Myths of Evolution (5): Kin Selection

As with “the selfish gene”, kin selection is a valid perspective which manages to become misleading by missing the point of what it is trying to explain. The idea behind it, developed by J.B.S. Haldane and W.D. Hamilton in the late 1950's and early 1960's, is simply that the behaviour of a great many organisms include the risk or sacrifice of the individual for the benefit of relatives – for instance, alarm calling in squirrels (which puts the caller at greater risk from predators), or sterile worker bees who will never breed. According to conventional Darwinian thinking, natural selection should (supposedly) eliminate such behaviours (at least when considered from the gene-centric view), so kin selection was introduced to explain the apparent anomaly.

But there is no anomaly to explain here: the assumption that natural selection cannot favour the risk or sacrifice of individuals rests on a chain of inferences linking behaviour to strict genetics, which as already discussed is a flimsy argument. When we think of “advantages persisting” instead of “selfish genes”, it becomes clearer that co-operation will not be eliminated by natural selection – it will in fact be selected for whenever it is advantageous. As Peter Kropotkin observed in the 1890s, co-operation and mutual aid contribute significantly to the survival of species, since a group of animals can achieve more than the individual members can on their own. The rise of humanity as a species appears to have been significantly facilitated by its capacity to act as a group.

The power of co-operation even extends into the world of plants: Allen Herre and Elizabeth Arnold discovered that plants defend themselves from fungal infection by hosting a population of symbiotic fungal lifeforms (known as endophytes). Throughout the natural world, examples of beneficial co-operation – symbiosis – can be found. In almost every environment, one can find examples of genetically distinct organisms which co-operate, and thus gain advantages in survival and reproduction.

Furthermore, as Lynn Margulis has explicated, all multi-cellular life operates as a consequence of symbiotic co-operation between micro-organisms which have become so reliant upon one another that we mistake their colonies for individual animals. This viewpoint, which is referred to as endosymbiotic theory, originates in the work of the Russian botanist Konstantin Mereschkowski, who in 1905 observed that cell division in chloroplasts mirrored that in free-living cyanobacteria.

You and I are the product of co-operation between billions of cells (closely related to bacteria) which came together to act as one – the mitochondria (cellular powerhouse) and other organelles in each of your cells descend from bacteria which began to work closely together between 1.6 and 2 billion years ago. This astonishing evolutionary development, along with other symbiotic adaptations, have had as great a role in the development of life on Earth as competition between species.

On the macroscopic scale in which we are more familiar, the existence of social animals who co-operate for mutual benefit is another overt sign of the benefits of co-operation. Although it is misleading to conflate biological mechanisms with behaviour, the hormone and neurotransmitter oxytocin appears to be the constitutive mechanism by which trust operates. Almost all vertebrate species possess a chemically similar protein, which appears to have originally been involved in water regulation. From amphibians onwards, however, there has been a trend of increasing capacity for trust between animals – something we are most familiar among mammals, who form packs and other groups for mutual benefit, but which is also found in the flocking of birds, and even in reptiles and fish. Oxytocin and other related proteins also seem to play a critical role in the formation of families (a large quantity of oxytocin is released during childbirth), one of the most important survival benefits found in the natural world.

The extent to which co-operation affords advantage is apparent in the prevalence of mutual aid in the ocean. Coral reefs, which are nature's cosmopolitan cities, contain species such as cleaner wrasse (pictured above) and cleaner shrimp which remove parasites and dead tissue, even from within the mouths of other fish who do not harm the cleaning animal. Recent observations have shown dozens of sea snakes co-operating not only with each other but also with goatfish and trevally in an astonishing display of multi-species collaboration, dramatically depicted in the recent Planet Earth nature documentary. This is not to deny that competition between and within species does not also occur, nor to claim that savage acts do not happen in the natural world, but there is no reason (other than a metaphysical commitment) to deny that symbiosis and other forms of co-operation are also a key part of the story of life.

All of which goes to demonstrate why co-operation is so well represented throughout biology: it is an advantage so powerful, whenever species stumble upon it, it tends to persist. Rather than trying to shoehorn this observation into older models (as with kin selection), perhaps we might consider a new story which captures the incredible role of co-operation in evolution. Simple lifeforms co-operate because it's in their mutual benefit, and advantages tend to persist, but more complex life (such as humans) co-operate not just because it is immediately advantageous but also because we have inherited a biological mechanism which facilitates the formation of trusting relationships – primarily with our relatives, but potentially with anyone. The story that trust is an advantage runs contrary to the usual mythologies of evolution, but it is just as valid, and far more optimistic.

Alternative myth: Trust is an Advantage

Next Week, the Final Part: Myth #5: Survival of the Fittest

Comments

[Isegoria]

According to conventional Darwinian thinking, natural selection should (supposedly) eliminate such behaviours (at least when considered from the gene-centric view), so kin selection was introduced to explain the apparent anomaly.

I believe you've missed the key insight of "the selfish gene" or the gene-centric point of view.

A naive "suvival of the fittest" point of view would be individual-centric and thus unable to explain (unreciprocated) altruism toward siblings, cousins, etc. If it costs the individual anything, it makes that individual less fit.

A gene-centric point of view recognizes that a gene that promotes altruism towards kin can increase the prevalence of that gene, even if it hurts the individual making the effort to be helpful, because siblings share half their genes, cousins a quarter, and so on.

Reciprocal altruism, on the other hand, does not require any sophisticated explanation, because helping others in order to receive help is a win-win situation; it improves the fitness of both parties. The trick there is knowing whom to trust — not a trivial problem.

[Peter Crowther]

I believe you've missed the key insight of "the selfish gene" or the gene-centric point of view.

Seconded.

[Keith]

Isnt that what he said?

[Chris]

Isegoria: "A gene-centric point of view recognizes that a gene that promotes altruism towards kin can increase the prevalence of that gene, even if it hurts the individual making the effort to be helpful, because siblings share half their genes, cousins a quarter, and so on."

Sorry, but hypothesising a gene that "promotes altruism towards kin" is all well and good, but eventually you have to cough it up. How do genetic relatives know whom they are related to, exactly? Can they smell their kin and thus know to behave altruistically towards them? :) There's no known mechanism to bridge this gap that I've seen - if you have something to offer in this regard, please share!

I'm saying: if there is a gene which accounts for this allegedly anomalous behaviour, it is likely to be the gene that codes for oxytocin, and this gene doesn't produce "kin selection", it facilitates trust between organisms *irrespective* of their genetics.

"Kin selection" is only required as an explanation if you are locked into the gene-centric view, and believe that everything is explained by genes - the teleological account I argued against last week. (Although I recognise that you don't acknowledge it as teleological).

Do I need to dig into the cases of individuals sacrificing themselves for non-genetic relatives? I have a box full of cases of dogs risking or losing their lives for their not-genetic "family", and even some cats too. This isn't best explained by kin selection, which presupposes a close genetic connection not present in these cases - I believe it is sufficient to call upon the trust mechanism as an explanatory device in this regard.

As I say in this piece, kin selection isn't wrong, it's just unnecessary. You only need it if you have predicated a teleological account of evolution in which everything originates in the genes. But sorry, that isn't enough - there are many other factors that must be taken into account which are not reducible to the genes, and without a complete account for behaviour originating in the genes the whole house of cards falls flat. (And frankly, even this wouldn't be enough, because ecology has a central role in the conditions for life and the confluence of species in any given area is not in any way defined by genetics).

You are free to believe that genes provide a full account, and if you do you will share beliefs with Dawkins, Dennett et al. But this is not the only interpretation, nor am I convinced this is the most common explanation among the scientists in the various fields which intersect with evolutionary theories - it might simply be the most well known explanation.

It's ironic, really, that having the Young Earth Creationists to shout at has kept the scientific community seeming to be in agreement about all manner of issues, when in fact there are serious disagreements about fundamental issues under the hood.

I am far from a lone voice arguing that kin selection is an unnecessary construct. Most of the debate about this right now is on the question of kin selection versus group selection - I am swaying heavily towards the group selection camp in this regard. Their arguments strike me as more compelling. However, as long as you hold the belief that genes can be used to explain everything, you will presumably remain wed to the kin selection camp.

If you want to look into this further, I recommend investigating the work of David Sloan Wilson and Elliot Sober (whom I linked to earlier). I believe the multilevel selection theory they advocate has to be taken quite seriously at this point.

As ever, happy to discuss further if there are any relevant points of discussion!

[Chris]

I had some further thoughts about this while I was failing to get to sleep last night (damnable insomnia!)

Firstly, where kin selection *does* work rather well as an explanatory device is in insect colonies. Explaining why a (sterile) termite drone will sacrifice itself for its colony can appeal to the idea of kin selection - since these insects *do* have a signalling mechanism with a genetic origin, namely the chemical messages some (but not all) eusocial insects use as a primary form of communication. It's only as we "scale up" that kin selection suddenly finds itself without any foundation (or rather, with a foundation which contradicts its claims).

But then, there is more than one way to look at this. Hofstadter and others suggest we should think of insect colonies as self-organising systems similar to the way you or I are self-organising systems - thus in this view the termite colony is a superorganism that can be seen as one "individual", comprised of other individuals.

And remember: you and I are more alike here than it at first seems. We are not made of ants, but of the descendants of bacteria who co-operate to form a "superorganism" of another kind. There are more scales at which discrete components can be identified in the eusocial insects (cell, ant, colony vs cell, animal) but either way we are talking about *systems* at different scales as well as genes and individuals. This, in part, is why Leroi argues for "evolution of systems" as a perspective, and I tend to agree with him in this respect.

Again, I'm not saying kin selection is wrong - it's probably a valid explanatory principle in the context of eusocial insects, and presumably other colonial animals such as the portuguese man-of-war, a colonial siphonophore consisting of four distinct and highly specialised polyps/medusoids. (They are clones, so kin selection works especially well for this case!)

But when you seek to apply it at all scales, I must call shenanigans. Alarm calling in squirrels, for instance, which is an oft-cited case of the individual risking/sacrificing itself for "relatives" (even though the squirrels form semi-social groups irrespective of genetic closeness), falls apart on a number of fronts, but none so crucial as the fact that even squirrels of the same species have different alarm calling behaviours from one population to the next (calls, chittering, teeth-gnashing, tail-shaking etc.) and in situations where there simply isn't the opportunity for genetic variance to account for this. It's hardly surprising, since behaviour isn't perfectly described by genetics, and the biological systems from which alarm calling occurs (including the fight-or-flight response, the trust mechanism etc.) are all vastly older than squirrels as a species. From my rather considerable observations of squirrels around the world I must conclude that some crucial aspect of alarm calling is effectively learned behaviour, and hence cultural (not genetic), although I acknowledge considerable similarities therein.

I suspect the value of the kin selection explanation ends at the colonial superorganisms.

Too harsh to call it a myth, then? Not if you want to use kin selection as an explanation for human behaviour. That, I contend, is a thoroughly metaphysical spin on the scientific ideas involved.

Finally, a brief clarification on my use of the term teleology, a term which means "the study of design and purpose". A concept can be stated as teleological whenever a final cause or purpose is identified.

Once upon a time, the preferred teleology among intellectuals was (in effect) God - and this comes more from Plato than from Christianity, which was in fact *hugely* influenced by Plato's ideas since Jesus principally provided an *ethical* message, not a complete metaphysical model to attach it too, and early Christians in Rome were working on ideas within a principally neo-Platonic context.

These days, intellectuals seem to prefer a teleology in which the final cause is natural selection (and *just* natural selection - c.f. Dennett's "universal acid") and the related purpose is the propagation of genes. It is not a mistake to call such a viewpoint teleological - it is a wholly accurate use of the term.

This doesn't mean that you can't wield the idea of natural selection without falling into a teleological trap - it merely notes that there is a trap to fall into here (as Gould observed), and it happens when you believe that genes and natural selection are the *whole* story rather than a key aspect of that story.

Or, for that matter, when you take it further and imagine (for it is pure metaphysics) universes developing under the principle of natural selection. You are free to believe this, but as with all metaphysics, there is never an appeal to proof - it is simply a matter of personal beliefs.

And when we get to different metaphysical camps arguing amongst themselves, well, that's where I turn to my trusty freedom of belief to settle the matter. :)

Well, hopefully that exhausts my ramblings in this regard until you players poke me some more. :)

I look forward to reading your thoughts!

[Isegoria]

Sorry, but hypothesising a gene that "promotes altruism towards kin" is all well and good, but eventually you have to cough it up.

Why? I can't point to a gene for height, but we've known for a long, long time that height is strongly inherited.

In fact, we've known for a long, long time that height is strongly inherited and strongly dependent on nutrition.

You seem caught up on mechanisms that are peripheral to evolution and on a teleological view that a gene must be for something.

All that matters is that a gene has some effect — physical or behavioral — that can be passed along.

I'm saying: if there is a gene which accounts for this allegedly anomalous behaviour, it is likely to be the gene that codes for oxytocin, and this gene doesn't produce "kin selection", it facilitates trust between organisms irrespective of their genetics.

You've taken a very narrow view of a particular mechanism. If oxytocin produces feelings of warmth and trust irrespective of the target's relationship, that means very little, if, say, oxytocin is produced in large amounts by passing a child through the birth canal.

Oxytocin's potential to work its magic with respect to non-relatives is much less important than the fact that it tends to work significantly more often in situations that involve close kin.

[Chris]

Hello again!

I don't believe a gene must be *for* something at all - in fact, that isn't how biology/genetics works at all in practice. Individual proteins turn out not to be anywhere near as clear cut, most having many different functions, and almost always in many different contexts (didn't I put something about this up in the context of the human genome project a short while ago?).

But I *do* believe that a scientific proposition that postulates behaviour emerging from a genetic basis eventually has a burden of proof therein.

"Oxytocin's potential to work its magic with respect to non-relatives is much less important than the fact that it tends to work significantly more often in situations that involve close kin."

So let me get this straight: you agree this is a likely biological explanation for what kin selection was posited to explain?

Then I don't see your point of objection.

I am not saying kin selection is false - clearly, the effect on allele frequency it posits aren't problematic to determine. But kin selection theory goes farther than this (usually): it says non-reciprocal altruism can be *explained* by the allele frequency bias that occurs with relatives.

So what I do is look at the natural history to assess this claim, and what I find is that in more complex creatures, the trust mechanism is the best explanation for non-reciprocal altruism. The key gene here (not the only one, mind you) - the gene for oxytocin - evolved a long time before any of the non-reciprocal altruism behaviours that kin selection is posited to explain. The individual behaviours are highly varied, but the genetic element remains essentially unchanged in all cases. (Oxytocin appears to have originally been a water-regulation hormone which only later acquired its additional new role).

At that point, kin selection *as an explanation for these behaviours* (but not as a mathematical phenomena) is bust wide open. Non-reciprocal altruistic behaviours do not appear to have resulted as a consequence of the allele-frequency effects inherent in genetic relatives.

If the evolution of non-reciprocal altruistic behaviours *doesn't* relate to the allele frequency skew of relatives, then kin selection's claim to explain these behaviours is bogus. Do you see the thrust of my argument here?

Remember that I concede kin selection as a more plausible explanation in the case of eusocial insects. But as an explanation for non-reciprocal altruism in mammals? No, I'm sorry, this doesn't appear to have anything to do with kin selection as far as I can ascertain.

If all the animals involved in the "original trust mechanism" already had all the genes involved because they constituted mandatory elements of the genetic programme and were not situations where allele variation can occur (c.f. Kimura's neutral theory) - which is what I'm claiming on the basis of the available evidence - then allele frequency skews had nothing to do with the arrival of the trust mechanism.

There may still be some plausible objection, but I don't see it myself.

I strongly recommend you look into the kin selection versus group selection debate, if you aren't familiar. It's one of the more fascinating battlegrounds in evolutionary science right now, and watching group selection gain ground after having been previously "counted out" is highly entertaining for me!

Hope this comment has clarified my position somewhat! Thanks for continuing to engage me on this subject.