seven months, the Serengeti wildebeests abandon 

 them soon after the long rains end in May. The herds 

 then spend the drier months mowing the tall grass- 

 es of the so-called woodland zone — which actually 

 encompasses typical savannas (tree-studded grass- 

 lands) as well as open plains interspersed with wood- 

 lands of varying density. Here, too, armies number- 

 ing many thousands of wildebeests keep on the move. 

 In contrast, hyenas and lions, the main predators, are 

 tied to specific territories. They cannot simply follow 

 the wildebeests wherever they go without trespass- 

 ing on property defended by rival clans and prides. 



How, I've long wondered, do the wildebeests of 

 the Serengeti synchronize their reproduction? 

 After all, the dramatic changes in the length of the 

 day and night, which can 

 serve as clues to the sea- 

 sons in the temperate lat- 

 itudes of southern Africa, 

 are hardly noticeable in 

 the tropics. More than 

 forty years ago I hypoth- 

 esized that the calling of the bulls triggered and syn- 

 chronized estrus. The noise and confusion of the 

 Serengeti rut, which coincides with the migration 

 away from the short-grass plains and into the wood- 

 land zone, is a spectacle unequalled among land mam- 

 mals. The sound and fury come from bulls staking 

 claims to territories and competing to round up and 

 mate with cows. Indeed, the bulls inseminate some 

 600,000 females in less than a month. How they man- 

 age to do it, though, is mystifying. All the action an 

 observer is likely to catch goes into getting and hold- 

 ing females coveted by every neighboring bull. 



My hypothesis about the source of synchronized 

 estrus is being tested as part of a study of the repro- 

 ductive physiology of Serengeti wildebeests. In 2003, 

 a month before the rut, two groups of captured cows 

 (one group penned with a bull) were continuously 

 exposed to the recorded calls of rutting bulls, while 

 a third, control group was kept isolated. The analysis 

 of the thousands of samples of the captive cows' dung, 

 collected throughout the study, is still underway. If 

 my hypothesis is correct, the levels of reproductive 

 hormones in the dung will show that estrus was 

 more closely synchronized in the cows exposed to 

 rutting calls than it was in the control group. 



But assuming I was right, what sets off the bulls? 

 Maybe the fecal-hormone assays will show some 

 changes in the females' hormones that could arouse 

 them. Or perhaps the males are simply reaching top 

 condition at the end of the rains, attaining maximum 

 testes size and testosterone production. Several 

 months of increasing territorial activity precede the 



The wildebeest's short calving 

 season gluts the predators, 

 enabling many calves to survive. 



rut, and mating continues for as long as three months 

 after the rut peaks. Indeed, some territorial activity 

 is seen year round, which is good evidence that ac- 

 tive bulls are ready and willing to breed. But one way 

 or another, male aggression and sexuality reach fever 

 pitch during the few weeks of the rutting peak. 



Whatever stimulates the bulls, the dynamics of 

 how they manage females and establish ter- 

 ritories, both on the move and in more permanent 

 locations, are complex. The prevailing system, in 

 which large aggregations of wildebeests migrate 

 over long distances, has presumably grown out of a 

 nonmigratory system, in which a population's food 

 and water requirements are satisfied more locally. 

 I observed both patterns during my 1960s doctor- 

 al research in Tanzania's 

 Ngorongoro Crater. Out 

 of some 14,000 wilde- 

 beests, as many as 3,000 

 behaved as permanent 

 residents. They could be 

 found in specific areas 

 that were not regularly invaded by the main popu- 

 lation. There, discrete small herds of females, calves, 

 and yearlings, numbering, on average, about ten an- 

 imals each, circulated within a network of fixed ter- 

 ritories defended by breeding bulls. Some of the res- 

 ident bulls I kept track of maintained the same ter- 

 ritory for more than two years. 



The rest of the Ngorongoro wildebeest popula- 

 tion migrated between pastures affording the best 

 grazing on the hundred-square-mile crater floor, in 

 aggregations of several thousand animals. The ag- 

 gregations included competing, mature bulls, as well 

 as noncompeting bachelor males (ranging from 

 yearlings to oldsters), females, and young. When such 

 an aggregation settled on a pasture, the accompany- 

 ing mature bulls began staking claims to territories, 

 pushing noncompeting males out to peripheral — 

 and often substandard — grazing grounds and round- 

 ing up groups of females and young. Temporarily, 

 then, the aggregation would be fragmented much 

 like a resident herd, until the time came to migrate 

 in search of new pasture. 



In a resident population, one bull controls, on av- 

 erage, a territory of perhaps one hectare (a bit big- 

 ger than two football fields). That is actually a re- 

 markably small area for such a big animal (a harte- 

 beest bull, in contrast, may control a territory a 

 hundred times as large). When an aggregation of 



Crocodile exacts a toll from a herd of wildebeests at a river 

 crossing. Spotted hyenas and lions are among the other 

 predator gauntlets that wildebeests run. 



32 



NATURAL HISTORY September 2006 



