Photo: Large flock of European Starlings (a bird)
Group selection is the hypothesis that organisms have the traits they do (including altruistic traits) because selection has produced traits that assure that species survive. Although this is intuitively an OK idea, it turns out that it doesn't work.
Have you ever noticed large "roosts" of birds in trees around town. Roosting birds gather by the hundreds or thousands in one, or a few, trees (maybe you have mistakenly parked you car underneath a roost and suffered the consequences). Biologists are interested in understanding the causes of roosting behavior. People who support the group selection hypothesis have proposed that the reason that these birds are roosting is that it gives them an opportunity to examine how large their population is. Becasue the birds do not want to overpopulate their environment, because overpopulation could lead to a loss of all of the food so that the entire species dies, birds want to know how many other birds are there so they know how much to reproduce. If birds see that the roosts are large then they know that the population is large so they decide to produce only a few babies. However, if the birds see that the roost is small then they are decide to produce many babies. Thus, the population never gets so large that they eat up all of the food.
Unfortunately, the math required for group selection just doesn't work out. Imagine a species of birds that mated monogamously for life. If the parents wanted to keep population sizes constant than their best strategy would be to produce two offspring during their life so that they make just enough kids to replace themselves. For this to happen all females would have a gene that said "make two babies". Imagine that a mutation occurs that says "make three babies". This mutation would quickly spread througout the population so that eventually all females would produce babies. If mutations that said produce 4 or more babies occurred then these mutations would also spread. It is thus possible to imagine that each female would make so many babies that the population would indeed get large enough to consume all of the food which would cause the population to go extinct. Thus, the math of natural selection does not allow organisms to artificially reduce their fitness for the "good of the species".
The observation that led group selectionist to thinking that roosting and reproduction could be explained by group selection was that females produced fewer eggs when more individuals were at the roost than when fewer individuals were at the roost. Can you think of another hypothesis to explain this observation?
So why do birds form roosts? There are at least two hypotheses. First, some scientists propose that organisms roost because they are safer from predators when living in large groups. Others think that organisms form roosts because they can benefit from information gained by living with lots of other individuals. For example, if you flew to the south to look for food and didn't find much and you noticed that those birds returning to the roost from the north looked well fed, then you might head out to the north the next day.
Although most of this makes sense to me, one aspect doens't. You mentioned that group selection wouldn't work in the rosting example because mathmatically it didnt make sense. Since mutations in genes could cause the birth of 3, 4 ect babies eventually causing the population to be overpopulated overtime. Well in a population where both the wildtype(gene that codes for only 2 offspring) and the mutated gene were present how will the mutated gene spread over the entire population causing overpopulation and ultimately starvation and death? Wouldnt nature favor the gene that only codes for the birth of only 2 offspring since it gives the population the best chance to survive and reproduce effectively? The mutated gene would first target the species at the individual level starting with parents feeding their offspring, if food began to become limited then the parents wouldnt be able to find food and feed their offpring thus causing the offspring to die before reaching a mature age for reproduction. This would consequently cause the "mutated gene" to reach its genetic dead end. Overtime the population would stabalize the way nature has intended.
ReplyDeleteAdam, thanks for your post. Because I was trying to illustrate how group selection doesn't work as simply as possible I have imagined a very simplified situation. In my example, I assunmed that all offspring produced had equal probability of surviving. If this was true, then my explanation was accurate.
ReplyDeleteYou are correct, that in reality this assumption may not always apply. In bird species that produce altricial (helpless) young that they feed it is possible that parents that produce more babies may actually have less surviving offspring because they are not able to give their babies enough food. It is well know that birds will adjust the number of eggs that they lay (clutch size) according to the environmental condiditons. When environmental conditions are poor parents lay fewer eggs because that maximizes ghe individual fitness of the parents (not as sometimes suggested it assures the survival of the species).
You suggested that organisms should have traits that give "the population the best chance to survive and reproduce effectively". It is important to realize that selection doesn't work to give populations the best chance ofd survival. There are lots of examples of local populations going to extinction. We will talk more about population regulation starting next week so hopefully this will make a little more sense then.
Thanks for your comments. Often while trying to make a point I present a very over simplfied example. Often it is possible for students who are thinking about the issue in more detail to get confused. Hopefully, I this has helped clear some of this us. If you have more questions let me know.