In the last post I pointed out that when studying evolutionary history we take the same approach as do all scientists: we try to explain observable natural phenomena in terms of other natural phenomena. In this case, we extrapolate what we observe in human genealogies back over hundreds of millions of years, to explain the history of species, which we cannot observe directly.
One large problem with this approach is that using human genealogies (family trees) as a model for species phylogenies is not necessarily a good thing. We need to think carefully about what phylogenies might look like, and thus how to describe and display phylogenetic history.
Perhaps the most important aspect of this problem is confusion about the words "evolution" and "phylogeny". Rather unfortunately, in biology the word evolution has two quite distinct meanings, whereas it usually has only one of those two meanings outside of biology. I am not sure that non-biologists realize the importance of this fact. Both of the meanings are involved in a phylogeny.
Outside of biology (e.g. in physics), "evolution" refers to a change in an object through time. That is, there is a (possibly predictable) transformation of the object through two or more distinctly different states.
For example, we might talk about the evolution of a star, which changes from a main sequence star through the stages of sub-giant, red giant and finally white dwarf. Alternatively, we might refer to the evolution of a website's homepage, as discussed by Ann Smarty at this link. We might even read about the evolution of the Haggunenons, those beings who, in the work of Douglas Adams, could change into armchairs, escape capsules and even other species.
In biology, the most obvious transformational changes occur during our own lifetimes, where we start out as infants and then progress through childhood, adolescence, adulthood and old age. However, biologists rarely call this "evolution", even though that is exactly what it would be called by non-biologists. (Biologists call it "ontogeny".)
In biology, "evolution" usually refers to variational change rather than transformational change. Here, a group of objects appear to change because some of the objects survive and some do not. Each object differs from the other objects in some way but each also remains unchanged. The group as a whole changes because through time the proportion of the different types of object changes. There is a change in the group without a change in any one object within the group.
In biology, the most obvious variational changes occur in domesticated organisms. Over the millennia, we have selected specific variants of dogs and cats and roses, for example, and those variants are now more common than they were. So, modern dogs, cats and roses look very different from what we know dogs, cats and roses looked like several hundred years ago, even though each individual dog, cat and rose remained the same for its whole life. What has changed is the group rather than any member of the group.
What is equally important is the obvious increase in the biodiversity of domesticated organisms. There are now more types of dogs, cats and roses (and cows, horses and apples) than there were in the past. That is, a single type of ancestor has "evolved" into many types of descendants. This is a branching process, just as in a family tree.
So, variational evolution is a branching process whereas transformational evolution is a linear (non-branching) process. This needs to be represented in a phylogeny.
In addition, some variants do not survive. Genealogical lineages come to an end when there are no descendants left. This happens repeatedly in human pedigrees; and the fossil record is usually interpreted as showing that the same thing happens for species, too. Indeed, geological time-periods are actually defined by dramatic changes in the fossil record, indicating what are referred to as "mass extinctions".
So, variational evolution is a process of continual branching and extinction, apparently without end, whereas transformational evolution is a simple linear process, usually leading to a predictable end. This is what a phylogeny must represent.
The idea that evolution in biology is a more complex process than that in non-biology is usually credited to Charles Darwin, although Alfred Russel Wallace independently developed the same idea at the same time. Darwin did not use the word "evolution" in the first edition of The Origin of Species (it made its appearance only in the 6th and final edition, in 1876), which suggests that he could see the potential confusion between the ideas of transformational evolution and variational evolution. The subsequent history of biology shows that we would all have been wise to follow this lead, as the confusion has been widespread and irksome.
Nevertheless, in a phylogeny we might also like to represent transformational evolution as well as variational evolution. For instance, if we treat each species as a separate object (instead of a collection of individuals) then if a single species changes through time, without any branching or extinction, then this would look like transformational evolution. This was the idea of Jean-Baptiste Lamarck, who didn't believe in extinction of species but instead thought that each biological form transformed into another form through time. Thus, Lamarck's version of evolution was essentially a transformation series among species, with occasional branching into new forms.
Where does this leave us? Well, there is a lot of potential confusion here, with two different types of evolution, and the differences between family trees and species phylogenies. One possible solution is to drop all of these terms, and to create a model of a phylogeny using new words with explicit meanings. This is basically what phylogeneticists have done, as shown in the next figure.
Cladogenesis refers to the branching process of evolution, represented by the diverging lines in the diagram.
Phylesis (or sometimes also called anagenesis) refers to unbranched evolutionary change through time, represented by the length of the lines (longer lines = more change).
Taxon (the plural is taxa) refers to each observed object in the phylogeny, which may be a species or a genus or even an individual.
The phylogeny is the unobserved history connecting the taxa, which consists of a series of inferred ancestors, linked to their descendants, leading backwards from the observed taxa to their inferred common ancestor.
Note that the diagram does go from the past to the present (or vice versa), but the branch lengths do not represent time. Thus, the fact that the branches end at different places does not indicate that some of the taxa became extinct — they are all observed taxa that exist in the modern world.
Hopefully, it is clear from the phylogeny that the closest relative of Taxon 2 is Taxon 3. We might think of them as "sisters" (or siblings, if you want a neutral term), because they share the same "parent". Taxon 1 is a "cousin" to the other two taxa, because it shares their "grandparent" (but not their parent), in the same way that we interpret a family tree. This is the information that a phylogeny contains — it shows who is the closest relative to whom, by indicating the branching order of their genealogical history.
Here is an example of a real phylogeny for eight animal species, which makes it more concrete. This history was constructed from the DNA sequence of a single gene, which explains why the fruitfly (Drosophila) branch is longer than the others (it would be shorter than some of the others if all of the genes were included). Do not be confused by the fact that the phylogeny proceeds from bottom to top, whereas the previous one proceeds from left to right. The information is in the branching pattern and the line lengths, not in which way round it is drawn.
Note that the branching order shows that Rat and Mouse are the closest relatives. Rat is not historically derived from Mouse, and nor is Mouse historically derived from Rat — instead, they are derived from a common parental species (or ancestor). Human is shown as the next closest relative, sharing an ancestor with both Rat and Mouse. Obviously, there are a lot of other mammals that are not included in the diagram (such as the dog, cat, cow and horse referred to above), but we would expect them to also appear in this part of the diagram. This interpretation of the diagram proceeds until we get to the ancestor that all eight animals share, called their most recent common ancestor.
The phylogeny does not end there, of course. All organisms have a history, and so all of them fit into the phylogeny somewhere (including the rose and apple referred to above). The phylogeny simply extends further and further back in time, including more and more species as it goes. Any one diagram only includes a subset of the organisms, for convenience' sake if nothing else.
This is what a phylogeny communicates, and nothing more — the historical information is in the branching pattern (which tells you about the relationships) and the line lengths (which tell you about the amount of evolutionary change).
The problem is what happens when people try to add more to the phylogeny, or to read more into it than is really there. This is covered in the next post, which will be about what happens if we confuse transformational and variational evolution.