AI reveals African Great Migration has only 533K wildebeest, not 1.3M

A new AI-powered analysis of satellite imagery has revealed that East Africa’s Great Migration involves far fewer wildebeest than previously estimated, with researchers counting only 324,000-533,000 animals instead of the widely cited 1.3 million. This finding challenges decades-old population estimates and demonstrates how artificial intelligence can provide more accurate wildlife census data crucial for conservation efforts.

What you should know: Princeton University researchers used deep-learning models to analyze thousands of square kilometers of satellite imagery, producing the first comprehensive AI-based count of wildebeest in the Serengeti-Mara ecosystem.

• Two AI models—U-Net and YOLOv8—were trained on a dataset containing 70,417 manually labeled wildebeest to identify the animals in satellite photos.
• The models analyzed over 4,000 square kilometers of high-resolution imagery captured in August 2022 and August 2023.
• Both AI systems produced consistent results, counting between 324,202-337,926 wildebeest in 2022 and 502,917-533,137 in 2023.

Why traditional counts fall short: Conventional aerial surveys have significant limitations that can lead to overestimation of wildlife populations.

• Traditional methods survey only small areas at a time, requiring statistical models to extrapolate densities across unsurveyed regions.
• Herds are unevenly distributed and constantly moving, creating opportunities for double-counting or missing animals entirely.
• The current 1.3 million estimate has remained largely unchanged since the 1970s, despite these methodological challenges.

The big picture: Satellite-based AI counting offers superior accuracy and consistency compared to traditional wildlife census methods.

• “Automation through AI does provide greater consistency and accuracy in counting,” says Isla Duporge at Princeton University.
• One satellite photo can cover hundreds of thousands of square kilometers, eliminating the need for extrapolative projections and reducing double-counting errors.
• The approach represents the first time AI has been used to conduct a census of individual mammals across a large, distributed population.

What they’re saying: Researchers believe the true wildebeest population lies somewhere between their AI count and traditional estimates.

• “Based on our findings, I suspect the true population size is closer to around 800,000 individuals, if we were able to count every individual with zero error,” says Duporge.
• “I think the aerial count is an overestimate and ours is a slight underestimate. We miss those under the trees and there will be some outside of the study area, but it’s very surprising that we can see no more than 533,137.”

Conservation implications: Accurate population data is essential for protecting wildebeest and their ecosystem amid mounting environmental pressures.

• Lower counts don’t necessarily indicate population collapse—animals may have altered migration routes due to habitat changes.
• Wildebeest face significant challenges including habitat loss and fragmentation from agricultural expansion.
• The research team is currently developing similar AI methods to detect and count African rhinos.

Looking ahead: The researchers advocate for broader adoption of satellite-AI methods in wildlife monitoring and have made their technology publicly available.

• “I do believe we should be moving toward satellite and AI-based methods for counting wildlife populations, particularly for species that are widely distributed across large landscapes,” says Duporge.
• The team previously used AI to recognize elephants from satellite data and has made their wildebeest detection code available at https://github.com/sat-wildlife/wildebeest.