the spring runoff flows. Islands with trees might divide them. The streambeds are species. And there are narrower channels in the summer. These are genes.
The streambeds might split, the speciation occur, further down the valley, later in time. And a gene might have split higher up the valley, and earlier. This is called deep coalescence.
Schematically represented:
In the branching river illustration, the two channels on the left diverged long before the physical trees, representing a species barrier, had forced it. And another species split happens in between. This kind of split allows incomplete linage sorting. Gene trees and species trees do not always line up.
Incomplete Lineage Sorting
Incomplete lineage sorting can produce discrepancy between the phylogenetic tree for a specific gene or a genomic segment and the overall species-level phylogenetic tree.
Comparative primate genomics: emerging patterns of genome content and dynamics, Jeffrey Rogers and Richard A. Gibbs
The glacier valley photo is from Wrangell-St. Elias National Park in Alaska. But the streambeds might be called, left to right, the Chimpanzee, the Human, the Gorilla, and the Orangutan.
The Chimpanzee and the Human are the most recently speciated. But the Gorilla and the Human are the most recently diverged, in the flow of one particular gene.
We have many genes. Different genes can flow through the streambeds on different paths.
Here, schematically, it is the gorilla and the chimpanzee that are closest in the flow of one gene. Gene and species trees again do not line up, but in a different way.
At the level of human, chimpanzee, and gorilla, incomplete lineage sorting leads to more than a quarter of gene genealogies differing from the species phylogeny.
ILS shows in the alignment of the genomes for human, chimpanzee, and gorilla, where it leads to gene genealogies different from the species phylogeny for >25% of the genome
Incomplete lineage sorting patterns among human, chimpanzee, and orangutan suggest recent orangutan speciation and widespread selection, Asger Hobolth et al
Incomplete lineage sorting can be thought of mathematically. The larger an effective population size, the longer it takes genetic variation to sort out, or working backwards in time, to coalesce. When speciation is more rapid than the sorting of genes, the sorting along species lines can be incomplete.
In the river metaphor, incomplete lineage sorting is more likely to occur if the streambed is wide. Wideness of the streambed represents effective population size. Genes flowing in a wide streambed have more freedom to wander about.
Incomplete lineage sorting is also more likely to occur if the distance between branchings is short.
Speciation as a Process Not an Event
In evolutionary biology, speciation has always been seen as a process not an event*. Representations of cladogenesis as bifurcating at a point are not in the least intended to be contrary to this.
To some extent, at least, a taxon is what taxonometric methods measure. In recent decades, taxonomy and phylogenics have more and more become computational molecular genetics by method. Molecular genetic datasets are becoming larger. The reach of molecular genetics is extending deeper into the past.
Molecular genetics can now see evidence of deep convergences, of the type that are farther back than a single speciation. The process of speciation has been extended. The start of it can be detected deeper in the past.
It is hard to see in the following phylogeny, that the speciation process of humans and chimpanzees started before the splitting off of gorillas. That it started before the splitting off of orangutans.
That conceivably, the first tiny swirl or eddy of the speciation process dividing humans and chimpanzees started before the splitting off of gibbons.
That gene flows had already started then, which would later end up going more to the chimpanzee or the human side.
There are deep convergences with flows separately joining humans and chimpanzees, and humans and bonobos. The form of representation does not make these separate flows easy to imagine, coursing in parallel along the thin line before the chimpanzee bonobo speciation. It is not easy to imagine, in this form of representation, that the speciation process of those two had really started further upstream.
The Unclear Phylogeny of Bees
It is hard to make a hierarchical phylogeny of the seven families of bees. Some 60 different attempts have been published.
The differentiation occurred rapidly, and long ago. Early incomplete lineage sorting and hybridization obscure our view.
[T]he uncertainty of early bee history appears to be due, in part, to short branches among families. This suggests that major lineage differentiation occurred within a relatively short amount of time early in bee history. Incomplete lineage sorting or hybridization between lineages early in bee history could also obscure bifurcations.
The bee tree of life: a supermatrix approach to apoid phylogeny and biogeography, Shannon Hedke et al
This study speaks in metaphor, about a bee tree of life. It speaks in computer and information science terms, of sparse supermatrixes and of signal strengths. It speaks of biogeography.
By overall approach, taking advantage of the large DNA datasets now available, to address longstanding phylogenic controversies about the long ago, it ends up speaking in probabilistic terms.
There is talk of thinking of phylogeny, not as nodes and branches, but as probabilistic clouds.
The Adaptive Radiation of Birds
Alexander Suh et al looked at the diversification of neoavian birds. We might be reminded here of the adaptive radiation into ecological niches of Darwin’s Finches, after landing in the Galapagos.
Astoundingly, we found that near the K-Pg boundary, speciation events were accompanied by extreme levels of ILS, suggesting a near-simultaneous, star-like diversification process that appears plausible in the context of instantaneous niche availability that must have followed the K-Pg mass extinction.
The Dynamics of Incomplete Lineage Sorting across the Ancient Adaptive Radiation of Neoavian Birds, Alexander Suh et al
Given high to extreme levels of incomplete complete lineage sorting with rapid speciation, bifurcating trees are not a good way to visualize the relationships.
We provide evidence that, given the extreme level of up to 100% ILS per branch in super-radiations, particularly rapid speciation events may neither resemble a fully bifurcating tree nor are they resolvable as such. As a consequence, their complex demographic history is more accurately represented as local networks within a species tree.
The radiation of birds might look like this:
Don’t be overly mislead by the nice messy neighbor-networks of bird adaptive radiations, though.
The study is looking at a limited set of genetic markers. But there are many genes and many ways of looking at them, and other hierarchies beyond that. Many different networks might be drawn.
Because life is organized hierarchically, reticulation can occur at different levels: chromosomes, genomes and species.
Network (Reticulate) Evolution: Biology, Models, and Algorithms, Randal Linder et al
This is just a simplified drawing, a marching down many paths, to help us make a sense of the bewildering array of it all.