Festering into disrepair at the end of an abandoned military airstrip in Panama was a small and lonely toilet. If Hell has toilets, they are probably more inviting than this one: three grime-slicked cubicles that hadn't seen action since the Cold War. I lent my ladder against a crusty urinal, and hundreds of tropical wasps sprang into defence formation. Bristling from the sink, a phalanx of butch workers glowered, and rudely flicked V-signs at me with their wings.

Pieter – my fellow wasp-botherer – was waiting anxiously for news outside the loo. I said what was required of me as a biologist.

“Well, er, it’s a big sample size in here.”

Wasps love apocalyptic ruins. In the quest for the perfect wasp populations, Pieter and I had become connoisseurs of abandoned Panama. We had spent months scouring derelict and overgrown army bases, crawled under bridges in mangrove swamps, gingerly stepped over needles in the graffitied shell of a forgotten hospital, climbed through the ruins of a submarine-watching radar station (now inhabited by large crabs), counted wasp nests on the precipice of a rainforest dam, and set up shop in ominously empty houses being reclaimed by tropical vines. This is par for the course. On the trail of wasps in various parts of the world, members of the lab have discovered three rabid dogs, one corpse, and the disembodied mummy of a two-headed Zambian cow. Being a tropical wasp biologist means venturing into the unknown, where 'the unknown' is the sort of place that was known once but is now widely agreed to be not worth knowing at all.

In the toilet, I got to work. We had a routine. I propped the ladder against a cistern, climbed up so my face was inches from my first clients, and shone my torch into the hexagonal paper cells that made up the nest. The infant wasps glistened innocently like satanic jelly-babies.

“Large larva… pupa, egg, egg, pupa, egg, flesh fly parasite – ooo, big one –, egg, pupa…”

I called out the current occupant of each cell in the nest, and Pieter recorded it in shorthand on a hexagonal map. In this way – week in, week out – we had diligently worked through twenty thousand brood-rearing cells on repeat. Improbably, we had racked up over a quarter of a million records documenting the development (and frequent demise) of thousands of baby wasps in nests at the Caribbean mouth of the Panama Canal, whose short lives would otherwise have passed unnoticed by anyone. Pieter’s writing-hand was beginning to suffer, and we had both acquired an intimate knowledge of wasp stings.

A reasonable question is ‘why?’. Here’s the answer. Despite their bohemian choice of lodgings, wasps are worth getting to know personally. Their social lives are immensely rich. As W. D. Hamilton – a hero of evolutionary biologists – put it, a Polistes wasp nest is...

...a world human in its seeming motivations and activities far beyond all that seems reasonable to expect from an insect: constructive activity, duty, rebellion, mother care, violence, cheating, cowardice, unity in the face of threat – all these are there."

Most importantly, wasp workers toil unto death to raise their siblings, offering a window into the evolution of something magical. Most animals live solitary, mutually hostile, and narrowly self-interested lives. And yet, in the lush tropical world of the Cretaceous, wasps started to cooperate with one another. They evolved to sacrifice their personal survival and reproduction to help other members of their close family. As career changes go, this was a big one - much as if Vlad the Impaler decided to take up child-minding. Around 100 million years later, an otherwise unextraordinary primate wondered why wasps had bothered. The many enigmas of wasp societies have now captivated biologists for decades.

Polistes wasp nests can be observed directly. Workers adopt a budget approach to building, without any covering envelope to hide the larvae. Photo: PK.

Pieter inspecting a small nest in a tree. Photo: P. Botha.

Searching for nests on the coast near Colón. Photo: P. Botha.

We were here to unpick a specific puzzle about wasp altruism. The wasps of Central and South America are misbehaving. The job description for a wasp is supposed to be simple: help your close family in your own nest, and be mean to everyone else. But these wasps had gone rogue. They were graciously, diligently, and apparently stupidly, also helping the neighbours. In 2005, in a leprosy hospital on a sea-cliff near Panama City, one of my PhD supervisors (Professor Seirian Sumner) had attached minute radio-tags to hundreds of wasps, revealing that – far from forming mean and mutually-murderous colonies – wasps were cheerfully moving between nests and helping to raise supposedly-enemy kids. Wherever wasps of this species set up camp, they form inexplicable networks of vagrant babysitters. This poses a striking challenge for inclusive fitness theory (the established theory for the evolution of social behaviour). Why would a wasp scorn its closest family and help the neighbours instead, who are more distantly related and so are (in the calculus of wasp value) less-useful vehicles for shared genes?

An outbuilding covered with wasps in a clearing. Photo: P. Botha.

“I’m going in deeper!” I called to Pieter, edging along a narrow gauntlet of wasps brandishing stings. The final cubicle was the most waspish, but could be reached only by shimmying past a moody nest on a disintegrating hinge.

Why workers ‘drift’ between nests in Neotropical wasps has, in fact, been a paradox since the beginning of the field. In the 1960s, W. D. Hamilton noticed Polistes wasps moving between nests in Brazil. In his explosive 1964 paper introducing inclusive fitness theory to biology, Hamilton included his observations as an apparent ‘anomaly’ to his own predictions. Since then, three adaptive hypotheses have been proposed. Broadly, these are:

1. Wasps might become slightly useless at home when there are too many helpers, so it’s worth switching to help more distant relatives after a point. You’re still helping family, after all – just more distant family. We called this the ‘diminishing returns’  hypothesis.
2. Wasps live in a shockingly risky world. Nests are failing all over the place. If you invest all your effort in one nest, you’ll feel pretty stupid when it is clobbered by a bird. Like an investment banker diversifying investments into different stocks, you should hedge your bets by investing in multiple relatives on different nests – even though that means being less related to the recipients of your help overall. We called this the ‘bet-hedging’ hypothesis.
3. Wasps might work better with distant relatives. Who wants to run a business with their sisters? Wasps might exchange workers with other nests, and then everyone benefits. Send workers to nests who are sending workers to other nests! We called this the ‘indirect reciprocity’ hypothesis.

In a long series of animated meetings in Bristol, Dr Andy Higginson, Professor Andy Radford, Professor Seirian Sumner, and I bounced ideas around and debated the possibilities. Probing the shadowy past of natural selection becomes easier when you employ evolutionary models. Models can’t tell you what has happened, but they can tell you what hasn’t happened. If a good idea no longer works when translated into cold algebra, you can reject it. And in our case, the models issued a verdict: the ‘bet-hedging’ and ‘indirect reciprocity’ hypotheses were unlikely to work in the real world. That left the ‘diminishing returns’ hypothesis. But do these wasps face diminishing returns strong enough to select for drifting?

Empirical estimate of the benefit term in Hamilton's rule. Left: The highest predicted benefits (yellower colours) are found on colonies with substantial brood-rearing challenges - those with many brood and relatively fewer workers. Right: 95% credible intervals.

Our observations suggest they do indeed. Back in England, we spent months transcribing the field notes into a form suitable for R (the current version can’t read hundreds of pages of sweat-smeared hexagonal pencil marks). Analysing the data with techniques often used in medical population studies (under the expert eye of Professor Nicky Welton), we found that as the number of workers rises on a colony, the usefulness of a worker declines. The range of nest sizes in these wasps is large – from below 10 workers to over 200. This creates a sizeable gradient in payoffs. In principle, sharing the workforce with more distant kin makes sound evolutionary sense. In the paper, we show that sufficiently large diminishing returns to cooperation occur in wild populations, and we use evolutionary models to show that diminishing returns create a simple selection pressure favouring the evolution of generosity towards the neighbours.

I retreated from the final cubicle into the warmth of the Panamanian afternoon, gratefully started breathing again, and looked out across the bay past the lines of palms. Container ships were queuing to enter the Canal. The guttural groans of howler monkeys emanated from the nearby canopy, and energetic highways of leafcutter ants bustled past my feet with their bobbing cargo like tiny green yachts. It wouldn’t be long before the fireflies appeared. Hell probably has better toilets, but I doubt they have a better view.

The paper can be found in Nature Ecology and Evolution here.

Kennedy, P., Sumner, S., Botha, P., Welton, N.J., Higginson, A.D. & Radford, A.N. Diminishing returns drive altruists to help extended family. Nature Ecology and Evolution, DOI: 10.1038/s41559-020-01382-z