Before getting started, one disclaimer: there will be nothing about humans in this post. If that’s what you’re looking for, you have to search elsewhere. Human social evolution is a big messy topic and I’m not qualified to talk about it because humans are boring. If you’re interested in a general look at Hamilton’s seminal work and examples of cooperation in nature and the theory underlying how we explain them, read on.
Cooperation is defined as any behaviour that benefits another individual; to really count, this behaviour must also have evolved, at least partially, due to its cooperative effect. The latter part is a critical distinction: an elephant shitting can be seen as cooperation with dung beetles, but excretory behaviour obviously did not evolve in elephants because of the beetles. The existence of these behaviours is a problem for evolutionary theory. Imagine meerkats (pictured above): they live in groups of up to 30 adults with their offspring. Of the adults, there are the dominant ones who breed and their subordinates who help raise the dominants’ young. If one of the subordinates finds something tasty (a large scorpion), it gives it to one of those young instead of eating it. Why?
In evolutionary theory terms, at face value, the subordinate meerkat (termed the actor, since he intiates the cooperative behaviour) is reducing its relative fitness, therefore it has less chances of surviving.
It’s relatively easy to think of animals cooperating, but it actually is a fundamental interaction, right down to the level of cells. The meerkat is only surviving because its cells, tissues and organs are cooperating; those cells in turn depend on the cooperation of their mitochondria, once free-living organisms. And let’s not forget the unicellular life forms: bacteria need to cooperate with each other in order for the colony to grow.
One area that is genuinely interesting and that, at first sight, seems to have nothing to do with cooperation, is the evolution of multicellularity. This is a topic that is extremely hard to study, since any evolutionary signal of how it might have happened has long been erased – except in one group. The volvocine algae have recently evolved multicellularity and they still show a range of transitional forms. Some species are unicellular, some consist of four cells, whereas other are sizeable 50000-cell colonies, with differentiated cells to boot. We can therefore examine these different algae and, using phylogenetics, reconstruct the intermediate steps from a single cell to a multicellular organism.
It is beyond the scope of this post to go into this process completely; I’ll just emphasise the role of cooperation. Obviously, a prerequisite for multicellular development is that the cells stick together. Some specific cells of the multicellular volvocines produce an extracellular matrix (ECM), containing a cocktail mix of glycoproteins that binds cells together. Although this is a costly thing to do, the cooperative benefit – it allows multicellularity – greatly outweighs the cost, and therefore is kept.
Consider again the meerkat (or any other multicellular organism). If we go back through eons of time, this meerkat is but a descendant of some unicellular eukaryote, competing against other unicellular eukaryotes. At some point though, these cells started cooperating, and thus multicellularity was born.
From molecules joining together to form compartments at the beginning of life, through replicators joining together to form chromosomes, through prokaryotes combining to form the eukaryotes, cells multiplying into multicellular organisms and individuals forming colonies, cooperation is ubiquitous. We cannot chalk up its existence to a few exceptions and it must be incorporated as a building block in evolutionary theory. This was a groundbreaking insight from William D. Hamilton, one of the greats of evolutionary theory of the 1960s. It would not be an overstatement to say that most of the things in this post follow more or less directly from his work.
There are two broad explanations, from theory, about the evolution of cooperation, both stemming from Hamilton (1964)’s formalisation of inclusive fitness (although their names were introduced in a review book, Natural Selection and Social Behaviour: Recent Research and New Theories (1981)). Inclusive fitness basically means that the effect of an action is calculated by its effect on the actor’s and the recipient’s fitness, and weighed with the genetic relatedness between them.
There is the direct fitness category, whereby cooperation directly increases the actor’s reproductive success. The best example of this comes from the long-tailed manakins. The females of that species enjoy highly-coordinated ritual dances between two males. Of the males, one is dominant and the other is subordinate – a younger bird, taken in as a sort of apprentice. The pair perform the dance, but only the dominant male gets to copulate. This continues until the dominant one’s death (10+ years!), at which point the younger bird becomes the dominant one (and takes in another subordinate), and becomes the one who copulates. The subordinate therefore does get direct benefits by cooperating (albeit much delayed).
Direct benefits can be divided into two categories. The first is when the benefit is an automatic by-product of the cooperative behaviour. Zebras are an example of this: the larger the herd, the more confusing it is for predators. Each individual is still at risk, but the group as a whole has a high chance of survival. The same principle applies to the eusocial wasp Polistes dominulus. Here, high-ranked dominant females lay eggs while lower-ranked subordinate females forage for food. It was found that 35% of the subordinates were not related to the rest of the colony – not part of that family (and wasps can recognise themselves very easily). However, these unrelated wasps stay because while the danger in foraging is great, there is always a 10% chance that by the time the new offspring emerge, they will be dominant females (when the old dominant ones die).
The other category of direct benefit is when cooperation is forced. Meerkats offer a nice example of this: if a subordinate female gets pregnant, the dominant one will attack her or even throw her out of the group, leading to abortion, as punishment for not cooperating. The reason this is evolutionarily selected for is, on that single species levels, that subordinate females are pricks and will attack and kill the dominant young; on the theoretical level, punishment and the such allow for only cooperative individuals to reap benefits, while cheaters will just die alone.
The meerkat example is an example of punishment due to illegally gaining direct fitness benefits. Cooperation can also be enforced if the cheater stands to gain indirect fitness. An example of this comes from some eusocial hymenopterans. Only queens are supposed to lay eggs, but sometimes, some unique workers will break the rules and lay their own eggs. Other workers will eat those eggs, to prevent the genetic relatedness of the colony from sinking. Those that cheat once tend not to cheat again, and so cooperation is maintained.
The second category used to explain the evolution of cooperation is indirect benefit. That is when a cooperative behaviour leads not to the reproductive success of the individual actor, but to the reproductive success of its genes. In other words, if it can help another individual with similar genes, it will indirectly achieve reproductive success, since it has still passed its genes on to the next generation. Kin selection (a term coined by Maynard Smith) is the most widespread embodiment of this.
One class of cooperative behaviour that can only be explained using indirect fitness is altruism, i.e. behaviour that is costly to the actor but beneficial to the recipient. Hamilton’s Rule summarises how altruistic behaviour can evolve: “an altruistic trait can increase in frequency if the benefit (b) received by the donor’s relatives, weighted by the relationship (r) to the donor, exceeds the cost (c) of the trait to the donor’s fitness.” In mathematical form, this basically means that altruism will be successful when rb > c. Or, as Haldane prophetically and amusingly put it, he “will jump into the river to save two brothers or eight cousins.” It’s prophetic not because he did that (to my knowledge at least), but because he uttered it years before Hamilton established the rule.
The main implication of that formula is that the only way for altruism to evolve is by having a population of genetically closely related individuals. This brings up another problem: how is this high genetic relatedness achieved? Again, we turn to Hamilton’s work for answers. He proposed two mechanisms.
The first is kin discrimination. As the name suggests, this basically means that an individual can recognise relatives and behave cooperativbely only to them. The best example from the natural world is parental care. Other examples include some bird species where those that fail to reproduce will help in raising only their relatives’ offspring. But the coolest example comes from Dictyostelium purpureum, a slime mould found in forests, where it feeds on bacteria. When it runs out of food, the individual cells come together by the thousands and form a moving ‘superorganism’ that looks like a slug. It migrates to the surface, where a fruiting body on a stalk (akin to a mushroom) forms. The stalk cells seemingly commit evolutionary suicide, since only those cells in the fruiting body get to produce spores. However, indirect fitness and kin selection provide perfectly reasonable explanations for how this happens: experiments show that there is quite strong kin discrimination that happens when these slime mould cells come together and only highly related cells will form a slug. Therefore, even the stalk cells are benefitting by passing on their own related genes, and not helping strangers.
Let’s go back to rb > c. So far we’ve seen only cases where genetic relatedness (r) is high. But there is also the case of the benefits (b) being the defining factor. To see this, we will bring about the notion of limited dispersal. In basic terms, this is when an individual will cooperate indiscriminately with its neighbours, since they are most likely to be relatives. An example of this comes from bacteria. Individuals can spend a lot of energy to release siderophores, iron-binding molecules. It is costly to the individual, but benefits the whole population, since anyone can then take in the iron. Even when unrelated cheating members are added to a colony, the individuals will continue to release the siderophores, and at no cost to the colony: the cooperative effect outweighs the cheating effect of the foreign strains. In other words, (b) is very high and (r) does not matter much.
Now let’s look at a case where the cost of helping (c) is high. To see this, we turn to the eusocial hairy-faced hover wasps. Here, females queue up in order to reproduce. The higher up in the queue the female is, the higher ranked she is. Lower-ranked females must forage for food (a very dangerous, thus costly, job) and take care of the young. As kin selection predicts, higher-ranked females go out on less foraging trips: c becomes too high and cooperative behaviour decreases.
Of course, just because we can abstractly split these interactions into the direct and indirect benefit camps, it’s important to note that it is most often the case that both direct and indirect benefits are involved. Many of the fights for which this field is notorious can be attributed to differences in opinion on just how much of each is involved.
And another thing that’s important to state very plainly is that Hamilton’s Rule may be old (1964), but it is fundamental. When reading the literature on this topic, one comes across hundreds of analyses of ‘games’, which can be seen as mathematical analyses of cooperation (originally derived from economics). These are, to be a bit provocative, near-useless at this point – their underlying assumptions are so unrealistic that they cannot be taken in by biologists. At best, they only apply to humans, where a special form of cooperation, indirect reciprocity, takes place. Note that I say reciprocity; the term ‘reciprocal altruism’ is meaningless, since reciprocal behaviour results in direct fitness benefits for both actor and recipient – altruism needs to be costly to the actor, therefore the term is an oxymoron.
You may notice that I have not mentioned group selection explicitly much during this post and instead have taken it for granted. That’s because it is only a controversial topic in the minds of a few antiquated scientists. Let’s review what group selection is in an unbiased fashion.
The original idea of group selection, as espoused by Wynne-Edwards in 1962 (although it originated in a different form much earlier, with Darwin giving it some thought (see the passage in Appendix 1), and was first truly examined by Sewall Wright in 1945), was that in groups filled with selfish individuals, everyone would be acting selfishly, eventually resulting in starvation and overexploitation of resources. In groups where individuals were cooperative and breeded at reasonable rates to preserve resources, the survival rate would thus end up higher. Therefore, cooperative behaviour developed.
As intuitive as that sounds, it is contradicted by all the theoretical and empirical data, as demonstrated by the works of Williams and Maynard Smith. On the empirical front, species that limit their own reproductive potential are almost unheard of. On the theoretical front, the conditions for this type of group selection to occur are absurd – as an example, if only one individual migrates away from the population, the model is ruined. Good luck finding a population where no emigration takes place.
And so it was that the old, 1960s view of group selection was demolished (see Appendix 2 for a passage from George C. Williams). In retrospect, it became known as naive group selectionism, and its principle is aptly demonstrated in Appendix 3. In the 1970s and 1980s, no doubt spurred by E.O. Wilson’s Sociobiology (1975), a new idea of group selection emerged. In this one, individuals interact with each other during their life cycle. It is not a matter of different groups competing as in the old view; it is competition within the group. It is a much more individualistic approach. Instead of the old view’s implication that group selection is the main way of natural selection to manifest itself, individual selection is at the heart of the new view. This is why it is also called multilevel selection (coined by Price in 1972): all levels, from the gene to the group, can be accomodated.
This is where I may have made a false step though. That last statement is very influenced by George C. Williams’ view of group selection. As you can see in Appendix 2, he did not oppose it by principle. He simply said that it was too weak to be practically detectable in natural populations and thus can safely be ignored in favor of individualistic evolution. David Sloane Wilson thinks (and I may be putting words in his mouth here, but that’s the feeling one gets from reading his work) that this is when group selection got doomed. We don’t need group selection anymore, since everything can be explained by Hamiltonian inclusive fitness, kin selection, etc. The history of group selection is a delightful topic (if one is more than slightly masochistic), but I will leave it for another post. The next paragraph will tell you the jist of why the debate is more or less pointless.
If you use a framework of kin selection to examine a problem, and you use multilevel selection to examine that same problem, you will get the exact same results. Once you simplify both theories to their core, you will see that they are the same; multilevel selection is just a rebranding of kin selection, as Hamilton realised in 1975. Hamilton’s Rule also lies at the heart of both; the terms used are just different. And this is what has also led to much confusion, from within the literature to books to public debates. Again, DS Wilson will probably slap me if he reads this, and I’m aware that group selection does have a lot going for it and actually separating it from kin selection (take animal punishment as an example – the ant that kills the worker eggs. Why is the police ant doing it? What’s its motivation?). But I’ve already deviated from the post topic enough!
This post is meant as an introduction into this unwieldy topic. It is a historically very active field and has since its inception been subject to much confusion, mostly caused by faulty use of terminology. This post is by no means a manifesto; I just want to make sure that if you want to get into this topic, that you can orient yourself a bit. The post is also written at a basic level, mostly because there is no clarity beyond there, since the conversation tends to dissolve into semantic bitch fights (or scientific debates, as they are diplomatically called).
“It must not be forgotten that although a high standard of morality gives but a slight or no advantage to each individual man and his children over the other men in his tribe … an increase in the number of well-endowed men and an advancement in the standard of morality will certainly give an immense advantage to one tribe over another.”
– Charles Darwin, 1871. In The Descent of Man.
“It is universally conceded by those who have seriously concerned themselves with this problem … that such group-related adaptations must be attributed to the natural selection of alternative groups of individuals and that the natural selection of alternative alleles within populations will be opposed to this development. i am in entire agreement with the reasoning behind this conclusion. Only by a theory of between-group selection could we achieve a scientific explanation of group-related adaptations. However, I would question one of the premises on which this reasoning is based. Chapters 5 and 8 will be primarily a defence of the thesis that group-related adaptations do not, in fact, exist. A group in this discussion should be understood to mean something other than a family and to be composed of individuals that need not be closely related.”
– George C. Williams, 1966. In Adaptation and Natural Selection.
“The probability of survival of individual living things, or of populations, increases with the degree with which they harmoniously adjust themselves to each other and to their environment. This principle is basic to the concept of the balance of nature, orders the subject matter of ecology and evolution, underlies organismic and developmental biology, and is the foundation of all sociobiology.”
– Allee et al., 1949. In Principles of Animal Ecology.