A conversation with Aneil Agrawal about how evolution works in a fragmented population.
Paul Craze: Okay, so one of the themes of this year’s European Society for Evolutionary Biology meeting is how evolution and selection works in subdivided or fragmented populations. So, populations where the organisms are in small groups, something like fragments of woodland in agricultural landscape where the species that exist in the woodlands are fragmented and they’re not in a single population. That can influence how evolution works, and one of the people who has been looking at that is Aneil Agrawal from the University of Toronto, and you’ve been looking at how deleterious mutations are affected by this subdivided population structure, so what are these deleterious mutations and why are they important?
Aneil Agrawal: Well a deleterious mutation is any mutation that reduces the fitness of an organism, and we know that the vast majority of mutations that affect fitness at all are deleterious. Now, one might tend to think that these things are going to be removed by selection, and they tend to be, but they’re constantly occurring, so all populations contain deleterious mutations at some frequency. And we, as humans, are well aware of this and we all know people that have genetic diseases that are obvious, but also we know that there are genetic factors that contribute in more subtle ways to things like problems with eyesight or high blood pressure or obesity, etc.
Paul: Okay. So these always occur in populations. So how does the structure of those populations, how does this fragmentation of the populations, how does that affect them?
Aneil: Well, it affects them in two ways. So, the first way is that because of stochastic processes we know that the frequency of deleterious mutations is not the same in all subpopulations.
Paul: Okay, can I just jump in there, so a stochastic process, what’s that?
Aneil: That’s any kind of random process. In evolutionary biology we refer to genetic drift, which is changes in allele frequency that occur just by chance. So, for example, when an offspring inherits alleles from its parents, it receives one from its mum and one from its dad, but which one it receives from its mum and which one it receives from its dad…
Paul: Okay, this is the sort of classic thing of taking coloured balls out of a bucket or something?
Paul: And you don’t know if you’re going to get red ones or green ones and…?
Aneil: Exactly. And so because of that allele frequencies will differ a little bit among populations, so not all populations will be exactly identical genetically. And one consequence of this means that deleterious alleles will be at a higher frequency in some populations than others, and that means you’ll tend to see them as homozygotes more often than you’d expect.
Paul: So that’s when you’ve got two alleles, two of the same alleles together and not two different ones?
Aneil: Exactly, and when that happens, so that’s the first effect of population structure, it creates an excess of homozygosy, and when you have homozygotes it’s easier for selection to see those alleles and more efficiently remove them.
Paul: Oh this like the thing with very rare human diseases that are recessive; they’re usually not seen but in very small populations you can get two of them together, they become homozygous, and then suddenly you’ve got that disease appearing?
Aneil: Exactly, or it’s really the same process that you see with inbreeding that you expose these deleterious alleles when you have inbred offspring. So that’s the first effect of population structures, it creates homozygosy that allows selection to be more efficient at removing these things and thus reducing the effect of these deleterious mutations on the population in general.
Paul: Oh so once they’ve gone from the population, then that selection’s removed them.
Aneil: That’s right.
Paul: It’s a better way of removing them from the population?
Aneil: Exactly, so that’s one effect of population structure; however, there’s potentially this other effect which is local competition. In many populations, individuals compete for resources, and it might not be so bad to have a deleterious allele if you're competing against other individuals that also carry that deleterious allele.
Paul: Okay. So if you're all bad, if you're a little bit better than all of them then, then you do well?
Aneil: Yes, exactly.
Aneil: But this second effect of the population structure really depends on the ecology of selection, so, and it’s rated to classic ideas of what’s known as hard selection and soft selection. So, hard selection is the type of selection where the fitness of an individual depends only on its genetic quality, and it doesn’t depend on who its neighbours are. With soft selection, the fitness of an individual depends on its genotype relative to its neighbours, and this is what we sort of normally think of as competition; it doesn’t matter if you're a bad competitor if the individuals you're competing against are also bad competitors. So that’s soft selection. So, the difference between hard selection and soft selection really is whether your fitness is independent of your local surroundings or very dependent on your local surroundings.
Paul: Okay, could you maybe give an example of each of those?
Aneil: Sure. So, a good example of hard selection would be a gene that affects whether you will successfully hatch out of your egg. So that kind of a gene, it’s not going to matter what other eggs are around you, whereas a gene that affects how quickly you're able to find food will have a big effect on your fitness if you're in an environment with others who are quickly removing resources from the environment. But, if you're in an environment where other individuals are also bad at finding food then there should still be lots of food to find and it won't affect you so much. So, that type of gene we might expect to experience soft selection.
Paul: So soft selection is the thing which is important in these fragmented…?
Aneil: That’s right, so if ecological circumstances cause this soft selection then you can get these deleterious alleles being sheltered from selection and increasing in frequency as a result of population structure, rather than decreasing as you would expect from that inbreeding or homozygosy effect that I first described.
Paul: Okay right, so putting this back into the broader context, so there is this mechanism then, for allowing soft selection to maintain these higher levels of deleterious alleles in these divided populations. So what’s that going to mean, ultimately?
Aneil: What it can mean is that the average genetic quality of individuals in subdivided populations can be considerably lower than one would get in a well mixed population, in a population that wasn’t so fragmented.
Paul: Okay, and this could have conservation implications or, for extinction or…?
Aneil: It certainly could in, so one case is if a population gets fragmented, and you get this kind of thing where the average genetic quality of an individual declines, and then is exposed to competition from an external, from some second species that, say an invasive species, then it’s in a situation where the native species is starting off in a state of being in reduced condition because of the population structure making it easier for that invasive species to get in and potentially displace it.
Paul: Okay, well thanks very much. It looks pretty clear that we’re going to have to really start thinking about the fine detail of ecology and how populations are put together if we’re fully going to understand evolution. Thanks very much.
Aneil: Thank you.