Here is a sample phylogenetic tree from a textbook publisher:

enter image description here

We often ask questions like "Which organism is most closely related to the pig, the hedgehog or the cow?" I don't like these particular questions, and maybe that's because I misunderstand what "most closely related" means in phylogeny.

I have been told and have read that the most closest relatives share the most recent common ancestor. For my question, we would have to say the cow is the closest relative of the pig. My gut feeling says we cannot say that is truly the case unless we have sufficient DNA evidence. For example, what if we compare the pig and hedgehog genomes and find, say, 80% match. We might then compare the pig and cow genomes and find, say, a 78% match. Would this not mean that the hedgehog is more closely related to the pig than the cow is to the pig, despite the heuristics we have been handing students about reading phylogenetic trees?

Here is an extreme example to make my case: Is the oldest species of archaea more closely related to humans or to the species of bacteria which most recently preceded the archaea? If we use the "most recent common ancestor" answer, we must say the archaea is more closely related to the human, but I would find its temporal proximity (and thus likely % of shared DNA) to the bacterium to be a much more convincing argument for relatedness.

Question: How should we interpret and approach these sorts of questions, or are they bad questions due to their lack of context?

  • 1
    $\begingroup$ Can you draw out a simplified example where the pattern you propose would occur? Perhaps using just one gene with a short sequence as an example, or imagining genes as something that are collected over time in a lineage the way someone collects rocks. That is, start with a common ancestor of hedgehogs, pigs, and cows, which then goes through changes to result in the hedgehog lineage on one side and the pig and cow on the other, which then branch again to separate pig and cow lineages. $\endgroup$
    – Bryan Krause
    Commented Mar 11 at 19:18
  • $\begingroup$ The phylogenetic trees are constructed based on a model that links observations with an ancestral relationship. So I think you need to specify which evolutionary sequence you're considering (as Bryan Krause says above), and which types of observations you're interested in correlating to it. Your question focuses on genetic sequences as the main observations. The whole "Woesean revolution" that established the three domains was based on overturning an older phyologeny based on sequence data. That our lineage includes endosymbiosed bacteria is a different fact of evolution. $\endgroup$
    – Ryan
    Commented Mar 12 at 18:22
  • $\begingroup$ @BryanKrause I do not entirely understand your request. Are you and Ryan saying that a phylogenetic tree is based on only one observation at a time, like only one gene? Regardless, I'm not exactly sure what it practically means to say "the cow is the closest relative of the pig" in any phylogenetic tree. What traits, observations, or characteristics are we referring to when we make such a statement? $\endgroup$ Commented Mar 12 at 21:46
  • $\begingroup$ I'm saying that this is how things actually diverge, one step at a time. So, if you know what the "family tree" looks like (from your picture), I'm challenging you to produce something that looks like your hypothetical numerical relationship. Of course not exactly "78%" or "80%", but I'm suggesting that you run through a hypothetical stepwise process, starting from the common ancestor, and see what you need to do to get that result where you would "calculate" a closer relationship between species not shown as closer according to the tree. $\endgroup$
    – Bryan Krause
    Commented Mar 12 at 21:49
  • $\begingroup$ You can do this with a totally made up and hypothetical case; you could start with a gene like AAABBBCCCDDD if you wanted and then change, add, or remove letters one step at a time, or you could use traits/attributes as long as you modify them stepwise. Re-create something like the process that would create these relatives as long as you follow that branching pattern. $\endgroup$
    – Bryan Krause
    Commented Mar 12 at 21:51

2 Answers 2


Phylogenies are an attempt to answer precisely the question of what things are most similar, where "similar" can mean one of a number of things. In statistical terms, a phylogeny is a hierarchical clustering of available organisms based on some similarity criterion. This attempts to make use of second-order information to make guesses about closest relatives based on more than just pairwise similarity; pairwise similarity can give you incorrect answers due to convergent evolution / homoplasy.

A phylogeny based on mitochondrial DNA only will differ from a phylogeny based on full proteome sequences, will differ from a phylogeny based on morphological characters, and so on and so forth. These are all different measures of similarity with advantages and disadvantages. We can further pass these data through one or more models of how the data in question might evolve through the tree. These might include the maximum parsimony model based on Jukes-Cantor distance, maximum likelihood based on the F84 model, or something else.

I believe that the specific measurement that you would be most interested in is the branch length. Branches in phylogenetic cladograms are calibrated to show evolutionary distances according to the evolutionary model fit to the available data.

If you are interested in the nearest relative, you may therefore simply measure the length of the branches differentiating any two tips of the tree. Whatever path is shortest from a tip (organism) $A$ to any other tip will be the nearest relative of tip $A$.

This provides an unambiguous method of identifying a closest relative in a tree. Of course, it does not account for organisms that may be missing from the tree that you consider!

For a clearer guide on reading phylogenetic cladograms, I would suggest this resource.


The common scientific figure of speech is "The closest living relative", because the species are like a puzzling family tree, a phylogenetic tree. We can also theorize about a well documented fossil as "the closest known relative".

We can also use close relatives in a plural and general sense:

The closest living relatives of hyenas are mongooses and civets.

The red panda is most closely related to skunks and raccoons.

Armadillos are most closely related to sloths and anteaters.

It's an interesting and practical way to reason about the tree of life. It is best measured with genetic similarity.

Some species have mysterious ancestry, sea cows, red pandas, hyenas, and so it's useful and interesting to know what family of species they belong to.

The notion of relatedness can be referred to as genetic distance. To research biology trees on the web it's usually best to search images i.e. "amphibian phylogeny"

This notion indicates the evolutionary relationship and genetic similarity between organisms, highlighting their proximity on the evolutionary tree.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .