What is it about?
Brucellosis is a bacterial infection that moves between wildlife, farm animals, and people. We know quite a bit about how it affects cattle and goats, but we know surprisingly little about how it touches wild members of the dog family, like wolves, foxes, coyotes, and jackals. That is a problem. These animals often roam right up to farm gates and through the edges of towns. They might be spreading the bacteria without us realizing it, or they might be acting like canaries in the coal mine, quietly signaling that the disease is present in the landscape. The trouble is that studies on wild canids use all kinds of different tests, and no test is perfect. Some cry wolf when there is none, giving false positives. Others stay silent when infection is actually there, giving false negatives. With so many studies using different tools, it becomes almost impossible to piece together a clear picture of how common brucellosis really is across the world. So we took a different approach. We gathered every study on brucellosis in wild canids published between 1962 and 2025. Then we built a statistical model that accounts for test errors. We treated antibody tests, which show past exposure, separately from DNA tests and bacterial cultures, which show active infection. This let us estimate true prevalence more honestly. What we found surprised us. Exposure is surprisingly widespread, showing up in about 8 percent of wild canids globally. But confirmed active infections are much rarer, only about 4 percent. Geography matters too. South America had the highest exposure levels, while Europe had the lowest. And here is the twist: much of the variation we saw across studies had less to do with actual disease patterns and more to do with which test researchers happened to use. Older tests tended to inflate the numbers, while newer confirmatory tests told a more measured story.
Featured Image
Photo by Naveen Naidu on Unsplash
Why is it important?
This work matters for three reasons, each building on the last. First, it gives us the most reliable global picture yet of brucellosis in wild canids. By correcting for imperfect tests, we move closer to the truth. For wildlife managers and public health officials, that means decisions can be based on numbers that actually reflect reality, not just the quirks of a particular test. Second, this study shines a bright light on where we are flying blind. Africa and large parts of Asia have almost no data on wild canids. That is troubling because those are often the very places where brucellosis in livestock and people is most difficult to control. Our results map out exactly where future research should go. Third, we show that older, simpler tests consistently overestimate infection compared to modern confirmatory methods. That insight gives practical guidance to researchers designing new surveys. It also helps veterinarians interpret older studies with a more critical eye. At its heart, this work is about connection. Wild canids live at the crossroads of wildlife, livestock, and human communities. Understanding their true exposure to brucellosis is a perfect example of One Health, the idea that the health of people, animals, and the environment is woven together. We cannot understand one without paying attention to the others.
Perspectives
What drew me to this project was the messiness of the data. When we first laid out all the studies, the numbers looked like a wild rollercoaster. Some papers reported 30 percent of animals exposed. Others found zero. It would have been easy to shrug and say brucellosis is just unpredictable, or that some regions are riskier than others. But something about that explanation felt too simple. So we built a model that could look past the surface noise and ask a deeper question: what if the differences were not in the animals at all, but in the tests we used to look at them? That turned out to be exactly what was happening. The rollercoaster smoothed out once we accounted for test performance. It was a reminder that sometimes the most interesting discoveries are not about the disease itself, but about how we have been seeing it all along. Wildlife disease surveillance is chronically underfunded. Researchers often have to make do with whatever tests they can afford, not necessarily the ones they would choose if money were no object. Our framework offers a way to still extract reliable insight from imperfect tools. I hope this work encourages more studies in the places where data is scarcest, especially Africa and Asia, where the boundary between wildlife and people is closest. More than anything, I hope it serves as a reminder that in science, getting better data is important, but understanding the limitations of the data we already have can be just as powerful.
Dr Arman Abdous
Tehran University of Medical Sciences
Read the Original
This page is a summary of: From tests to truth: A misclassification-aware machine learning framework for estimating brucellosis seroprevalence in wild canids, PLoS Neglected Tropical Diseases, March 2026, PLOS,
DOI: 10.1371/journal.pntd.0014029.
You can read the full text:
Contributors
The following have contributed to this page
