Fake Gnus: Five Myths About The Serengeti Wildebeest Migration

At times in Serengeti National Park, one can peer out and see only vast herds of wildebeest in all directions, a dark and humming sea that floods every grass-covered hillside and valley bottom. This spectacle – the Serengeti migration – is arguably amongst the most awe-inspiring sights in the natural world. It involves roughly 1.5 million wildebeest, not to mention 250,000 zebra, 400,000 Thomson gazelle and tens of thousands of eland. Unfortunately, the migration also abounds in misunderstanding and fake gnus (pronounced ‘news’). Superficially, the Serengeti migration may seem like a coordinated movement, involving an innumerable army of clones travelling in synchrony towards some distant destination with greener pastures and fewer lions. But closer inspection suggests a more complicated story: one that emphasizes complexity and variability rather than uniformity.

Our research group at the University of Glasgow in Scotland examines the individual stories of wildebeest and zebra in great detail, scrutinizing where they go, whom they associate with, what drives their movements and how they die. Our approach uses a suite of methodologies, including GPS tracking collars, stable isotopes, camera traps and a variety of other techniques. Because myths are bad business for scientists, in this piece we set out to debunk the five most common falsehoods about the Serengeti migration.

As far as the eye can seen. Wildebeest in the Serengeti. Author provided.


First, a point of clarification: people often use “the migration” to refer to both the behaviour exhibited when animals travel over long distances and the particular collection of herbivores on the move (i.e. the wildebeest, zebra, gazelle and eland). This causes countless problems when people use the meanings interchangeably. For instance, a tourist in Serengeti who comes expecting to witness the behavioural migration may leave disappointed, because, despite seeing tens of thousands of animals, they may not have seen animals sprinting en masse across the savanna. In fact, wildebeest spend most of their waking hours either eating grass or chewing cud, so an expectation of seeing the migration is both unlikely and grossly impertinent.

I understand that such misconceptions have been fuelled by footage of large herds of animals galloping across the plains in wildlife films, so I cannot blame anyone who believes them… but it is important to note that the animals you see in such scenes are almost always running because a helicopter or drone is scaring them senseless, so don’t count on witnessing that behaviour in real life. For the purposes of clarity and expectation management, our preference is to use the term “the migration” to refer to the behaviour, and to use other terms to refer to the animals.

Cool? Now onto our myths…


1. There are 3 million wildebeest in Serengeti

False. Our best recent estimates put the number at around 1.5 million wildebeest, plus or minus about 200,000 animals. Over the last 30 years, the number has hovered closer to 1 million animals. Some websites perpetuate the idea that Serengeti has 2.5 or 3.0 million wildebeest, which likely confuses the total number of migratory animals (wildebeest, zebra, Thomson gazelle and eland) in Serengeti (about 2.5 million) with the total number of wildebeest. Why does it matter? Well, why does it matter if you had three hands instead of two? Because functionally, numbers matter.

How do we know there are ~1.5 million wildebeest? A Tanzanian governmental agency called TAWIRI leads periodic aerial censuses during the wet season when wildebeest are grazing on the short grass plains in southeastern Serengeti. Wildebeest tend to be relatively evenly distributed during this time, making them easier to count. Researchers fly in small planes across the entire section of the park where wildebeest are grazing and take regularly spaced photos from a standard altitude. These photos serve as a sample of the population from which the total number (and measure of error!) can be estimated.


2. The migration has a beginning and an end.

Unlike some organisms that migrate (i.e. passerine birds), herbivores in the Serengeti continue to move throughout the year, never staying in a given location for more than a week or two. The high densities of animals in Serengeti compel them to move constantly in search of food (and sometimes water). Over the course of an entire year, these small daily movements generate the larger movement that makes Serengeti famous. This differs from the classical view of migration in which individuals make rapid, long-distance movements between two seasonal ranges but short movements within seasonal ranges. In fact, on most days (57% of days), wildebeest move less than 5 km (see Figure).

However, this isn’t to say wildebeest don’t occasionally make long distant movements. Each year, wildebeest spend about 13 days in which they travel farther than 25 km, sometimes traveling up to 100 km in a single day. These longer distance movements often occur at the very beginning of the rainy season, when new patches of grass have emerged far away. How wildebeest (and other species) perceive and find these fresh grasses is still a question of great interest.

Animal frequency versus daily distance travelled. Author provided.

Other migrations need greater preparation before undergoing long-distance travel, and thus a start and stop is more obvious. Migratory birds, for instance, have to plan for their migration by completely rewiring their physiology before they go. Individual birds store extra fat and strengthen their flight muscles prior to departing their seasonal ranges. In an extreme example, migratory godwits (an intertidal bird) fly non-stop for ~10,000 kilometers between Alaska and New Zealand. No wonder they need big muscles.


3. All wildebeest are the same.

False. In fact, every animal has a different story. It becomes interesting to think that each individual moves to different places at different times, which means they each eat slightly different food and that they each have unique experiences encountering predators. For instance, our work is trying to understand the physiological consequences of moving to areas outside of the national park, in areas shared with humans.

Beyond experiential differences, these animals also have discernible physical characteristics that allow us to tell them apart. For example, every wildebeest has unique stripe patterns on their shoulders, as evidenced by these photos of the same adult female, photographed one-year apart.

Photographs of the same adult female, taken one year apart, to demonstrate the unique stripe patterns on a wildebeests shoulders. Author provided.


These patterns can be used to identify individuals, similar to fingerprints in humans. In past research in the Tarangire Ecosystem (just East of the Serengeti), we used photographic identification to help track the movements of individual wildebeest across their migration cycle.

Because of the large numbers of animals in this population (about 6000-10000 animals), we developed an open-source software package in JAVA (called ‘Wild-ID’) to assist with the identification of photos.

Photographic identification used to help track wildebeest movements.


While we would love to apply this technique to identify wildebeest in the Serengeti, the sheer number of animals involved makes it simply infeasible.

Back of the notebook calculations suggest that the effort would require dozens of vehicles to take the photos, several hundred people to search through images, and large amounts of computing power to run the identification algorithms. Alas.

Zebra and wildebeest line up side by side to rehydrate. Author provided.


4. (a) All individuals move together in a single migration; (b) zebra and wildebeest migrate together.

Both false, but let’s talk about (a) first. All wildebeest do not necessarily move together. In fact, our research shows that individual animals tend to return to places they have previously visited with high probability (in other words, they have high spatial fidelity). This is surprising given the size of herds (often tens of thousands of individuals make up an aggregation).

As for (b), zebra and wildebeest do co-occur in many areas, often leading people to assume that the two species migrate together (that they are besties!). Only recently have we found that, at the individual level, wildebeest and zebra do different things. First, many zebra in Serengeti have shorter migrations than wildebeest (see below), though some zebra do migrate the full length of the ecosystem (from the short grass plains in the south to the Masai Mara in the north). Second, it seems that wildebeest and zebra do not have a strong association at the individual level. We observed this firsthand in a small experiment (n=3) where we placed GPS collars on pairs of wildebeest and zebra found in the same herds. After a few weeks, the individuals moved in different directions, suggesting that individual wildebeest and zebra may not share the “bestie status” after all. Nonetheless, living in herds is still beneficial for avoiding predators and there are known advantages to associating with other animals. Perhaps, they just don’t care with whom.

5. Death by predator is the most common cause of mortality for wildebeest.

The Discovery Channel would lead you to believe that all wildebeest enter life as calves on the Serengeti plains (mostly true), and die in the awesome jaws of hungry crocodiles or lions (false). In fact, most wildebeest die from starvation, or complications surrounding starvation. We know this for several reasons. First, we can roughly estimate the number of predators in the ecosystem and the number of herbivores they eat each year. Secondly, during periods of drought, wildebeest die in such large numbers that you can easily drive around and find carcasses. In fact, Dr. Simon Mduma, director of the Tanzania Wildlife Research Institute (TAWIRI) did just this about 20 years ago and found that about 75% of all wildebeest deaths were related to lack of food. One can look closely at carcasses (see below) and find that animals often wear down their teeth to the point they can no longer masticate food properly, so they slowly die from starvation. The condition of their bone marrow also shows whether or not they died in a state of starvation.

However, being hungry also makes individuals more likely to take risks to get food, and so there is an important (and difficult to separate) interaction between food and predators that shouldn’t be ignored.

Out in the field. Author provide.

6. Wildebeests are flesh-eating robot clones from the future.

That’s a ridiculous idea. Also, weren’t there only supposed to be five myths?