The Lowveld has experienced average to above-average rainfall over the past six years. These ‘years of plenty’ often give us a false sense of security. As game numbers increase, we try to create a sense of ‘anticipatory awareness’: the dry times will return, but we cannot predict when, how long they will last, or how severe they will be. In fact, it appears that the increased variability in climatic conditions will make prediction increasingly difficult.
The Rangeland Ecology Group of the Agricultural Research Council has over many years presented potential trend scenarios to land users based on current veld condition and animal numbers under varying rainfall conditions (including up to 25 years of historical data) and the predicted response of the grass sward to these variables. We don’t want any unpleasant surprises and we must be proactive rather than reactive in management decisions related to animal numbers.
Losses during times of drought
As land fragmentation has barred natural wild mammal movement to the higher rainfall areas and forage resources in
the west near the Drakensberg range, we inevitably see animal losses in drought years. A decline in numbers of especially larger grazing species such as buffalo, zebra and wildebeest would vary from minimal to steep, as in the 1982 to 1983 droughts where the populations of species were reduced to between 10% and 20% of their pre-drought numbers following extensive mortality in perennial grasses.
The mortality of some of these grazing herbivores is part of a longer-term cycle; droughts are also times in which predators, in particular lion, feast on weakened prey animals. The question is whether we are prepared to allow drought-related mortality to occur and whether the deterioration of the veld would be acceptable if numbers were allowed to increase unchecked. Management decisions are also linked to whether or not a protected area is fenced to bar animal movement to more favourable grazing areas.
High-density grazing herds
The relationship between grass production and standing crop is highlighted by recent favourable rainfall seasons in the eastern Lowveld (mean or above-average rainfall since 2008/2009, as in the example given), resulting in an increase
in grass standing crop – the portion of production remaining after utilisation – as depicted in Figure 1. The latter is due to a favourable perennial grass sward and cover as well as improved soil moisture conditions promoting grass growth (Figure 1).
This has steadily increased herbivore numbers in the Lowveld Protected Areas (Figure 2), largely reflecting three favourable grazing conditions, as will be discussed below. Over the past few years, the grass layer has in general not been limiting for grazers (Figure 1). As grazers such as buffalo move in large herds over extensive areas and are localised around a single water point, they generally have a beneficial effect on the vegetation for the following reasons, among others. At a high density, large-hooved animals:
- break up the soil crust by their hoof action; promoting a good soil surface to seed contact;
- reduce the height of moribund grass, allowing sunlight to penetrate to the shorter vigorous grass tufts and reducing the temperature of the soil to make it more suitable for rainfall infiltration;
- depositing concentrated amounts of dung and urine.
The above promotes seedling establishment, particularly in bare areas, resulting in a healthy and productive perennial grass sward. In turn, increased plant density, a better layer of organic litter and stable soils reduce evaporation and improve rainfall infiltration with a lower soil temperature, less rainfall runoff and reduced silting up of streams. The presence of predators, in particular lion, causes buffalo herds to bunch up when chased, thus intensifying these positive impacts.
As these large herds are mobile, they seldom ‘camp’ on a patch for long but continually move through different landscapes. Unlike selective water-dependent grazers, buffalo utilise an area and then move on to reduce overgrazing, a function of time and not necessarily of number.
Veld needs rest. For example, artificially supplied surface water results in a high density of sedentary water dependent on species such as impala. So where and when do we control animal numbers? Even in unfenced areas, this option may be necessary where localised water provision has increased animal numbers and resulted in insufficient forage during dry periods. This is obviously more critical in fenced situations.
The alternative is to allow the population to fluctuate according to prevailing resource conditions, including a die-off of weaker animals in drought. This may be acceptable in unfenced, ‘open’ situations, but is it appropriate in fenced areas where animals cannot migrate? If the laissez faire option is adopted, the tricky issue of the long-term effect of overgrazing on the resource base arises.
A hypothetical example from a fenced area
We examine the effect of resource use by grazers by inserting the resource requirements for grazing species and determine whether the grazing population can maintain itself under varying environmental and resource conditions. The model is based on a fenced Protected Area (PA) using real data (the major grazers rounded off: buffalo 1 000, blue wildebeest 550, zebra 250, impala 3 100). The grass standing crop (about 1 700kg/ ha) in Year 1 provides some residual for the Year 2 season’s standing crop, as a worst-case scenario projected to yield only 600kg/ ha (the lowest standing crop for some 18 years on this specific PA).
This would be insufficient for the grazing animals present on the PA. Such information is crucial for early animal management decisions, depending on the degree of risk a manager is willing to take. Such management would be aimed at preventing excessive animal die-off and veld degradation.
Know which species is being managed
Species such as wildebeest and zebra that were showing encouraging increases should not be reduced unduly (Figure 2). The lion population can relatively quickly push these and other more sensitive species (including waterbuck) into a ‘predator pit’, as happened to wildebeest and zebra between 1997 and 2002 (Figure 2) under high predation pressure. The latter situation required predator – in particular lion – management.
Removing species such as impala should be considered with caution, as impala are an important buffer for other prey species that may be under pressure. But the grazing resource is crucial. To address this situation, removing around 20 buffalo would ensure just sufficient food for the grazing population, an obvious oversimplification used here purely for illustrative purposes.
With a good Year 2 season, the stressed grazing situation never materialised. If we feed the Year 2 standing crop (about 2 100kg/ha) and project an increase in animal numbers minus predation (actual data obtained from the Protected Area concerned), anything less than 680kg/ha would result in a shortage of grazing.
This is because a grazer population on a resource close to ‘ecological carrying capacity’ seldom increases at the rate achieved on surplus resources (on the fast part-logarithmic section of the growth curve). The point at which grazing stress becomes an issue increases from 680kg/ ha, assuming that reduced animal increment levels for the abovementioned reasons result in more grass but are still stressed to ‘break-even’.
At 600kg/ ha, it would be difficult to reduce the number of buffalo in one exercise to get to the ‘break-even’ point, as this number would have to be reduced from around 1 150 to around 900 (a 10% increase in buffalo from 100 is 110). The other species would also increase in number.
Is this logistically practical? We must also consider other species. For example, 700 impala could be removed to stabilise the situation. As stated above, we must be wary of reducing species such as wildebeest and zebra (both increasing) as well as waterbuck, due to their susceptibility to heavy predation. But the above assumes a drought situation and we are coming to the end of a run of good seasons.
The good news is that sufficient grazing and offtake, depending on rainfall, should be aimed at maintaining this situation. A staggered offtake is logistically preferable but things can quickly get away. In a fenced area where animals cannot move from, the situation is even more critical.
An active adaptive management approach means that, in a worst-case scenario:
- We suffer a drought;
- We lose animals;
- We have to take pressure off the veld;
- We have to consider feeding in some cases;
- We recoup economically from the offtake.
The best-case scenario would be that:
- We do not suffer a drought;
- We lose animals through natural attrition;
- Pressure is taken off the veld;
- The veld remains in good condition;
- We recoup something from offtakes.
In unfenced protected areas, another option in terms of management is obviously possible – a laissez faire or hands-off approach. However, a population cannot increase at a consistent rate under stressed conditions, so a drop off in natural increments is inevitable. So we use adaptive management to grasp opportunities (allow numbers to climb) and avoid hazards (large scale die-off related to veld degradation).
In many Lowveld protected areas, stocking rates for different species would require relatively large-scale management to reduce the herbivore numbers to adapt to any decline in veld condition. As the grazing resource is generally limiting, grazing species in particular require constant monitoring (removal, feeding or no action). These ‘managed’ animals would be those
not removed by predation but are necessary to remove for ecological reasons.
At the same time, we must be careful not to push prey species into a ‘predator pit’, but still strive to achieve the ecological and economic objectives of the protected area in question.
Email Mike Peel, Rangeland Ecology Group programme manager at the Agricultural Research Council, at