Back

Predators at the Viral Gate: Multi-Species Foraging at a Marburg Virus Reservoir

Atukwatse, B.; Cornille, O.; Muhereza, J.; Nsabimana, W.; Ssemakula, Y.; Braczkowski, A. A.

2025-06-19 ecology
10.1101/2025.06.16.659814 bioRxiv
Show abstract

Understanding how zoonotic viruses spill over from wildlife to humans requires direct ecological observation at reservoir-host interfaces -- yet such events remain rare in the literature, and no such interface has been documented at scale. As part of a broader long-term study on African leopard (Panthera pardus) population ecology in Queen Elizabeth National Park, Uganda, we deployed camera traps on animal trails and at Python Cave, a known roost site of Egyptian fruit bats (Rousettus aegyptiacus) and a confirmed Marburg and virus reservoir. This serendipitous deployment yielded the first visual documentation of a multi-trophic predator and scavenger guild interacting at a filovirus reservoir site. Over a five month period (February 16th - June 5th 2025) across 304 trap nights, we recorded at least 14 different vertebrate species, including leopards, primates, raptors, and small carnivores, engaging in bat predation, scavenging on bat remains, guano foraging, or cave exploration across 261 temporally and spatially independent events (>1 hour apart). These instances were structured, repeated and the species continuously made contact with the bats or entered their roosting space. Camera traps also recorded an estimated 400 human individuals visiting the cave -- including school groups, tourists, and local trainees -- the majority with no personal protective equipment. The shallow, accessible structure of Python Cave appears to collapse the typical spatial buffers between reservoir species and both predators and humans. These observations constitute the first ecological confirmation of a dynamic, multispecies exposure network at a known Marburg virus site, and may represent a Rosetta Stone for interpreting the real-time mechanics of zoonotic spillover.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.