It’s August in Australia, and a small, mouse-like creature called an antechinus is busy killing himself through sex. He was a virgin until now, but for two to three weeks, this little lothario goes at it non-stop. He mates with as many females as he can, in violent, frenetic encounters that can each last up to 14 hours. He does little else.

A month ago, he irreversibly stopped making sperm, so he’s got all that he will ever have. This burst of speed-mating is his one chance to pass his genes on to the next generation, and he will die trying. He exhausts himself so thoroughly that his body starts to fall apart. His blood courses with testosterone and stress hormones. His fur falls off. He bleeds internally. His immune system fails to fight off incoming infections, and he becomes riddled with gangrene.

He’s a complete mess, but he’s still after sex. “By the end of the mating season, physically disintegrating males may run around frantically searching for last mating opportunities,” says Diana Fisher from the University of Queensland. “By that time, females are, not surprisingly, avoiding them.”

Soon, it’s all over. A few weeks shy of his first birthday, he is dead, along with every other male antechinus in the area.

The technical term for this is semelparity, from the Latin words for “to beget once”. For semelparous animals, from salmon to mayflies, sex is a once-in-a-lifetime affair, and usually a fatal one. This practice is common among many animal groups, but rare among mammals. You only see it in the 12 species of antechinuses and a few close relatives, all of which are small, insect-eating marsupials. (Although they look like rodents and are colloquially called marsupial mice, antechinuses are more closely related to kangaroos and koalas than to mice or rats.)

Why? Why do these marsupials practice suicidal reproduction, and why are they the only mammals that do so?

The question has vexed biologists for three decades, and many have offered answers. Some say that females don’t survive very well after breeding, so males are forced to hedge their bets by mating with as many as possible. Other suggest that it’s just a feature of the group, which have become locked into a weird breeding system through some unknown quirk of their evolutionary history. Yet others think the males are being altruistic, sacrificing themselves to leave more resources for the next generation.

But Fisher, who has been studying antechinuses for decades, favours a different idea. Her team gathered data on the lives and environments of a wide variety of 52 insect-eating marsupials, from the fully semelparous antechinuses, to relatives where a small number of males survive past their first sexual liaisons, to species that breed repeatedly.

It’s their diet that matters. These animals feed on insects, and some experience a glut of food once a year but very little at other times. This seasonality increases the further you get from the equator. The species with the most seasonal menus also had shorter breeding seasons, and their males were more likely to die after mating.

Fisher thinks that as the ancestors of antechinuses spread south through Australia and New Guinea, they encountered strong yearly fluctuations in their food supply. The females were better at raising their young if they gave birth just before the annual bonanza, and were well-fed enough to wean their joeys. Their mating seasons shortened and synchronised, collapsing into a tight window of time.

That probably wouldn’t have happened if they were placental mammals like shrews or mice, which could have produced several litters during the peak of food. But they were marsupials: their babies are born at an incredible early stage and rely on their mothers’ milk for a long time. A baby shrew suckles for days or weeks; a baby antechinus does so for four months. The females could only fit in one litter during the annual peak.

This had a huge impact on the males, which were forced to compete intensely with each other in a matter of weeks.  They didn’t fight. Rather than using claws or teeth, they competed with sperm. The more they had, the more females they impregnated, and the more likely they were to displace the sperm of earlier suitors. Indeed, Fisher found a clear relationship between suicidal reproduction and testes size. The biggest testes of all, relative to body size, belong to species whose males die en masse, followed by those where a minority survive to mate again, and then by those with several breeding seasons.

The males that put the greatest efforts into sperm competition fathered the most young. It didn’t matter if they burned themselves out in the process, if they metabolised their own muscles to fuel their marathon bouts. These animals are short-lived anyway, so putting all their energy into one frenzied, fatal mating season was the best strategy for them. Living fast and dying young was adaptive.

This idea was first proposed in 1979 but Fisher’s data, although mostly correlative, provides fresh support for it. She certainly finds it more plausible than the idea that the males are selflessly sacrificing themselves for the next generation. After all, the males usually live outside the females’ home ranges, so are unlikely to compete with their own young for resources.

“Antechinus mating habits have appeared in many documentaries, and the explanation of males selflessly sacrifing themselves to increase food supply for young is the one given in all the ones I have seen,” says Fisher. “I hope that documentaries and textbooks now start to give an evidence-based explanation of sexual selection.”