Phylogenetic trees can tell us about more than just the hierarchical order of our dataset.
Taken together the position of internal nodes, the number of edges descending from a node and the edge lengths in the tree can be very informative.
Position of nodes
Below are some examples of tree shapes, note the node positions and the number of edges descending from each node:
Bifurcating
In a bifurcating tree each node splits into exactly two edges. The example above has two internal nodes, where each node splits into two edges, this is an example of a balanced tree.
Multifurcating
In a multifurcating tree a node can split into more than two edges e.g. you can find groups within a tree e.g (C,B,A) in the Multifurcating example or the entire tree can be multifurcating eg. in the Star example above.
Groups in Multifurcating trees that have more than two edges descending from a single node are called polytomies. A polytomy can be a result of:
1) Low phylogenetic resolution – this indicates there is not enough information in your data to resolve the true relationships of your taxa (otherwise known as a ‘soft polytomy‘).
2) The rapid expansion of a species or population at the same time – where the lack of diversity observed is a result of simultaneous divergence of species or populations (this is known as a ‘hard polytomy‘).
Ladder
A ‘ladder-like’, ‘pectinate’ or ‘comb-like‘ tree is an unbalanced tree shape.
The phylogenies of viral populations can be ladder-like, often the result of continual immune selection within a host, for example below is a phylogeny of Influenza virus samples:
When we are considering species phylogenies a ladder-like tree usually indicates speciation via anagenesis (see Glossary).
Tree shapes and population dynamics
Population phylogenies can provide other interesting clues of historic population dynamics such as how many lineages are diverging at different evolutionary times.
The examples below (A, B and C) represent characteristic population dynamic tree shapes, underneath are their associated lineage through time plots:
In these unlabelled tree diagrams (unlabelled because we have not assigned specific tip or node labels) we can recognise 3 population dynamic shapes, representing:
- (A) Even rates of evolution through time – the null hypothesis
- (B) Early burst of cladogenesis(from ecological opportunity)
- (C)Early extinction events or late burst of speciation.
(A) Even rates of evolution through time
The first example represents a population where diversification events are not overly skewed towards the root or the tips.
This is also known as the Yule model, which is the null hypothesis where ‘every lineage is equally likely to speciate at any given time’, in this case we end up with a completely balanced tree.
(B) Early burst of cladogenesis
This second example illustrates an early burst of cladogenesis, in this case many lineages arise at one time and we can see these nodes are situated closer to the root of the tree.
This tree shape usually indicates a rapid divergence of lineages in a short space of time as a result of ecological opportunities, this is also known as an adaptive radiation.
(C) Early extinction events or late burst of speciation
This last example indicate a late burst of speciation or early extinction of lineages where nodes are primarily situated towards the tips of the tree.
Something to remember
As mentioned in a previous post if we are looking at a phylogram we can only compare the evolutionary change in a single ancestral lineage at a time, we cannot make comparisons between different lineages.
Different lineages can have different rates of evolutionary change and so cannot be directly compared.
A real example!
It may surprise you to learn that many of our familiar veg fall within a single genus Brassica.
For example Cabbage, Broccoli, Cauliflower and Kale are all considered morphotypes of the same species Brassica oleracea.
10.3389/fpls.2022.932288
As we saw earlier, when considering multiple subspecies or populations tree shape can reveal ancient or recent diversification events to give us an overall idea of the complex evolutionary history of subpopulations within a species.
Before we look at a specific example let’s remind ourselves of some characteristic tree shapes:
Evolutionary history of Brassica rapa
In a study by Bird et al.,2017 individuals belonging to the Brassica genus were sampled from different geographic localities.
Below is a phylogenetic tree which has clustered these individuals quite nicely into subpopulations using genetic data.
The authors have coloured the subpopulations in blue (European turnip/ European oilseed), purple (Asian turnip, South Asian oilseed), green (yellow sarson/brown sarson), orange (Chinese cabbage), and red (bok choy, tatsoi, choy sum).
This tree contains a huge amount of information but we will focus in on a few interesting things going on here.
Labelling order
When morphological and genetic data of a species are reliably consistent we expect individuals from the same species to cluster togethor in the tree i.e we already have an expected labelling order before we build the tree.
In the example above we might expect to see all individuals from each subspecies of Brassica cluster within their own subspecies groups. On the whole this is what we observe (see Fig 7).
However, there are also a few polyphyletic occurrences.
To recap, a polyphyly, is when organisms with mixed evolutionary origin are placed in the same clade of a phylogenetic tree.
In Fig 7. above we can see some individuals identified as B.dichotoma spp. and B.oleifera spp. actually cluster within B.rappa clades (their placement is indicated by star symbols).
In short, some individuals are not clustered in the groups we might expect.
This could suggest that the species definitions in Brassica need to be revised or we may have individuals that were mis-identified.
The authors discuss this however and conclude that ‘the usage of oleifera as a subspecies seems inherently problematic’ so it is possible that Brassica has an even looser species definition than we first thought, uh oh!!
Tree balance
The B.rapa (Bok Choy) and B.rapa (Chinese cabbage) individuals cluster in the more balanced regions of the tree, in contrast the B.rapa (Turnip) clades are much more imbalanced, suggesting that there has not been an even rate of evolution in this group over time.
Diversification timing
There are a range of diversification patterns across the tree but we find that the Asian and European B.rapa (Turnip) clades deviate the furthest from a completely balanced tree shape.
Most of the nodes in the European clade (coloured in blue) are situated towards the tips, with late diversification times.
This could indicate early extinction events of the B.rapa (Turnip) morphotype in Europe or that there has been a late burst of speciation.
There is also a well supported division between European and Asian lineages.
If these populations had interbred, we would see a mixture of individuals from different geographic regions in the same clade, suggesting these geographic populations may have been historically isolated from each other for some time.
Note that the European sample does appear to be smaller than the other geographic samples – so you would need to sequence more individuals for a more definitive answer.
I hope you enjoyed this quick introduction to decoding tree shapes!