Carcasses of adult crocodiles do not usually signal the return of winter in South Africa, but mass death seems to be becoming the harbinger of the season. Rangers at the Kruger National Park have found Nile crocodiles floating in the Oli­fants River or bloated and decaying along its banks. Investigators are rushing to figure out the cause and worry that the deaths might be signaling the presence of toxins or pathogens that could threaten not only the croc population but also the livelihoods of the people living near the river.

The Olifants River runs several hundred kilometers through three South African provinces and into Mozambique. It supplies water to industrial agriculture operations that send food to Europe and to the local rural communities, which also depend on those waters for fishing and farming.

The first sign of croc trouble in the river came in the winter of 2008, when rangers collected 170 dead individuals, sometimes at a rate of 20 bodies a week. A survey at the end of this May showed nearly 400 crocs living in the park’s gorge, down from at least 1,000 in 2008. So far, as of Au­­gust 7, rangers and scientists have found 23 carcasses.

After slicing open some of the crocodile corpses last year, researchers determined some kind of pansteatitis—an inflammation of adipose tissue—was killing the animals. Specifically, their tails were swollen with the hardened, enlarged fat deposits, which had stiffened and immobilized the crocodiles and left them unable to hunt. Samples of the fat showed the deposits had oxidized to bright yellow.

[Slide Show: Photos of a croc crisis, including an image of the yellow fat]

The disease may not be limited to crocs. Scientists found the same kinds of fat deposits in fish in the Olifants River. And in the river’s gorge just upstream from Massingir Dam in Mozambique, which also has seen croc declines, birds were absent, raising the possibility that they, too, have succumbed to the same agent.

But the cause behind the strange fattening remains a mystery. In June a team led by Henk Bouwman of North-West University, Potchefstroom Campus, in South Africa reported test results from crocodile tissues at two European chemistry meetings. “Everything is there,” Bouwman says, referring to the detection of DDT, PCBs, dioxins and brominated flame retardants, “but nothing is screaming, ‘it’s me, it’s me, it’s me.’ ”

One possibility could be related to dinoflagellates and cyanobacteria found upstream in the catchment, which might be releasing toxins similar to those that cause red tides in marine environments, says Peter Ashton, a water resources specialist at the Council for Scientific and Industrial Research in South Africa and the University of Pretoria.

“It never is a quick, easy solution” in which it takes one test to find a culprit, explains Danny Govender, a disease ecologist for South African National Parks. She notes that samples taken from live crocs in 2007 showed that the fat of some crocodiles was beginning to harden. Along with Bouwman, she hypothesizes that all these toxins, found below harmful levels individually, could be acting together in a deadly brew.

Govender cites changes to the river’s ecosystem that stem from infrastructure outside the park, including hundreds of coal-mining operations upstream, where crocodiles have disappeared almost completely, and a dam downstream of the gorge. For the first time in the two decades since it was built, the dam’s reservoir was full last year, slowing down the Olifants’s flow through the crocodiles’ gorge. Govender wonders if the slowed water enabled toxins to build up along the crocodiles’ stretch of the river. Indeed, hydrogen sulfide, ammonia and other compounds from river sediments probably caused massive fish deaths in July, scientists have concluded, and crocs eating these contaminated fish could have been affected.

Even if researchers find the culprits, the impacts could reach further than suspected. “We really underestimated [the number of dead] crocodiles from last year’s count,” Govender adds, noting that their bodies could have been eaten by other crocs or sunk to the bottom of the river. “I suspect we’re losing a lot of breeding females,” whose carcasses are smaller and more easily scavenged. If that is the case, she says, the gorge’s crocodile population may not ever recover, even if scientists can pinpoint the cause of the die-off.

As for the people who depend on the Olifants River, “I don’t know what to tell them,” Bouwman says. His colleague Henrik Kylin of Swedish University of Agricultural Sciences in Uppsala wants to go into Mozambique to test fish there, and possibly people, to see if the croc killer kills more than just crocs.

Note: This article was originally printed with the title, "Hunting a Croc Killer."