Python Cave tours? The ways disease jumps from animals to humans are evolving

In July 2008, a 41-year-old Dutch woman developed a rapidly progressing illness after visiting Python Cave, a shallow groove cut into the Maramagambo Forest in Uganda that is home to thousands of Egyptian fruit bats. She’d come with a guide and three others to the cave, which because of its bats had become something of a tourist attraction. Visitors during that period reported seeing bats flying all around, darting overhead and even bumping into people. Bat guano smeared the cave walls. The Dutch woman would later die of Marburg disease, a hemorrhagic fever illness similar to Ebola. A US woman who had recovered from a fever after an earlier visit to Python Cave was also retrospectively diagnosed with Marburg.

Given its attraction of tourists, the Ugandan government and the US Centers for Disease Control and Prevention (CDC) set up an observatory a few hundred meters from Python Cave—so named because of the bat-eating pythons that live there. It was an attempt to keep people safe from the viral threats that lurked within. In early 2025, a team of biologists studying African leopards placed cameras at the entrance of the cave and recorded humans—hundreds of them—approaching the bat habitat. The cameras picked up tourists, school groups, and wildlife research trainees nearing the cave entrance and venturing outside the safe observatory without using any personal protective equipment.

Deadly as Marburg is—the fatality rate ranges from 24-88 percent, according to the World Health Organization—the disease is unlikely to cause a pandemic. The tourism at Python Cave, however, points to yet another way interactions at the animal-human interface—where outbreaks, epidemics, and pandemics start—are evolving.

Many pathogens affecting humans are zoonotic in nature, meaning that humans contract them from animal hosts. Even pathogens that we do not consider zoonotic nowadays—including HIV, most likely SARS-CoV-2, and most influenza strains that cause pandemics—started as such. When we come into contact with wilderness environments, we are exposed to their virome, microbiome, and fungome, risking contact with pathogens that possibly have pandemic potential. Given the varied and increasing interventions of humans in non-urban environments, tourism being just one form, the risk of pathogen crossover from other animals to humans is increasing. The methods by which this risk can be minimized are diverse, but they may carry their own secondary risks, require significant financial support, or even contradict one another.

A shifting interface. Marburg virus is less common than Ebola, the other notorious hemorrhagic fever found in Central African countries, but it appears to be more lethal. Egyptian fruit bats are the natural reservoir, primarily shedding the virus through oral secretions but also in blood and other bodily fluids. Humans can get infected directly, as was the case with the Python Cave visitors, or indirectly through an intermediate host that, once infected, can transmit diseases to humans. There were many such potential intermediaries recorded preying on bats inside Python Cave. Many—including L’Hoest’s monkeys, olive baboons, monitor lizards, and African civets—are hunted by humans as bushmeat.

Human-to-human transmission chains that begin with contact with intermediate hosts or via direct infection from a bat have caused outbreaks of tens or hundreds of Marburg cases in Democratic Republic of Congo, Uganda, and Rwanda, among other countries. Most of these stemmed from a single spillover event, often localized in a cave or mine.

Exposure may be common. A recent surveillance study in Western Uganda, demonstrated that 8 percent of villagers reported an illness after consuming bushmeat, including meat from primates and rats.

Marburg virus outbreaks are increasingly being reported in African countries that haven’t previously reported cases. One outbreak in Ethiopia ended earlier this year after causing nine deaths from 14 confirmed cases. One study estimated that 19 to 27 African countries with a total of 75-105 million people, are at risk of Marburg infection. These areas where Marburg may soon strike include large areas of Ethiopia, Cameroon, and Zambia. A Marburg virus pandemic spread from human to human is not likely: The virus is not considered airborne, and moreover, people infected from Marburg virus are typically infectious when they are severely and obviously ill. The most worrisome pandemic-capable pathogens spread through the air before an infected person shows symptoms. Still, Marburg can cause deadly epidemics that can spread around the world through infected travelers. And other zoonotic potential pandemic agents, known and unknown, are hiding at the human-animal interface.

Beyond Marburg, bats in Southeast Asia have been shown to harbor coronaviruses similar to SARS-CoV-2 that can likely infect humans. Bats are also animal reservoirs of henipaviruses such as Nipah, which causes fever, respiratory distress, encephalitis, and potentially death. Although such viruses aren’t transmissible until symptoms are advanced, one cannot assume that a henipavirus with different characteristics won’t emerge in the future and initiate a pandemic. This is, after all, what happened with SARS-CoV-2, which behaved very differently from its coronavirus predecessors.

An avian influenza pandemic is also possible, should a strain with mutations allowing for adequate human respiratory tract infection and human-to-human airborne transmission infect someone. The animal-human interface is not located only in jungles or forests or what we think of as the wild; it  also exists in human settlements and work sites. Mink farms are a typical example. Denmark ordered the culling of mink after novel strains of SARS-CoV-2 emerged in the animals during the COVID pandemic. Avian influenza strains with signs of mammal adaptation have also been found in fur animal farms in Europe.

Spillover. A spillover event is largely a product of chance; there are certain bottlenecks to be surpassed. A primate infected from Marburg virus, for example, might succumb to the disease and villagers searching for food might sample its carcass, generating human infections. The level of the subsequent outbreak is also dependent on chance factors such as the size of households, the existence of local medical facilities, or the awareness of Marburg and its transmission potential. Each can affect the eventual trajectory of a spillover.

Yet, these factors, largely social and thus theoretically modifiable, may be evolving.

While hunting animals for bushmeat has long been common in many sub-Saharan African countries, in recent years, hunters may have more reason to hunt. The illegal wildlife trade is the fourth most profitable criminal activity worldwide, with products worth more than $20 billion traded annually. In the Democratic Republic of Congo, for instance, some have started hunting for trade beyond personal needs.

Studies from Equatorial Guinea, Kenya, Zambia, Gabon, and Ghana have all demonstrated that hunting and selling bushmeat is a lucrative practice, often generating at least as much income as a government employee might make. When road infrastructure reaches isolated rural areas, so too does the opportunity to make more money in this trade.

As people enter the wild to degrees and in ways in which they hadn’t before, the wild is also coming to town, so to speak. Hendra virus in Australia is transmitted from bats to horses and then to humans. It has a mortality rate exceeding 50 percent. The virus was discovered in 1994 when an outbreak in the Brisbane suburb of Hendra sickened horses and workers at a horse stable. Habitat loss, such as from wildfires, can destroy forest habitat, forcing bats to seek food in trees in semi-urban areas. It’s there that free grazing horses, eating from fruit trees, can become infected through contact with bat saliva or excrement, subsequently infecting their owners or those who work with them.

Similarly, deforestation practices in Malaysia in 1997 to open land for pig breeding resulted in bats harboring Nipah virus no longer living in the middle of a forest, but on the edge of it, near pigs, and humans. This deforestation led to the first Nipah outbreak in humans in 1998-1999.

Guarding the interface. Given the increasing risks of zoonotic disease spillover, what should be done to prevent or at least mitigate it? Should we seek unknown next pandemic candidates in the wild, as some well-funded scientific efforts have? The wilderness can provide a buffer against spillover. Entering it in search of hidden threats in natural habitats, away from human activities, carries its own risk. Once a pathogen is found, there are risks to studying it. None of these risks may be necessary. The speculation that the COVID pandemic had a lab-based origin—research in Wuhan, China ahead of the pandemic involved finding and studying new bat coronaviruses—has cast these research efforts in a particularly critical light.

A better approach may be to study the coronaviruses or other pathogens that have already emerged, like SARS-CoV and MERS. Research on the two viruses wasn’t as robust as it should have been ahead of COVID. A suitable antiviral for MERS would likely have been effective on SARS-CoV-2, as well.

Along these lines, the ReVAMPP approach, for example, involves a collaborating network focusing on research and development on a representative pathogen in each of nine virus families, including, for example, viruses in the Zika and measles families. ReVAMPP is one of the initiatives of the “prototype pathogen approach” initiated by the US National Institute of Allergy and Infectious Diseases (NIAID). Understanding a specific pathogen from certain viral families, developing rapid diagnostics, vaccines, novel antivirals, and monoclonal antibody archetypes for these may allow us to be prepared when something new from one of these virus families emerges.

Aside from better understanding worrisome viral families, surveillance for early recognition is the most convenient approach to guarding against emerging pathogens. But monitoring health care facilities or sampling wastewater are not easy in disease hotspots in low- and middle-income countries that may lack the infrastructure or funding to deploy surveillance. Given the increasing (if so far theoretical) risk of synthetic biology-derived novel pathogens only surveillance (and particularly novel and flexible diagnostic approaches) can guarantee early recognition and response.

Another practical approach would be to focus not on the pathogen but on the host, the natural reservoir, e.g., study bats instead of pathogens. Understanding how bats interact in novel environments after major habitat alterations may allow us to preemptively recognize increasing risks. Such an approach, taking as granted that bats may be the reservoirs of unknown pathogens, might minimize spillover—it could have helped avert, for example, most cases of Hendra virus, by keeping horses (the intermediate host) away from a novel bat habitat. Focusing on the behavior of reservoir species like bats might not tell us about all the viruses that might be out there, but we would be better prepared for the ones that matter.

The risk of spillover continues to increase as the animal-human interface expands and even takes perhaps unexpected forms like tourism. Economic factors like poverty and limited access to public health facilities and other diagnostic and therapeutic infrastructure allow for outbreaks to expand. The absence of health literacy and indeed overall literacy also increases risky behaviors. Disease spillover happens by chance, but from research and surveillance to economic, urban, and tourism policies, there are intervention points and ways for governments and public health authorities to minimize the chance of a pandemic.