Understanding Phylogenies: Agh! There’s a polytomy in my data!

In the first Understanding Phylogenies post, I outlined some basic phylogenetic terminology using a candy bar phylogeny. The main point being that the layout (aka- topology) of branches and nodes illustrated how candy bars (aka- taxa) were related to each other based on their traits (in this case candy shape or flavor). However, you may have noticed a number of spurious relationships.

For example (Figure 3A):
• Are York Peppermint Patties and Cherry Mash really sister taxa just because they are round?
• What unites the peanut and coconut clades as sister taxa?
• In the peanut clade, is peanut butter (Reece’s Peanut Butter Cups) sister to peanuts (PayDay) or to “peanut butter-y” filling (Butterfinger)?
• What is the common ancestor between “Drops” and “Flat Bar” candies?

The answer is that some of the relationships in the original tree don’t make sense. I forced the taxa into a fully bifurcating tree (Figure 3A). However, we can display ambiguity by not forcing the tree to bifurcate and allow polytomies on the tree (Figure 3B). Polytomies can either represent uncertainty in the data or a lack of information. Discerning whether a polytomy represents uncertainty or lack of information is up to the researcher and their knowledge of the system and the data. They may be able to collect more data (ex- extra taxa or more character traits [or for molecular phylogenies, additional sequence data]) to resolve the tree. However, some polytomies cannot be resolved and simply reflect uncertainty which researchers must account for as they move forward with their analyses.

Figure 3- Two versions of the candy bar phylogeny: fully bifurcating (A) and with polytomies (B).

The weight given to the traits contributes to the difference between the bifurcating (Figure 3A) and polytomy (Figure 3B) trees. When candy shape was considered (Figure 3A), flavor groups taxa near the tips but shape groups clades closer to the root. However, when shape was not considered, groupings were only done by flavor and there was insufficient data to fully bifurcate the tree near the root as some clades did not share multiple flavor profiles.

While polytomies may be frustrating because all of the relationships between taxa are not described, they are a very real part of building and interpreting phylogenies. Therefore, polytomies are also a real part of interpreting evolutionary patterns. (You can see many polytomies across the Tree of Life on this interactive website.)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s