Natural selection is often described as “survival of the fittest.” But fittest is an adjective–what are the fittest things that are doing the surviving? The knee-jerk answer most of the time will be individual organisms, and this is true, but perhaps not the complete truth. More on the flip.
I should mention at this point that much of this post is based on the writings of Richard Dawkins; the books of his I’ve read are all good, including the most recent one, the Ancestor’s Tale, and should be accessible to anyone with basic knowledge (undergrad or good high school level) of biology.
So, what does natural selection operate on, if not organisms? One possibility is genes. To see what I am talking about, consider a thought experiment. Suppose that genes could be copied and distributed to others at will, perhaps for a price, perhaps out of the goodness of your heart. In this exchangeable situation, the best genes would become very common very quickly; in other words, they would undergo positive selection, and multiply. Defective mutant genes would go away–who would want the gene for hemophilia? Or colon cancer? Not I.
Now, in reality, you can’t replace your genes with other ones, but you have some choice as to which genes your children get through your choice of mate. Of course, genes are generally only revealed through phenotype (outwardly observable characteristics) unless you happen to have a DNA sequencer handy and your potential mates submit to screening. So, to some extent, you select the genes, or the set of genes, that gets passed on to part of the next generation.
What about organisms that reproduce asexually, in which each offspring is essentially a clone of its parent? Here, there isn’t as much choice involved, but to some extent gene selection is still occurring in that it is the combination of genes and environment that determines whether an organism is successful/fit. Gene redistribution just will not occur as quickly in this situation most of the time.
These ideas have been around for a while–I first heard of them in high school back in the early 90s. What I just became aware of (in the Ancestor’s Tale) is something called clade selection, which is likely related to the evolution of evolvability–the ability of a lineage to change (adapt) more rapidly over time. This is clearly a kind of selection that cannot operate on a single organism.
I may have lost some of you in that last paragraph, so let me go through it a little more slowly. A clade is a group of all the species descended from a common ancestor. Mammals are a clade, as are insects. Animals are a clade too, that contains many smaller clades (however, reptiles are not a clade, but reptiles plus birds plus mammals are; mammals and birds descend from reptiles). The idea is that the clades most able to adapt to changing situations (usually on geological time, but sometimes there are earth-shattering events like the one that wiped out the dinosaurs) will expand, or radiate, out and become more common. So, you see lots of different mammals, because mammals were well-positioned to fill the niches made available when the dinosaurs vanished. The ability of mammals to change rapidly under appropriate selection can be seen with dogs, who admittedly have undergone intense artificial selection, but show an absolutely amazing range of diversity that appeared in a matter of a few centuries–from ratlike chihuahuas to big great danes, which in the absence of humans could reasonably be expected to diverge into different species if given enough time (assuming they survived).
What makes a clade more or less adaptable? I think a lot of the answer to this is the absence of limitations at near-fundamental levels. Compare mammals and insects. Mammals are warm blooded (endothermic) with internal skeletons, while insects have exoskeletons and are exothermic. Neither one of these is easy to change, although obviously it may have happened in the past. (With endo/exothermy, it must have; the skeletons may have evolved separately from an original wormlike ancestor.) However, endothermy gives a lower limit to size–if you’re too small, the rate of heat loss becomes unsustainable. Exoskeletons have the opposite effect–it’s hard to grow because you’re stuck inside this case of armor, and if you want to get bigger, you have to get rid of it and make a new one, which is costly. So, mammals smaller than mice presumably don’t do well, but insects don’t get as large as mice, perhaps with a very few exceptions (I hear there are some big cockroaches in Hawaii). Reptiles have neither limitation, and so you see lizards smaller than mice, and also big things like crocodiles. However, it looks like at the ant-size range, there are no reptiles, and perhaps this is because insects are too effective and would outcompete any newcomers, or perhaps there are other aspects of reptile physiology or anatomy that become problematic. I believe exoskeletons are more efficient in terms of promoting muscle strength, so that may be part of it.
One of the things Dawkins says about the historical trajectory by which animals became more evolvable is that the appearance of segmentation was likely a watershed. If the number of segments is expanded, they can then be diversified; a simple example is the difference between various segments of the human vertebral column. This also has a pleasing analogy to conventional evolution, in which new genes often arise through accidental duplication, leading to specialized variants; there are multiple hemoglobin genes in our genomes, some of which are expressed only in the embryo and designed to outcompete maternal (adult) hemoglobin for oxygen, hence allowing “breathing” through the placenta. Similarly, muscle myoglobin is more distantly related to hemoglobin but is also able to grab oxygen from the blood.
One last example which may surprise you is DNA. It is now thought that the original life forms that arose billions of years ago did not use DNA as their genetic material; one leading theory is that RNA was the original molecule, which is attractive because RNA also can catalyze chemical reactions, including potentially its own replication. So, the first life may have been RNA-based, but when DNA came on the scene (or more generally, the genetic material was separated from the bulk of cellular metabolism) this was a huge deal–DNA is a better genetic molecule than RNA because it forms a much more regular structure (conversely, the lack of that regularity is why RNA can be a catalyst for multiple different reactions) and the regularity of DNA means that mutations can occur that do not adversely affect its structure. In essence, DNA opened up a much broader range of possible diversifications, meaning the organisms that used it were much more adaptable.
So, in conclusion, it now appears that natural selection operates on multiple levels: genes, organisms, and clades. I’m sure more ways in which evolvability has been improved will be brought to our attention in the near future.
If you are so inclined, I would be interested to hear if anyone has ideas about what other features of <insert organism here> make it particularly evolvable/adaptable.