FEATURE ARTICLE
Water, Migration and the Serengeti Ecosystem
Understanding the mechanisms that control the timing of wildlife migrations may prove vital to successful management
Eric Wolanski, Emmanuel Gereta, Markus Borner, Simon Mduma
The Hydrologic Setting
Although the Serengeti ecosystem includes areas outside Serengeti National Park, we chose to restrict our study area to within park boundaries, where data are most available and reliable. The park covers approximately 15,000 square kilometers, about the area of Connecticut, and is drained by the Mara, Grumeti and Mbalageti Rivers, all of which flow westward to Lake Victoria (see Figure 2). The Mara passes through the park in the far north, at 1,300 to 1,500 meters of elevation, and drains a total of 10,300 square kilometers, most of which is in Kenya. The Grumeti, with a total catchment area of 11,600 square kilometers, drains much of the wooded savanna of the central and northern hills much inside the park's boundaries. To the south, in an area of treeless grasslands and hills lying between 1,600 and 1,660 meters in elevation, the Mbalageti drains 2,680 square kilometers, nearly all within the park.
In general, river flows, as indicated by data collected daily from 1970 to 1974, are not indicative of drainage area. The Mara, which was gauged at its entrance to the park, had a peak flow of 1,000 cubic meters per second. The Grumeti, with a similar drainage area, saw a maximum of 200 cubic meters per second at its exit from the park. To the south, the Mbalageti peaked at only 40 cubic meters per second where it left the park. These flows vary greatly with the season. After the rains, discharges decrease exponentially to base flows maintained by groundwater seepage. The Mara reaches its base flow a few months after the rains end, whereas the Grumeti and Mbalageti decline to essentially zero flow within weeks and days, respectively. Thus during much of the year, the Grumeti and Mbalageti consist of a series of stagnant pools tens to hundreds of meters long and less than a meter deep, except where hippopotami constantly stir them and prevent siltation. These stagnant pools are the only source of drinking water for wildlife during the dry season.
We analyzed monthly rainfall data from January 1960 to June 1989 from 232 stations, not all of which were operational continuously. Rainfall peaks are reached in April and minimums in August or September. There is also large interannual variability, although this has little effect on the seasonal pattern. The amount of annual rainfall does not correlate significantly with the Southern Oscillation Index (a measure of climatic patterns in the tropical Pacific) in this part of Africa, although major El Ni?o Southern Oscillation (ENSO, a disturbance in climatic patterns in the tropical Pacific) events were clearly reflected in increased rainfall (see Figure 3). Sub-Saharan climate is also known to be influenced by quasi-biennial oscillations of the atmosphere (QBO) and by interactions between the global ENSO signal and tropical Atlantic and Indian ocean processes, such as monsoon convection, with time lags of 12 to 18 months.
In general, rainfall is greatest in the northwestern and extreme southeastern corners of the park, exceeding 110 centimeters per year in the northwest, and least in the east, at about 50 centimeters per year (see Figure 4). A rain shadow downwind of the Ngorongoro Mountains makes the southern grasslands the driest area of the park. When the rainfall data are computer animated, it becomes apparent that rainfall generally intrudes into the park as fronts from the north and the southeast. On an interannual basis, the southeastern rainfall dominates in very wet years (top 10 percent), whereas the northern rains dominate in dry years.
Rainfall was negligible in the southern grassland in 1969 but abundant in 1989. These are extreme years in an overall pattern wherein mean annual rainfall was about 50 percent higher during the mid-1970s and late 1980s than it was during the early 1960s. These patterns are clearly reflected in vegetation, with the boundary between wooded savanna and grassland moving southward during the wetter decades. Nonetheless, other factors may influence this boundary shift, including the discouragement of man-made fires and the poaching of 70 percent of all elephants during the 1980s. Elsewhere, in Congo's Virunga National Park, the killing of elephants led in only 20 years to the transformation of grasslands into densely wooded savanna, despite continuing wildfires.
Rainfall and river flow clearly influence the movement of migrating species of herbivores in the Serengeti ecosystem, but the mere timing or quantity of water arrival cannot precisely explain when the animals will begin to move. They will be in the south during the wet season and back in the northwest by the dry season, but, as noted previously, the move out of the south may vary by as much as three months from year to year. There is clearly more at play in this equation.
» Post Comment