A team led by scientists from the University of Oxford has developed a more accurate method of counting wildebeest herds crossing Tanzania’s Serengeti and Kenya’s Masai Mara each year. For years, population counts relied on aerial photographs of the herds and extrapolation techniques that were prone to error if herds moved between survey areas or clustered unevenly. The new approach trains artificial intelligence (AI) models to count wildebeest and applies this knowledge to satellite images taken above Tanzania and Kenya.

The traditional estimate of the wildebeest population was 1.3 million, but the new AI models estimate the population to be between approximately 325,000 and 525,000. Researchers do not believe the population has collapsed. The difference may stem from altered migration patterns due to human encroachment on migration corridors and changes in grazing areas caused by climate change. Nevertheless, accurate population data is crucial for conservation planning.