It is another peaceful day on the savannah. Herds of buffalo mill around, chewing the cud and flicking their ears to keep the flies at bay. On their backs, tiny oxpecker birds hop here and there to remove tiny skin parasites, unnoticed by their large hosts.

But there is a dark undertone to this tranquil scene. Closer examination has revealed that oxpeckers do more than simply clean their hosts' skin. Some of them are poking away at open wounds and feeding on buffalo blood.

It seems like even this famous symbiotic arrangement between large herbivores and grooming birds is hampered by organisms' inherent tendency to be selfish.

In fact, this might be a general rule. Mutualistic interactions in nature are often regarded as unstable, precisely because one partner will generally take advantage of the other given half a chance.

This suggests that selfishness is a dominant force in nature; an inevitable symptom of the "survival of the fittest" mantra. So is there any such thing as a truly selfless act? And how should we interpret human acts of charity?

Darwin was a diligent thinker. He agonised over any example he came across of organisms behaving in a way that seemed to run counter to his idea of evolution through natural selection. He soon recognised altruism as "one special difficulty, which at first appeared to me to be […] fatal to the whole theory".

"He who was ready to sacrifice his life," he fretted in The Descent of Man, "would often leave no offspring to inherit his noble nature." How can genes for selflessness propagate if the heroes carrying them spend more time doing good deeds than ensuring their own survival?

Not only are these animals placing the safety of others above their own, they are playing with very high stakes

All sorts of animals seem to be capable of acts of kindness. Primates groom their peers, birds warn each other when they see a predator, and African wild dogs will look after pups belonging to their fellow pack members.

These altruistic behaviours do not necessarily carry a significant cost in terms of an individual's chances of survival, but some other examples do.

Take Belding's ground squirrel. This rodent is found in the mountains of the north-western US, where it lives in sociable groups and provides a delicious snack for various birds of prey and predatory mammals.

When a predator approaches a colony of ground squirrels, the individual who sees it will delay fleeing in order to emit a series of noises, alerting its fellows to the imminent danger.

This is exactly what concerned Darwin. Not only are these animals placing the safety of others above their own, they are playing with very high stakes. Research has demonstrated that making certain calls increases the likelihood of an individual ground squirrel ending up as lunch.

About 100 years after Darwin first published his theory, biologists finally formalised a hypothesis that could explain such baffling behaviour.

Alarm calls are in fact a function of entrenched nepotism in squirrel society

A potentially apocryphal story helps explain the basis of this hypothesis. When J.B.S. Haldane, co-founder of the "modern synthesis" of evolution, was asked if he would throw himself into a river to save his brother, he replied: "No, but I would to save two brothers or eight cousins."

This is the concept of kin selection, which was formalised in 1963 by the highly influential biologist W.D. Hamilton.

According to Hamilton, the altruistic behaviour of an individual ground squirrel is easy to explain when the selfless act is put into its proper context. One animal might be prepared to sacrifice its life – and its ability to reproduce – in order to raise the chances of its relatives surviving and breeding.

Relatives share many of your genes, so they can be considered as proxies for your own evolutionary success and – as per Haldane's statement – the closer the relative, the better.

Kin selection might explain the most extreme form of self-sacrifice in nature: eusociality

Ground squirrels, it turns out, are a textbook example of kin selection.

Paul Sherman, a researcher at Cornell University, spent three years in the 1970s monitoring populations of ground squirrels, paying particular attention to their familial relationships. His conclusion was that alarm calls are in fact a function of entrenched nepotism in squirrel society, with individuals more likely to emit them if there are relatives nearby.

Hamilton also suggested that kin selection might explain the most extreme form of self-sacrifice in nature: eusociality.

The average honeybee, ant or termite has pretty limited options in life. They inhabit highly cooperative colonies where, unless you are queen, you must pay the ultimate evolutionary price: your reproductive ability. The societies these animals form are what biologists call "eusocial" ones.

Since Hamilton, kin selection has been the most popular explanation for this counterintuitive phenomenon, although there have been various modifications to his original proposal. The fact that ants in the same colony are all sisters explains why one member might sacrifice her reproductive capacity, and even her life, for the good of the colony.

A vampire bat cannot go for more than 36 hours without a meal

Arguments have been traded back and forth over whether or not eusociality results from kin selection alone, and indeed over how significant kin selection is in general.

But whatever the conclusion, the idea of genetic ties lying at the heart of selfless behaviour is now an important tenet of evolutionary biology.

However, there is another key theory concerning altruism that has proved rather more contentious.

In 1984 Gerald Wilkinson from the University of Maryland, USA, reported on food sharing in vampire bats. This is a fairly disgusting process that involves bats regurgitating blood into each other's mouths, but it is also a necessary one. A vampire bat cannot go for more than 36 hours without a meal, so regular food sharing can save some animals from starving.

Wilkinson's observations were of note because while the bats do live in groups with other family members, they were preferentially helping those who were the hungriest, not just those with whom they shared genes.

What Wilkinson had uncovered was the best example to date of a concept, previously proposed by eminent evolutionary biologist Robert Trivers, known as reciprocal altruism.

Reciprocal altruism is a model wherein animals help unrelated individuals and accept a short-term loss if it means a long-term gain. So if a starving bat is helped by a well-fed one, the action is done because of some deep, unconscious instinct that in the future the tables might be turned: today's starving bat might be tomorrow's well-fed one.

This suggests that all forms of altruism evolve for selfish means, at least in terms of the survival of genes

"You can think of this as 'social capital' – when vampire bats go hungry they depend on a network of family and friends," says Gerald Carter, a researcher at the Smithsonian Tropical Research Institute and the University of Toronto, who previously completed a PhD under Wilkinson. Carter has since worked on strengthening the case for reciprocal altruism in these bats, responding to claims that even this behaviour can be ascribed to kin selection.

While kin selection has seen wide acceptance both on theoretical and empirical grounds, reciprocal altruism has been harder to pin down, not least because it involves long-term – even lifetime – monitoring of animals to determine costs versus benefits.

But the real issue with both kin selection and reciprocal altruism is that, by our standards, they are not particularly altruistic. Both seem to exist only because they actually benefit the "altruist", albeit in a roundabout sort of way.

This suggests that all forms of altruism evolve for selfish means, at least in terms of the survival of genes. Trivers acknowledge this point when he stated that models of altruistic behaviour in terms of natural selection "take the altruism out of altruism".

So where does that leave humans? Surely we are capable of no-strings-attached charitable acts?

Perhaps we are, says Jonathan Birch, a philosopher of science at the London School of Economics and Political Science.

I have seen nothing special about the human brain

"We should distinguish 'biological altruism' from 'psychological altruism'," he says. "Biological altruism is action that has consequences for reproduction – it causes the altruist to reproduce less and the beneficiary to reproduce more."

"Psychological altruism is action motivated by concern for others – something humans do a lot."

The question is whether our psychological altruism – what we might call "true altruism" – is uniquely human or whether it has deeper evolutionary connections to the altruism seen in other animals.

Michael Platt, a neurobiologist at Duke University, USA, who has spent years investigating the neurological mechanisms that underlie decision making in animals, thinks that animals and humans have more in common than we might imagine.

The factors motivating us to be "psychologically" altruistic are probably shared with at least our primate cousins

"From my own point of view, the notion of 'true altruism' is probably a bit misplaced," says Platt. "I have seen nothing special about the human brain that distinguishes motivated behaviours that benefit another individual compared with monkey brains, or even rat brains."

Platt and his colleagues have undertaken behavioural experiments with rhesus macaques that reveal a degree of charitable behaviour to their fellows. Monkeys, like us, are social animals, and such behaviour might be an adaptive strategy that allowed us all to build and maintain social bonds.

When the researchers conducted brain scans on the monkeys, they identified brain cells that are associated with giving rewards to another individual. Crucially, these cells inhabit a part of the brain that is active in humans when we are empathising with other individuals.

This type of observation suggests that the factors motivating us to be "psychologically" altruistic are probably shared with at least our primate cousins, and maybe other animals as well.

"When people are motivated to help a refugee from Syria, or give to Oxfam, I think that motivation is deriving from this brain circuitry," says Platt.

More than any other animals, we are products of our culture

"Of all the differences between man and the lower animals," wrote Darwin. "The moral sense or conscience is by far the most important." Does Platt's evidence that links our "true" altruism with the apparently selfish instincts of animals devalue this "moral sense", and the charitable actions that result from it?

"Some people have suggested that evolutionary explanations of morality show it to be a kind of illusion. After all, natural selection doesn't care about right and wrong," says Birch. "But I think a lot of this involves overselling the role of natural selection, and underselling the role of cultural evolution."

More than any other animals, we are products of our culture, and the addition of cultural evolution to the mix makes comprehending our own behaviour that extra bit more difficult.

"It's less clear that cultural evolution is a fundamentally amoral process," says Birch. "So a story of human altruism as the product of cultural evolution is potentially more 'morality-friendly' than a story based solely on genetic evolution."

But culture aside, there is a danger in entirely separating us from animals, and in writing off altruism by natural selection as inherently selfish.

The capability for altruism displayed by chimpanzees, for example, has been compared with that observed in young human children. A paper published in January 2016 even seems to show voles consoling each other in times of distress; behaviour that the researchers linked with similar neurochemistry to that of empathetic humans.

If we can understand why people do this, I think we will be aiming towards a greater good

"To suggest that monkeys and other non-human animals are just motivated by instinctive drives is belied by the evidence," says Platt. If we consider ourselves a species capable of selflessness, then we may have to conclude that at least some non-human species share that capability.

"Drawing a hard and fast line between humans and animals would impede progress in understanding how and why we are motivated to give to others," he concludes.

"Investigating this subject will hopefully tell us how to bring out those 'better angels of our nature'. If we can understand why people do this, I think we will be aiming towards a greater good."

Join over five million BBC Earth fans by liking us on Facebook, or follow us on Twitter and Instagram.

If you liked this story, sign up for the weekly features newsletter called "If You Only Read 6 Things This Week". A handpicked selection of stories from BBC Future, Earth, Culture, Capital, Travel and Autos, delivered to your inbox every Friday.