Together, the Serengeti and Masai Mara national reserves comprise the best game viewing area on earth. But it is not just their magnificent array of predators that draws visitors from all over the world. It is the chance to see the great migration of plains game; a spectacle involving the mass movement of animals drawn from at least 200,000 zebra, 300,000 Thompson's gazelles and 1.3 million wildebeest. Stories by Jonathan Scott.

It was the most immense visual spectacle one could ever hope to see. No words can adequately convey the impression of so many animals seen together: kilometre upon kilometre of plains blackened by the wildebeests' presence. And competing for space among them, the other migratory species: zebra, gazelles and shy parties of eland.

I waited with the herds throughout April and early May, watching as the wildebeest, zebra and gazelles moved in the wake of the rains. There were times when the Serengeti plains were so wet that it was virtually impossible to avoid grinding to a halt in the mud, but within a few days the sun emerged again and everywhere looked green and inviting.

As May drew to a close the herds seemed restless, as if they could sense that the long march to the woodlands was soon to begin. At first it seemed innocuous enough, nothing more than a pleasing breeze each morning, but the muddy waterholes were already retreating.

The dry north-easterly winds continued unabated, gusting night and day. The wildebeest turned ther backs on the stinging wind and, like a river in flood, began to stream off the plains, drawing to them tributaries of animals from every corner of the land. Dividing then merging again, they passed through the kopjes, damming up in black clusters to slake their thirst wherever they could still find water.

Lions and leopards looked down on the herds from among rock and tree; for as long as the animals continued to pour over the horizon the predators would have no difficulty finding food. Hyenas dogged the rutted pathways cut by the wildebeest. Many of these shambling hunters would soon abandon the barren eastern plains and move north-west to establish dens at the edge of the woodlands, commuting for days at a time between the Western Corridor and their den sites to prey on the migratory herds.

Early one morning in June, I met the great armies of wildebeest marching through chest-high grasslands that had remained virtually untouched for as long as the short grass plains could sustain the wildebeest during the rainy season. I pulled off the road and watched in awe as wave after wave of animals crested the rise, cropping the grass as they surged forward. Within minutes they were lost to the naked eye. The long march to their dry season pastures was in full spate.

Through binoculars I could see tens of thousands of wildebeest passing through the majestic Moru Kopjies at the edge of the woodlands in the west. Beyond them stood the hills of the southern and central ranges, like sentries across the wide mouth of the Western Corridor. Together the hills and water-courses give form to the westward movements of the herds, channelling them through a series of rocky gateways towards Lake Victoria.

Forty years ago a German, Bernhard Grzimek, and his son, Michael, conducted the first aerial census of the park's animals in an effort to determine more accurately the route followed by the migration. The Grzimeks were concerned about proposals to change the park boundaries to accommodate demands by the Maasai for more grazing land.

Tragically, Michael plunged to his death after a mid-air collision with a griffon vulture. But their work was virtually completed and in 1959 the publication of Serengeti Shall Not Die (and the film of the same name) became a landmark in the fight to save Africa's wild animals.

Though boundary changes were implemented, the land area encomassed by the park remained virtually the same (at present it is 14,763 km2). The addition of a corridor of protected habitat between the Serengeti and the Masai Mara, known as the Northern Extension, has proved to be of vital importance, safeguarding the migration route currently used.

Scientists stationed at the Serengeti Wildlife Research Centre have now refined and computerised the painstaking methods of aerial census work. Approximately 1,000 photographs are taken every two years and individual animals counted from the prints. When the first photographic census was completed in 1961, it revealed a figure of 263,362 wildebeest, nearly three times the 99,481 estimated by the Grzimeks in 1958. Subsequent counts have charted a five-fold increase in the population.

The primary reason for this dramatic rise in the migratory population is thought to be the eradication of rinderpest, a highly contagious viral disease of ruminants. The disease first struck the Serengeti towards the end of the last century. It had always been supposed that cattle contracted rinderpest from theirwild cousins, but when an intensive inoculation programme was initiated during the 1950s, rinderpest soon disappeared from the wildebeest and buffalo herds, suggesting that cattle had been the source of the disease.

Once released from the effects of rinderpest and aided by an increase in dry season rainfall during the 1970s, the Serengeti's wildebeest population exploded and now remains at about 1.3 million, held in check by the availability of dry season forage.

The migration we see today is a reflection of journeys travelled in times past; an event that has been happening for as long as wildebeest have roamed the Serengeti-Mara region. It is only the extent and timing of their wanderings that change. Their annual passage from the southern plains to the dry season holding areas is strongly directional and under suitable conditions follows traditional, well-defined routes; it may even be genetically incorporated into their behaviour. But the availability of food within wet and dry season ranges is constantly changing. To survive, wildebeest adapt their behaviour.

A remarkable feature of the wildebeest's nomadic wanderings is their ability to unerringly locate areas of good grazing, despite the fact these may be many kilometres apart. Ultimately, it is rain that stimulates regrowth during the dry season, so it is thought that rain is the guiding light to their movements. Rain can be falling 50 kilometres away, yet the wildebeest somehow manage to be there in time to meet the surge of green shoots the showers quickly unleash.

Do they hurry over the drying plains in response to the sight of lightning flickering across the darkening skies, the sound of thunder or the smell of the rain? It is probably a combination of all three senses that ensures they are in the right place on time.

But a wildebeest must also learn the easiest way to reach such places, a lifelong process that begins on the short grass plains. A calf accompanies its mother throughout its first year, in effect receiving a guided tour of the current migration route. This provides a broad outline of the path to be followed in years to come; the location of preferred feeding sites, watering places and river crossings.

The only stable form of association the wildebeest knows in its life is the one forged at birth with its mother, a relationship that dissolves soon after the arrival of the next generation of calves. It seems unlikely that any animal could easily remain within the same herd from one generation to the next. The fact that wildebeest do not form fixed relationships between individuals no doubt helps spread information more quickly through the population as to new and vacant areas providing good grazing, and the easiest routes leading to them. The herd is a storehouse of acquired knowledge which facilitates the flow ofinformation from one animal to another.

While living in the Masai Mara I had grown accustomed to a landscape dominated by the russet colours of waist-high red oat grass. But even during the height of the wet season, when the Mara lies buried beneath an ocean of long grass, the Serengeti's short grass plains do not sprout more than a few centimetres from the earth.

If you abandon your vehicle for a moment and press your face close to the ground you get more than a wildebeest's perspective of the plains: you find a carpet of little grass cushions, as neatly arranged as artificial turf.

The grasses are shallow-rooted perennials, able to grow unimpeded by the tree-defying hardpan and ideally suited to the alkaline soils. Their roots are covered with minute hairs which absorb every drop of condensation. Even in the driest spells the life of the sward is assured. In this manner the grasses can survive the withering effects of drought, yet erupt within hours of rain to nourish the wet season deluge of more than a million wildebeest.

Physically, the short grasses look quite different from those found in the long grass areas. They are dwarfed by comparison, bearing small fine leaves and short stems. It is grazing that keeps the grass short and the pasture growing. Proof of this can be found in the scientists' enclosures set among the grasslands. These square, fenced-off areas prevent the animals from feeding on the enclosed plants. Left ungrazed during the rains the grasses soon become mature and flower, growing to heights of up to 60 centimetres. Without the grazing herds, the composition of grass species would change and the taller grasses would crowd out the more prostrate forms in the battle for light. Thus the animals actually help create the conditions they prefer: short grass yielding a high leaf-to-stem ratio.

The wildebeest cannot afford to stay for long in any one part of their range. The rains arrive and depart in scattered thunderstorms. At times no more than a few drops wet the dusty soil, while barely more than a kilometre away I could see dark thunderheads stooping to touch the ground, delivering hour after hour of much needed moisture. This pattern ensures that a least some part of the plains is always receiving rain. The wildebeest respond by constantly rotating round these natural paddocks: revisiting areas every few weeks to recrop the grasses when they are at their most nutritious.

Surprisingly, perhaps, the presence of so many wild animals does not damage their environment. As they sweep from one part of the grasslands to another they constantly enrich the thin soils. Everywhere one looks there are dark piles of dung and urine: vital natural fertilisers. Even as the animals clip the grass, growth hormones pass from roots to shoots, promoting regrowth. And their saliva acts as a stimulant to the grasses. Plants and animals survive in a state of dynamic harmony, allowing both to prosper.

Time in the Masai Mara is measured by the arrival and departure of the migratory wildebeest; their appearance in June breathes life into the tall grasslands, their departure in October leaves the earth exhausted. Each year, as the dry season approaches, tour drivers and safari guides can be heard anxiously inquiring about the whereabouts of the migration and the movements of the herds.

Perhaps the greatest drama occurs when the great herds mass in their thousands at the Mara river. Sometimes they try and cross the river at suicidal places, plunging over cliffs and drowning in their hundreds beneath an insurmountable wall of mud. But what appears to be a difficult crossing place this year may have been far easier in the past. The river banks are constantly changing. Older, more experienced animals probably return to places where they have successfully crossed before. Regardless of how many other wildebeest died there in previous years, they survived to cross again.

Whatever the factors may be that prompt the wildebeest to cross - and sometimes it is simply a consequence of a build-up of animals wanting to drink - nothing deters them once the urge is established. If vehicles or predators disrupt a large crossing at a favourite fording place, the wildebeest simply move to the next, or establish a new one, even thundering through thick forest to reach the river if the need takes them.

Calves may become separated from their mothers during the pandemonium of the more difficult crossings. As the bulk of the herd heads off into the distance, scores of cows and calves gallop to and fro, grunting and bleating, mingling with those still trying to cross, sometimes even re-entering the river in their efforts to relocate each other.

Some years are worse than others. The height of the river, the place they choose and the number of animals crossing at any one time will determine how many animals die. In a bad year, the Mara claims thousands of victims and the course of the river is made visible from kilometres away by the constant swirl of vultures overhead.

But the river soon washes itself clean again. Crocodiles and hyenas, monitor lizards and catfish all play their part in completing the cycle. Vultures probe and prod, ripping out the softest tissues and within a few days the bloated carcasses collapse, leaving nothing but mud-spattered skin draped around the bony wrecks.

Visitors are often shocked by the sight of the aftermath of a difficult crossing; horrified at the suffering endured by wildebeest trapped in the mud. Some demand that ropes and chains be mustered to rescue the dying survivors, unwilling to accept the event as an inevitable part of the natural order of life on the plains. But the majority of imprisoned victims of the river crossing are weak with exhaustion or have sustained serious injuries and would be unfit to continue their journey after being pulled from a graveyard of bodies. Some people have proposed that difficult stretches of river bank be hacked open, or that ramps be built to make easier crossings for the wildebeest, thereby compensating in some measure for the tens of thousands of animals killed by meat poachers each year.

But it is not events such as these that ultimately control the size of the population. The migratory wildebeest have been crossing rivers for thousands of years, have died all manner of deaths, then replenished their numbers with the birth of hundreds of thousands of calves. If man tampers with the system, nature will simply redress the balance elsewhere. Far better to leave them alone than blunder around in the wilderness of our own ignorance.

During September and October hundreds of thousands of wildebeest leave the Mara Triangle (that part of the Masai Mara National Reserve situated to the west of the Mara River) and head back towards the Serengeti. The animals follow the rains south; if the rains falter the herds wait at the edge of the woodlands; if they continue, the wildebeest reach the short grass plains by December: so the cycle continues.

When Bernhard and Michael Grzimek first saw the Serengeti, they were bewitched. They perceived it as hallowed ground, a tract of primordial wilderness where the rest of creation was on equal footing with mankind.

And they were right. There is nowhere else on earth where you can see such an incredible array of animal life - more than two million ungulates, together with the predators that feed on them. Here the game still survives, unchanged for thousands of years despite the periodic devastations wrought by disease and drought and the ever present threat of the poachers. Perhaps the Serengeti and Mara will survive man's deprivation. But only if we really want them to.

Survival on the Move

Wildebeest have long been characterised as ugly fools; the clowns of the plains; a random collection of parts left over from the creation of other animals; God's joke. But in reality they are superbly adapted.

For instance, it takes them no more energy to run than to walk, thus they can take advantage of the scattered distribution of green grass, yet remain within range of water.

One of the great mysteries of the wildebeest migration is the way the animals synchronise the peak in calving. This is only possible if females ovulate at the same time each year and males restrict their sexual activity to a distinct rut in May/June. How this is achieved is unclear, although some think it is triggered by the lunar cycle.

Once the rut begins, the wildebeest are already moving from the plains to the woodlands. They are now at their most concentrated and the bulk of sexual activity is restricted to the space of a few weeks. Fuelled by rising levels of testosterone, contenders for territories challenge each other; chasing wildly through the herds, bucking and kicking. Those that are successful stake out a tiny patch of turf and mate with as many females as possible.

By the time the cows are ready to calve eight months later, the herds will have returned to the short grass plains and their high quality forage.

Almost all cows over three years old produce a calf annually but in recent years the birth peak has been less pronounced. Even so, about 80% of pregnant cows drop their calves within a few weeks, between late January and mid-March.

With the herds seething with 400,00 buff-coloured, black faced-calves, predators are soon glutted by the volume of easy prey.

A calf is also aided by its ability to gain co-ordination faster than any other ungulate (most calves are on their feet within five minutes), in being constantly on the move and in seeking the anonymity of the herd. Within a week calves can gallop at speeds of 50 kilometres an hour.

A more subtle form of protection is the time of day when most calves are born. Many are born at night, others around mid-to-late morning, when most hyenas and lion are seeking shade, giving the calves the rest of the day to gain their strength. However, many still die within their first year, victims of malnutition, disease - and predation.

The first minutes of life are vital in the mother/calf relationship. As the mother cleans and nurses her calf, she imprints its sight, smell and taste on her senses. It is possible that a cow even learns to recognise her calf's bleat of distress.

Born among such enormous numbers, it is all too easy for a calf to become separated from its mother. Predators, cars and low-flying aeroplanes sometimes cause the herds to scatter in panic. There are times when a mother and calf lose contact when they ford a river.

No cow will accept the calf of another. Orphans pass from cow to cow, each of whom smells them before nudging them on their way. Calves blunder towards hyenas already fat from feasting or trot blithely into the fatal embrace of lions. Others followed my car, galloping alongside for a while before turning away to stand alone in the vastness of the Serengeti.

The Migration in Motion

The migration may have been repeated annually for thousands of years, but have the migratory paths remained constant?

The history of the wildebeest migration is as ancient as man. Fossil evidence unearthed at Olduvai Gorge indicates that the modern wildebeest seasonally grazed the Serengeti plains more than a million years ago; its ancestors may have done so for millions of years earlier. Over countless aeons, the wildebeestpopulation has ebbed and flowed, their movements dictated by the ever changing mosaic of grass and woodland.

At present, the wildebeest move through a region that extends from Ngorongoro Crater Highlands in the east almost to the shores of Lake Victoria in the west, and from the Eyasi Escarpment in the south all the way north to the Mara country in Kenya - a total area of 30,000 km2 known as the Serengeti-Mara ecosytem.

Their annual journey through this vast region - which may involve a circuitous trek of up to 3,000 kilometres - has long been thought to retrace age old migration routes, implying a degree of predictability about their behaviour.

Yet I soon discovered you could spend a lifetime in the Serengeti-Mara waiting for a typical migration. The finer details of the herd's movements are always different. It is a dynamic process which defies prediction. Wildebeest are nomadic, responding in a highly opportunistic manner to the vagaries of rainfall and the subsequent availability of food.

The general direction taken by the wildebeest when they first begin to leave the Serengeti plains towards the end of May mirrors the rainfall gradient, which increases from south-east to north-west and is strongly influenced by the towering presence of the Crater Highlands. The dryer areas in the south-east lie directly in the rain shadow of the highlands and receive 500 millimetres of rain annually, almost all of which falls between December and May. Rainfall in areas to the north and west is more evenly distributed and may be as much as 1,200 millimetres annually.

Some years the transition between the wet and dry seasons is particularly marked. In these circumstances the stage is set for a truly dramatic exodus of animals, with hundreds of thousands of wildebeest departing for the woodlands at the same time, accompanied by zebra and Thomson's gazelles.

The trek from plains to woodlands is the most spectacular part of the journey only because this is the time when the herd must travel furthest and most rapidly to find water and grazing.

During the rinderpest years, when the wildebeest population was thought to have been no larger than 250,000, the migration had no need to travel as far north as the Mara Reserve. But as the population swelled, the herds began to leave the plains earlier and were forced to venture further north in their quest for food and water.

Nowadays, many of the Serengeti's wildebeest do not enter the Western Corridor at all after leaving the plains. Instead they travel directly north as the dry season sets in, tarrying a while in the Loliondo Controlled area to the north-east of the park after leaving the Serengeti plains. Meanwhile those animals that departed for the Western Corridor at the end of May eventually exhaust the best of the grazing in the valley pastures and open plains. Then they must move further north.

In some years more than half a million wildebeest advance across the border into the Masai Mara. The Mara and an area in the north-west of the Serengeti known as the Lamai Wedge now provide vital dry season pastures in the northern part of their range. Without these areas of higher rainfall, the migratory population would collapse and the effect would reverberate throughout the whole plant and animal community comprising the Serengeti-Mara ecosystem.

The Making of the Short-grass Plains

The origins of the Serengeti's plains - sometimes referred to as the ancestral home of the wildebeest - lie within the history of the Ngorongoro Crater Highlands, a group of extinct and dormant volcanoes that stand watch beyond the park's eastern gateway.

During periodic eruptions, ash from these volcanoes was swept to the north and west on the prevailing winds. For millions of years, layer upon layer of volcanic ash settled over the undulating plains to the west of the highlands. Today, only the granite kopjes stand as a reminder of the ancient land beneath.

In time, rainfall leached the salts from the porous upper levels to form a concrete-hard layer beneath the surface. This hardpan is impenetrable to all but the shallowest roots, inhibiting the growth of trees and allowing for the colonisation of the area by an uninterrupted sward of shallow-rooted grasses that reaches out to every horizon for 10,000 km2.

The volcanic origin of the Serengeti's powdery soils means that they are rich in potassium, sodium and calcium. A pregnant wildebeest could not wish for more nutritious pastures on which to graze and give birth. Perhaps this is why the migratory wildebeest return to these southern plains in the early part of each year.

Writer and film-maker Jonathan Scott has studied and photographed the migration for the last 20 years. Much of this time has been spent living in the Masai Mara. His books include The Great Migration, The Marsh Lions, The Leopard's Tale, Dawn to Dusk, and he is a past winner of the prestigious BBC Wildlife Photographer of the Year award.

Published in Travel Africa Edition Two: Winter 1997/8. Text is subject to Worldwide Copyright (c)