In Sri Lanka, ecological change is shrinking a park’s elephant population.
Elephants live in some of the most complex and resilient societies in the animal kingdom. Individuals die or move away, calves grow up, and newcomers join, but the local communities as a whole usually hold together. Long-term studies of African savanna elephants have shown that, even after poaching or other spikes in deaths, survivors tend to rebuild their lost connections over time. But a new study suggests that resilience may not keep the community intact when steady losses accumulate over time.
Anastasia Madsen, a behavioral ecologist at the University of California, San Diego, and her colleagues analyzed 12 years of observations of wild Asian elephants in Sri Lanka. They found that, once losses began to outpace newcomers, the elephants’ society didn’t just get smaller. It fractured.
Accidental savanna
Udawalawe National Park in southern Sri Lanka was created in the 1970s around a reservoir on one of the country’s four major rivers. “Before that, it was basically a tropical deciduous forest,” said Shermin de Silva, an ecologist at UC San Diego and senior author of the study (de Silva started the Udawalawe Elephant Research Project in 2006).
The reservoir had brought farming and settlers, some of whom lived inside what later became the park. When they were moved out, the land they had cleared for their crops turned into grassland that, according to de Silva, looks very much like an African savanna. With far more food than the original forest offered, the park became famous for its big elephant herds.
But the savanna was only a stage in the forest’s comeback. Without fires and clearing, invasive plants like lantana took over. “In some places, they’re like an impenetrable thicket. Not even elephants really like to go there because there’s no food for them,” de Silva said. This is a problem because elephants tend to move where the grass is. A park with less forage, the team says, stops attracting elephants from outside while, due to limited food, fewer calves reach adulthood.
Her team watched this transformation from a vehicle, driving random routes along the park’s roads about three days a week. The team identified elephants mostly by their floppy ears, which collect holes, tears, and cuts with age. “Their ears are very individually distinct, so it’s like a fingerprint,” de Silva said. Between 2007 and 2018, the team identified 443 adult and subadult females that roamed the park.
Networking elephants
The team focused on females because males lead more solitary lives, ranging widely in search of mates. Females live in a so-called fission-fusion society. “It’s kind of similar to humans,” Madsen said. “You might have your work friends, you might have your family, and you’re not always with those same individuals at the same time, but they’re all part of your social network.” Unlike African savanna elephants, where families follow older matriarchs, Asian elephants move more independently. This may be because Asia’s rainy environment ensures that food and water are rarely far apart and usually easier to find, so following an experienced individual brings less benefit.
Madsen reconstructed the elephants’ social network for each year, measuring the strength of each relationship by how often two animals were seen together. “You can think of it as a probability that you will see those two individuals together,” she explains. An algorithm commonly used for analyzing human social networks then picked out communities—clusters of elephants with many strong ties to one another.
For the first few years of the study, these communities seemed to hold very well. But then cracks started to appear.
The turning point
Around 2012, the elephants that died or left began to outnumber the newcomers. By 2018, about 69 percent of the females seen at the start of the study were gone, even though Asian elephants can live 60 years or more. Soon after the balance tipped, communities grew smaller, more numerous and more isolated from each other. “As individuals are being lost from the network, there’s nobody filling it back in,” de Silva says.
A shrinking population naturally makes for a smaller network, so the team checked whether this was just a numbers game. They looked at a population of 59 elephants seen every year of the study, which the team referred to as residents. “If it was just an artifact of the data, those core individuals should have fairly consistent relationships. But that’s not what we saw,” Madsen said. “They were changing their relationships with each other, even though they were still there.” Among these residents, the fracturing came one to three years after the decline began.
Madsen suspects some of the lost elephants were the glue holding groups together. “If you go to a city hall meeting, it’s maybe the same group of people that constantly comes in. You’re not friends with every single other person in that group, though,” she said. “There could have been some key individuals that were lost that were holding these subgroups together.”
The first ties to go were the weakest, the casual acquaintances. These weak relationships, Madsen says, could represent animals that merely tolerate each other: “Tolerance is really important in a society, especially in large animals. If they get into conflict, it could be serious.” These weak ties, researchers argue, may still be important because they can carry information between communities.
De Silva links this fracturing to the disappearing grass. “As the competition becomes more severe, your social sphere sort of shrinks,” she said. “It’s as if they’re emphasizing the quality over the quantity of the relationships.”
Flickering lights
Before the breakup, some network statistics fluctuated from year to year, which complex systems scientists call flickering. “If you have a light bulb that’s right on the edge of dying, it might start flickering,” Madsen said. “It’s on, it’s off, it’s on, it’s off. And then suddenly it dies.” Spotted early, she argues, such signals could serve as an early warning sign. “Hopefully, park managers and conservationists can detect that and intervene before the population collapses,” Madsen said.
The challenge is that this detection requires long-term data, and we don’t have that for many populations. What worries the researchers most is that, without early detection, we may only recognize problems after it’s too late. “If you are taking the boards out of a ship and you’re not replacing them, the ship starts to take on water,” Madsen said. “There was a point of no return where there was structural damage, but you might not have noticed it until later.”
The study’s data end in 2018, but de Silva says things haven’t improved. Her team predicted a decline in the park’s elephant population in 2019, and it now appears to be happening faster than expected. “They don’t appear to be breeding very well. A lot of the calves are not surviving,” she said. Many adult females die in their 30s and 40s, and they should be living roughly twice as long.
Part of the reason they die so young is conflict with humans. Female Asian elephants have no tusks, so ivory poaching isn’t the main threat. Instead, elephants get shot when they raid crops or startle people. Occasionally, de Silva says, they also eat explosives hidden in pumpkins some locals place to get rid of wild boars. But the bigger problem, she says, is that elephants can’t survive in the protected areas alone.
More than half of Asian elephant populations live in poor-quality habitats, and the species has lost nearly two-thirds of its likely historic range. “Management really has to be about entire landscapes,” de Silva said. “Roads and infrastructure, irrigation canals—those are the things that are hurting the elephants. It’s not really a wildlife management problem.” Her team now works with farmers outside the park, testing crops elephants are less likely to damage. “Their long-term persistence really depends on people and elephants being able to share the space safely,” she said.
Science, 2026. DOI: 10.1126/science.aef6714










English (US) ·