As herpetologist Steven Platt trudged through the seasonally flooded Nay Ya Inn wetland in Myanmar (formerly Burma) during a 2016 dry-season expedition, something strange caught his eye: Frisbee-sized pools brimming with clusters of frog eggs and wriggling tadpoles.

The watery pockmarks were old elephant tracks. Platt, who works at the Wildlife Conservation Society, realized that in the parched landscape these puddles may be a lifeline for the next generation of frogs. “It made me wonder how important these tracks—really, tiny little ponds—might be for all the smaller things that are out there,” he says.

Elephants are often cited as ecosystem engineers. They knock over trees, trample brush, prune branches and disperse seeds, enhancing biodiversity and helping maintain savannas and forests.

Many researchers focus on these big-picture impacts, but Platt realized other important ones may be right at the elephants' feet. When he returned to the site in 2017, he found tracks in the same spot—and the tadpoles and eggs were back, too. Resembling a series of frog-sized Jacuzzis, the tracks appear to act as small breeding sites linking together larger wetland patches during the dry season, Platt and his co-authors reported in May in Mammalia.

Such microcosms of life are probably commonplace, Platt says, but almost “no one bothered to look before.” A 2017 paper published in the African Journal of Ecology—possibly the only other study that has examined biodiversity in flooded elephant tracks—supports his hunch: its authors found dozens of invertebrate species and tadpoles in elephant footprints and artificially created puddles in Uganda.

According to Chris Thouless, who directs the Elephant Crisis Fund at the Kenya-based nonprofit Save the Elephants and was not involved in the new research, the Myanmar findings are “an amazing demonstration of the interconnectivity in the natural world, between the largest and one of the smallest creatures in the landscape.” But habitat loss and poaching threaten elephants throughout their range, Thouless says, and scientists do not know whether frog populations will crash if elephants disappear from the landscape—or whether “new ecological relationships will develop that re-create at least part of the lost complexity of the system.”

Platt guesses that at least some of that complexity is irreplaceable. “As the elephants go,” he says, “probably a lot of relationships we don't even know anything about at this point go, too.”