Let's Talk Phylogenetics

In my last post, I explained how the timely emergence of the white rot fungi during the Carboniferous period corrected an imbalance caused by the evolution of wood. I did not, however, discuss a big question related to that story: how do we know all of these details? The answer is that we are able to peer back through time using a method called phylogenetics, which makes comparisons using existing species to make inferences about the past. In this post, I'll explain how phylogenetics works, and in my next post I will explain how this strategy was used to understand when and how white rot fungi evolved.  

What a phylogenetic approach does is show us how closely a group of species is related. Say we take a human being, a chimpanzee, and a turtle, and we want to understand how they're related to one another.

The first step will be to choose an outgroup, a species that must be more distantly related to any of the members of our target sample than they are to each other, but still related somehow. In brief, we need something really weird, but not so weird we couldn't have it for dinner. Like, for example, a shrimp. Right off the bat, we can see that chimps, humans, and turtles have backbones and shrimp do not, but shrimp still have a few big similarities – they are symmetrical only one way (as opposed to a starfish) and need to eat. This is a sign we've got a good outgroup.

Now that we have our sample and our outgroup, the first thing to do is to total up the differences between each species using what we know about them based on observation. Let's take humans and chimpanzees:

This is all I can think about at the moment, so we'll say there are four differences between humans and chimpanzees. Let's move on to humans and turtles: 

There are six differences between turtles and humans. Chimps have all the traits that I mentioned for humans, except for the fact that, like turtles, they don't speak a language, so their number of differences is five.

Finally, let's take the shrimp:

It looks like there are 10 differences between humans and shrimps. Chimps have all the same differences that the humans have except for the ability to speak, so we'll give them a 9. Turtles, furthermore, do lay eggs and are cold blooded, so we'll say that there are 7 differences between turtles and shrimps.

The third step is to make a chart of all our data representing the number of differences we counted:

This is slightly easier to read than a meandering paragraph, but what we really want to do is represent these data graphically, so the next step is to construct a cladogram (a tree). To do this, we start with a star tree that looks like this:

The star tree represents the idea we want to disprove, called a null hypothesis. In this case, our null hypothesis is that all four of our species descended from the same presumably funny-looking common ancestor (marked by the red dot) simultaneously. This is obviously not true, which we can demonstrate using our table of differences: humans and chimpanzees have the fewest number of differences, so they must be the most closely related and therefore have a more recent common ancestor that they don't share with shrimp or turtles. We therefore join the human and chimpanzee branches to form a clade (a group of species and their most recent common ancestor or MRA), like so.

We'll call this the humanzee clade for shorthand. The next step is to find out which species (turtles or shrimps) is more closely related to the humanzee clade. This requires redrawing our table:

This table lumps together humans and chimpanzees, and so the number of differences between the humanzee clade and the other species is simply an average of the individual scores of humans and chimpanzees. Based on our table, turtles are more closely related to the humanzee group than shrimps, so we redraw our tree to reflect this information:

Now we have a humanzertle clade (with the ancestor marked by the first red dot) along with our outgroup, the shrimp, which completes the tree. From our data, we can now conclude that, out of the species we analyzed humans are most closely related to chimps and most distantly related to shrimps. More importantly, we can say that humans diverged from shrimps earlier in evolutionary history and chimps much later. This was the basic principle used to trace the lineage of the fungi I discussed in my previous post.

“But wait!” you might be inclined to interject - “I can think of way more differences between humans and chimpanzees, or chimpanzees and turtles, and some of the differences you listed were incredibly iffy.” This is very true, which is why physical and behavioral traits are no longer used to draw phylogenies. Instead, we use DNA. DNA analysis has several advantages over doing things the old way: for one, each base pair can represent a character or “trait”, meaning we can compare between thousands to millions of characters at a time. Second, this comparison is objective rather than subjective. Third, we can delegate this task to computers, which makes this process much quicker and easier than sitting down and thinking about all the possible differences between organisms. The result is incredibly detailed and much more powerful phylogenies.

Tune in next time, and I'll explain how phylogenetics helped us discover the evolutionary path of the white rot fungi, and how we can say with certainty that they played such a crucial role in the ecology of the late Carboniferous period.

Andrew Tomes is an MS student at SUNY-ESF in Syracuse whose work focuses on the mycorrhizal partners of the American chestnut. He earned his B.S. in Botany from the University of Maine, where he worked on wetland restoration. When not in the field, the lab, or the classroom, he can often be found in the kitchen baking bread.