A phenotypic view of evolution Evolution in Structured Populations

A quick review of the phenotypic perspective, pt 2

Continuing on with my brief reprise of the tenets of this blog:

Individuals are the level at which we assign fitness.

First a parable. If we go back to Mayr, he thought that the species was the only natural unit of organization, and argued that the biological species concept (BSC) was the correct definition of species. Since then we have realized that species are perhaps not that monolithic. There is gene flow between species, and it is not always clear whether two different populations are the same or different species.

I argue that the same is true of individuals. (part 1, part 2, part 3)(Not so long ago it was easy to argue that individuals were “organisms”, and that they were a natural unit. The problem, as always, comes when we look too close. Just like the BSC works until we look to closely, this classic definition of individual breaks down when we look closely. Evolution is always defined in a way that precludes changes within an individual, usually called development. Thus, evolutionary biologists are adamant that the change in an individual as it grows from zygote to adult is not evolution. On the other hand we would like to study distinctly cellular level processes, such as cancer development, as evolution. Similar problems come when we study proto-multicellular organisms like volvox, or organisms that grow clonally, such as Aspen. On the one hand a clean definition of “individual” as organism works fine when we are talking about vertebrates and not looking too closely, ignoring that the vertebrate organism is a multispecies entity. However, on the other hand, such a narrow definition results in our narrowing our definition of evolution to changes at or above the organismic level. Something has to give, and I would argue that the correct thing to do is to allow flexibility on what we call an individual.


Volvox: what is the individual? (http://www.dr-ralf-wagner.de/Bilder/Volvox-aureus-DF.jpg)

Allowing a flexible definition of individual has huge advantages. We can study evolution at any level. That is it IS legitimate to study cancer as an evolutionary process. It IS legitimate to study “species selection” in the fossil record even though you do not have access to the fitness of the underlying organisms. And, when we discover that the microbiome was not something that could be simply ignored we don’t have to through out all of the evolutionary studies that have been done previously. This last point is an important one. Our understanding of biology is expanding rapidly. We need concepts and models that can incorporate these new findings naturally without having to re-do our entire conceptual system. The classic concept of individual as a colony of genetically identical cells of the same species physically separated from other colonies (I made that definition up) is simply too rigid, and cannot be adapted to our expanding understanding of living systems.

The down side is also analogous to issues with species concepts. It is now understood that, in order to avoid confusion, when you talk about species and speciation you need to explicitly state what you mean by species. As with species concepts, our understanding of what is and is not evolution changes dramatically depending on the level at which we define fitness. Everything that is at or above the level of the individual can be studied as evolution. Everything below the level of the individual has to be studied as some form of development. This means that when Gardner argues that cancer cannot be studied as evolution (Gardner’s paper; My critique of Gardner; Gardner’s critique of me) and I argue that it can we are both right. In his view the individual is rigidly defined as the organism, and as such cancer must be treated as changes occurring within the organism. He can call it development, disease, what ever, but he cannot call it evolution. I, on the other hand, by embracing a more flexible definition of the individual, CAN call it evolution. What I am doing is defining the cell to be the individual, then the faster cell division of cancer cells is selection. This is opposed by “group” selection in that, with a few notable exceptions, the spread of a cancer is ultimately stopped by the death of the organism.

cancer and selection

Cancer as an evolutionary process (from http://www.nature.com/nature/journal/v481/n7381/full/nature10762.html)

Multilevel selection:

Viewed from the perspective of a flexible definition of the individual, depending on the level at which you assign fitness potentially ALL selection is multilevel selection. Classic “individual selection” is selection on organisms, which is a group of cells, and actually on a community of cells. It is group selection if we assign fitness at the level of the cell, it is individual selection if we assign fitness at the level of the organism. Given that genic view “individual” selection is actually group selection, and is nearly ubiquitous, is it really surprising that “group selection” works at other levels as well?

This raises a second point. The phenotypic approach follows the lead of quantitative genetics, and makes a sharp distinction between selection and the response to selection. Selection is the ecological process by which some entities leave more offspring than other entities. The response to selection is the evolutionary consequences of that. Organismal selection is the differential survival and reproduction of individuals. The evolutionary consequence is a difference in the distribution of offspring that is due to that differential survival and reproduction of the parents. The point is, regardless of the consequences, the level of selection is the level at which selection acts. Thus, if both individual selection and group selection cause exactly the same change in gene frequency, they are still different things, because they are different ecological processes. If I dye a shirt red it may be the same color as if I wove it out of red thread, but nobody would pretend that dying fabric and weaving fabric are the same thing, even if the outcome was the same. Of course, as with dying versus weaving fabric, selection at different levels has qualitatively different consequences. The result is that we do our understanding of evolution a huge disservice when we decide to dismiss levels of selection a priori. This is especially true when we realize that whether or not selection is acting at a particular level is an empirical question, and that we have the tools we need to answer that question.


Dyed or woven, you’re still dead. (from http://anartistcalledred.deviantart.com/art/Curse-of-the-RedShirt-173225300)


Leave a comment

Your email address will not be published.

Skip to toolbar