Elusive Form of Evolution Seen in Spiders

http://www.simonsfoundation.org/quanta/20141002-in-social-spiders-evidence-that-groups-evolve/

by digital55 • 12 years ago

73 26 12 years ago

gyom• 12 years ago

The spiders present a puzzle to evolutionary biologists. According to ordinary Darwinian natural selection, only the fittest individuals will pass on their genes. But if that’s the case, why do tangle-web spiders act in ways that might conflict with an individual’s drive to outcompete its neighbors? A spider that defends the nest might put itself at personal risk, jeopardizing its chances of producing offspring. And a spider that rears the young might have to wait to eat until the hunters are sated, so it might go hungry. These are not behaviors that would be expected to enhance an individual’s fitness.

Which is the whole point of "The Selfish Gene". It's not about individuals; it's about their genes. Genes win when their bearer helps out other individuals sharing the same genes.

Luc• 12 years ago

Which is the whole point of "The Selfish Gene".

It's one of my favourite books, but I wonder if it had much of a lasting influence academically. It seems this article doesn't even touch upon the selfish gene viewpoint.

azakai2• 12 years ago

It has, or rather, it reflected the strong trend.

Aside from a few prominent figures like E.O. Wilson, I believe that gene-level selection is accepted as the norm. And group-level selection has been almost completely out of favor for generations - gene-level selection killed it. That is why this article is interesting, it might show a new shift.

pesenti• 12 years ago

No, this article does go against the theories in "The Selfish Gene" (which only predict kin selection). See the whole row between Dawkins and Wilson over multi-level selection: http://www.theguardian.com/science/2012/jun/24/battle-of-the...

scott_s• 12 years ago

To me, the key point of "The Selfish Gene" was that the base unit of evolution is the gene. I don't see anything in this article which disagrees with this notion, and they did quote a biologist who said something similar:

Pruitt and Goodnight don’t propose a mechanism by which the colonies boomeranged back to their original state. And without such a mechanism, some researchers argue that the results could be due to ordinary selection acting on individuals. “I think they over-interpret [the results] as evidence of group-level selection,” said Andy Gardner, a biologist at the University of Oxford. “Natural selection may factor in the needs of the group, to some extent, when it hones the adaptations of individuals. But group fitness is not the whole story.”

I understand Gardner's point to be that even if the selection pressure starts at the group level, it can still go down to the individual, and then down to the gene. In that sense, we're still selecting the gene.

Actually, upon reading the actual article, the authors say basically the same thing in the first paragraph:

"In societies in which individual fitness is tightly linked with the performance of the group, the theory of group selection predicts that evolution will favour traits in individuals that aid in maximizing their group’s success—which, in turn, are predicted to increase individuals’ long-term evolutionary interests. Here we define group selection as selection caused by the differential extinction or proliferation of groups1. This represents a broad definition that is not in any way adversarial to the importance of kinship selection for social evolution."

Assuming I understand the authors correctly, they mean that the "fitness" of an individual cannot always be considered to be in isolation. In order to determine the fitness of an individual, we may have to consider the combination of the individual, and the group in which they live. I see that as consistent with the notion of a "selfish gene", as I understand it.

knodi123• 12 years ago

TL;DR: Different colonies of these spiders had different ratios of nanny spiders to warrior spiders, based on the specific pressures of the habitat they grew up in.

When these colonies were transplanted to a new habitat with different pressures, and then their ratio of nannies to warriors was forcibly changed to match the new habitat, the ratio quickly changed back to one that was suited to their old habitat, leading to the death of the colony.

This shows that the colony as a whole evolved to have a certain ratio, and is not merely adapting its ratio on-the-fly (pun intended).

taeric2• 12 years ago

I'm somewhat frightened by what this could imply about our cultures. :(

a_bonobo• 12 years ago

Nothing? Cultures are complicated things that may or may not be influenced by our genetic makeups. The life and death of culture involves thousands if not tens of thousands of factors, you don't directly inherit your parents' culture, you have to learn it first.

This behaviour in spiders, on the other hand, is highly heritable, so it's definitely linked to a handful of genes, or maybe just one.

bayesianhorse• 12 years ago

I don't follow this conclusion. Especially if the ratio changes back "quickly". I have yet to read the original article, but this sounds to me more like a simple dynamical system: Depending on the food "income", threats and other factors, either more nannies or more warriors hatch. The key may be whether or not the colonies revert to an ancestral composition in the wrong place.

SoftwareMaven3• 12 years ago

It's the opposite of that, and that's why it's interesting. You'd expect dynamics to produce matching warrior/nanny to their environment. Instead, even when their ratio matched the environment (ideal for individuals), the group reverted to its genetic roots (to the detriment of the individuals and group). Alternatively, when the group didn't match its environment, but the ancestors of the group did, the group reverted to matching (an important point left out of the GP).

tokenadult• 12 years ago

I'm eager to see biologists who write online regularly about evolutionary theory comment about this. I've asked Jerry Coyne of Why Evolution Is True to do so by sending him the link. Thanks for sharing the link here; I like Simons Foundation stories about mathematics, and I wonder whether or not this finding in biology will hold up.

tinco• 12 years ago

I've never understood how biologists could argue non-selfish evolution couldn't be a thing. If you look at groups of organisms as systems, and you pit the systems against each other, surely eventually the system that evolved to be superior would win in the end, regardless of whether that means any subspecies in that group of organisms would evolve to not reproduce?

Is it not just a question of whether evolution could produce specialized one-offs that don't reproduce but do serve a purpose? Wouldn't the significant population of homosexuals be a sign of this? Perhaps a biologist could give some info on why many biologists think this is unlikely?

lkrubner• 12 years ago

In "The Selfish Gene" Dawkins makes the point that you do not need to have children to pass on your genes. You get 100% of your genes from your mother and father, and you full-brothers and full-sisters also get 100% of their genes from your mother and father. Therefore, from the point of view of your parents, there is absolutely no difference between you having a child yourself, or you bringing in enough extra food that your sister's child lives through the next famine (whereas they would have died without the food you brought in). In both cases, your parents end up with 1 extra grandchild, because of you. And this is a possible explanation for the fact that in humans the more sons a woman has, the more likely it is for the son to be homosexual.

http://en.wikipedia.org/wiki/Fraternal_birth_order_and_male_...

Biologists do not normally regard it is "altruism" if you give resources to someone who shares at least "50%" of your genes. (The 50% is in quotes because any 2 creatures that belong to the same species will have well over 99% of the same genes, and often 99.9% of the same genes. The "50%" actually refers to only to those genes that vary in your species. Dawkin talks about this, too, in "The Selfish Gene".)

azakai2• 12 years ago

I've never understood how biologists could argue non-selfish evolution couldn't be a thing. If you look at groups of organisms as systems, and you pit the systems against each other, surely eventually the system that evolved to be superior would win in the end, regardless of whether that means any subspecies in that group of organisms would evolve to not reproduce?

Of course, all other things equal, when you pit groups against each other then group selection will occur. But the problem is that a huge body of science, both theoretical (game theory) and observational, argues that things are not equal: At the same time as the groups are competing, there is in-group competition. There has to be, because the groups consist of organisms that are being selected for, unless someone can propose a mechanism that would suppress that.

What this means is that group selection is unstable. A group behavior might be selected for, but individuals in the successful group will still be selected for "selfishly" - normally - breaking the pattern. If selection operates on the gene level - and almost all biologists believe it does - then it is inherently going to act on a scale much smaller than that of groups, and in an inevitable manner. And that is going to destabilize any group selection.

For these reasons it was thought that altruism couldn't be easily explained through evolution. That's why kin selection, reciprocal alruism, and a few other theories are so revolutionary - they show how useful group behaviors arise from selection at the gene level.

Is it not just a question of whether evolution could produce specialized one-offs that don't reproduce but do serve a purpose?

That does occur in the "eusocial" animals, such as ants and bees. But it is generally believed that happens due to how genes are passed on in those insects (different than we are used to among mammals), and they are a large exception. In practically every species we are aware of, group-level selection does not occur. Hence any evidence, as in this article, is very interesting.

jostmey2• 12 years ago

Well put! Selfless behavior may sometimes evolve, but it is unstable and will eventually be selected against.

Gravityloss• 12 years ago

Yeah, it's just extremely obvious and was discovered in the nineteenth century already by Kropotkin. I haven't actually read any Kropotkin (I haven't read original Darwin either) but it pops ups if you read for example Frans de Waal.

The whole evolutionary biology field is quite depressing actually. It doesn't seem to make any cumulative progress whatsoever.

http://en.wikipedia.org/wiki/Mutual_Aid:_A_Factor_of_Evoluti...

bronson• 12 years ago

Isn't it well known that natural selection operates on communities? Doesn't seem very controversial to me.

Douglas Adams described it years ago: https://www.youtube.com/watch?v=sCsHuoVABgI

drincognito• 12 years ago

I'm confused by the subhead. Natural selection occurs at the population level.

SoftwareMaven3• 12 years ago

Evolution occurs at the population level, but populations are generally not thought to evolve.

aaron695• 12 years ago

This experiment seems to prove the exact opposite to what they are saying.

Obviously many/most? animals have evolved group behaviours that benefit their species over other species which in turn benefits themselves.

Often these behaviours are not directly related to propagating their own genes ie protecting someone else's kid or having enough poison in you to scare something that kills you into not eating others like you.

How are these spiders different?

The 'group' that was suited to the environment failed.

This to me proves that groups don't evolve. If it was the group that evolved and not the individual then the group suited to the environment would have survived.

Instead it reverted back to combined individual behaviour.

The individuals had evolved to their environment and had genetic behaviour that allowed them in their natural environment to work as a team for the success of all members as per usual. Put them into another environment and they fail because their personal evolution was not perfected for that environment.

jostmey2• 12 years ago

Agreed! Whatever behavior is ultimately observed, it must be acting in a manner that is beneficial to the individual. The only reason such behavior exists is because the gains to the individual outweigh the costs.

SoftwareMaven3• 12 years ago

A group doesn't "evolve" overnight. That would be environmental pressures influencing it, not evolutionary. That the groups reverted showed heritability in the face of individual needs. If it were strictly individual needs, the colonies shouldn't have reverted; the evolutionary forces that directed warrior/nanny ratio should have reflected the environment instead of the ancestry of the group.

m_mueller• 12 years ago

They are discussing the results, but I have trouble understanding the methods and mechanisms.

* How did they change a docile group into an aggressive one? If I understand this correctly, they took a group of, say, 1000 spiders with a ratio of, say, 10-to-1 docile and removed about 890 spiders to create a new 1-to-10 ratio with 110 spiders. But doesn't such a change in group size also change the group itself significantly?

* How was everything tracked after they've been transferred to an environment?

Elusive Form of Evolution Seen in Spiders

Related Stories