Figure 4: There does not have to be a tug of war between cattle grazing on the one hand, and biodiversity conservation on the other. It is possible to have conservation and cattle grazing in mesic grasslands, but this requires careful management.

By Lize Joubert-van der Merwe
Ecological networks (ENs) among forestry compartments can conserve biodiversity and ecosystem services as well as connect nearby nature reserves if they are managed and designed appropriately. This article focuses on the effect of cattle grazing on biodiversity of grasslands in the wetter 
(>650 mm), eastern part of South Africa.

Sustainable management of resilient ecosystems
Grasslands are dynamic, diverse and resilient. A dynamic system has many ecosystem components with many potential movement pathways. Ecosystem resilience is the ability to absorb and respond to change, and resume original function after drivers of change are removed.

Historically, the external pressures that made species move were evolutionary drivers (e.g. plate tectonics) and ecological drivers (e.g. climatic changes) that operated over many, many years. Therefore, species had time to adapt genetically or behaviorally to the changing environment.

Nowadays, management acts as an additional external driver. It can be seen as the condensation of historical drivers into 2-3 human lifespans. So species do not have time to adapt. They are like professional athletes that do not have time to prepare for a race. They have to depend on muscle memory. The muscle memory of species lies in following evolutionary and ecological trajectories up to their limits, but not beyond those limits. To be sustainable, grasslands should be managed within the limits set by evolutionary and ecological drivers.

Recommendation: Management practices should operate within the limits set by evolutionary and ecological drivers to be sustainable. If management practices cause species loss or radical shifts in assemblage composition that result in the influx of alien species, chances are good that they are approaching their limits.

The importance of context
The effect of grazing on grassland biodiversity should be viewed within the greater landscape context of fire and climatic regimes. Grasslands change drastically in response to fire (as discussed in the previous article) as well as to weather cycles. In the same way that some species are fire-adapted, there are species that are adapted to drought.

Let us consider the two main growth forms: 1) grasses that dominate in grasslands and 2) forbs (or flowering plants) comprising the majority (>65 %) of species in grasslands.

Most of these forbs are perennial, and have large underground storage organs (e.g. rhizomes, bulbs or corms) that enable them to persist during adverse conditions (Figure 1). In fact, during the dry climatic period between 4 600 and 3 500 years BP, forbs dominated at the expense of grasses in the Drakensberg mountain range.

Figure 1: Many perennial forbs have large underground storage organs that enable them to survive during adverse environmental conditions.

Nevertheless, grasses have the competitive edge over forbs during humid climatic cycles, such as the one we are currently experiencing (See Box 1 at the bottom of this article: Historical perspective on grassland dynamics). The reason for the competitive advantage of grasses during wetter periods is linked to their greater above-ground productivity, and light sensitivity in most forbs. So when dry grass accumulates, it cuts off sunlight and prevents it from reaching the soil surface. The resulting cool conditions induces dormancy in forbs, which is broken when grasslands burn. To discuss grazing without considering other drivers of diversity, such as the current climatic regime and fire regime represents only one perspective on a multi-dimensional ecosystem.

Recommendation: Grasses (not forbs) should dominate grasslands in the relatively humid climatic cycle that we currently live in. Dominant forb cover (e.g. Acalypha or Eriosema species) could serve as an early warning system to adjust grazing or fire management (Figure 2).

Figure 2: Heavy grazing during a period of drought annihilated all grasses, resulting in a dominant stand of Eriosema distinctum. Dominant cover of forbs or unpalatable grasses can be used as an indication that something is going wrong.

Yes to moderate cattle grazing
To allow cattle grazing in areas designated for conservation seems ludicrous. Is cattle grazing not one of the primary drivers of landscape degradation in South Africa? Yes, indeed. One drive through communal areas will be enough to illustrate to anyone how continuous, heavy grazing can harm a landscape (Figure 3). However, all grazing regimes are not equal (See Box 3 at bottom: Grazing regimes).

Figure 3: Bare slopes with erosion gulleys and Aloe ferox dotting the landscape is a common sight in rural, communal areas in KwaZulu-Natal where continuous, heavy grazing is practiced.

Indigenous fauna and flora evolved with large game species trampling and grazing in the landscape. (See Box 2 at bottom: The historical vs. current grazing regimes.) So it is not entirely surprising that domestic cattle grazing (with less fire) outside the Hluhluwe-iMfolozi Park benefitted grasshoppers by simulating the effect of indigenous game that you would find inside the park (Samways and Kreuzinger 2001).

Similarly, in the KZN Midlands, cattle grazing inside the EN benefitted grasshopper species richness and abundance relative to the adjacent iMpendle Nature Reserve (Joubert et al. 2016). This also applied to plants, which cannot move around to find their required resources. So it appears that cattle grazing and trampling mimics the natural disturbance of grazing by indigenous game (Figure 4 at top of article).

Nevertheless, the presence or absence of grazing is not always the problem. What about different intensities of grazing? Moderately-grazed areas had higher numbers of plant species when compared to heavily-grazed grassland and even to reference sites in iMpendle Nature Reserve that only has light grazing by indigenous game (Joubert et al. 2017) (Figure 5).

Figure 5: Grasslands with moderate levels of cattle grazing have more plant species than natural grassland reference sites in a nature reserve with light grazing by indigenous game.

There are three reasons for elevated species richness in moderately-grazed areas. First, grazers consume grasses which are competitively superior to forbs. So by reducing the competition for forbs, it boosts diversity.

Secondly, moderate cattle grazing reinforces the natural variation among different habitat types in the landscape. So by strengthening the uniqueness of each ecological niche, it creates opportunity for a greater variety of species to co-exist.

Thirdly, especially in grassland with longer fire-return intervals of 3-5 years, moderate grazing stems the decline in plant species richness that naturally occurs in aging grasslands. Interestingly, none of the forbs associated with moderately-grazed areas were alien or indigenous, annual or weedy species. So astute grazing management can contribute to biodiversity conservation.

Recommendations:
1) Manage the competition between grasses and forbs in the same way that competition between palatable and impalatable grass species is managed. Impalatable grasses and forbs will always be part of grasslands, but they should not dominate entire landscapes.
2) Cattle naturally move around in a landscape to select the best grazing available. If rotational grazing is practiced in camps that are large enough, grasslands will maintain their vigor, while the optimal foraging habit of cattle will strengthen the natural variation in the landscape.
3) Managers instinctively know where to find distinct plant assemblages on their land. Guard against the homogenization of those unique assemblages. ‘No’ to heavy cattle grazing

Continuous, heavy grazing by cattle (similar to what was found in communal areas) resulted in elevated levels of bare ground, a shift in plant assemblage composition, an influx of alien forb species, and a decrease in total plant species richness relative to moderate levels of grazing. Together, these results indicate that continuous heavy grazing was not part of the evolutionary history of these grasslands. It is unsustainable, because it pushes grasslands beyond the limits set by historical drivers.

Recommendation: Signs of too heavy grazing include a shift in dominant grasses from palatable to impalatable species, the large-scale appearance of bare patches, and an influx of alien plant species (e.g. Bramble). Consider and monitor these signs carefully if the goal is biodiversity conservation.

Effect of drought
The 2015-2016 drought showed that no grazing regime can be reduced to following a set of numbers. Sometimes, the best plan on paper does not work in practice. Tim O’Connor (1995) found that the negative effects of a very severe drought on grassland productivity can be worse than the effect of continuous, heavy grazing. Specifically, the palatable grasses suffered severely in terms of basal cover, tuft size, and seed production. The good news is that Themeda triandra tufts recovered to its previous size under moderate grazing within two years, even with below average rainfall (See Box 4: The case of Themeda triandra for other interesting facts about this grass). The bad news is that recovery took longer in areas that were heavily grazed during and after the drought.

Many of the forestry plantations border upon communal areas that were already overgrazed before the drought. With poor recovery rates of grasslands in these communal areas, levels of illegal grazing in the adjacent ecological networks (and nature reserves) are expected to continue and increase at unprecedented rates. The control of illegal grazing in ENs will determine their future in terms of biodiversity conservation.

Recommendation: Identify high conservation value areas, and exercise absolutely zero tolerance to illegal cattle grazing in these areas.

Box 1: Historical perspective on grassland dynamics
It is hard to make sustainable management recommendations without an understanding of historical cycles that influenced these grasslands.

In coastal KwaZulu-Natal, humid climatic conditions prevailed 6800-3500 years BP causing an increase of water demanding forest Podocarpus (Yellowwood) trees, and rising water tables. This aided the spread of mangrove and swamp forests. After ~ 3500 years BP, drier conditions prevailed and sea levels lowered, which lead to an increase in grasses at the expense of forest trees.

Frank Neumann and co-workers (2014) also saw changes in climate and vegetation at Mahwaqa Mountain (1850 m a.s.l.), but over a longer period of ~ 18 000 years. Fynbos species, such as Ericas and Restios dominated during the first period (18 000-13 500 years BP) when cool conditions prevailed. These fynbos elements were gradually replaced by grasses and Asteraceae forbs in the next warmer and wetter period (13 500-7 500 years BP). These climatic conditions were maintained until a drier period (4 600-3 500 years BP) caused Asteraceae forbs to replace grasses as the dominant life form. Between 3 500 and 800 years BP, dominance of grasses and water-loving sedges were again restored at the expense of Asteraceae, which indicates a wetter cycle. This humid cycle lasted into the present age.

Box 2: The historical vs. current grazing regimes
Historically, Afromontane grasslands were grazed by small-bodied (< 80 kg) ungulates whose population sizes and spatial occurrence were influenced by seasonal availability and variability in quality of forage. Larger ungulates (e.g. blue wildebeest and eland) were either at the limit of their range, or migratory (O’Connor 2005). The ecological carrying capacity of the KwaZulu-Natal Drakensberg was estimated to be as low as 55 ha / LAU, which might have been due to either low forage quality in winter or due to specific habitat preferences of indigenous game species (Rowe-Rowe and Scotcher 1986).
In comparison, the current grazing regime is dominated by large-bodied domestic cattle that are generalist grazers without specific habitat preferences. Generally, grazing is year-long as opposed to seasonal, although seasonal rest once every 3-5 years is a recommendation for good veld management. The recommended official stocking rate (2.5-3.5 ha / LAU) is almost 20 times greater than in the period prior to permanent human settlement.

Box 3: Grazing regimes
Different grazing regimes have different effects on grassland biota.
• Type of animal: Sheep, goats and cattle have different effects on plant assemblages in grasslands due to their different ways of grazing. Sheep, in particular, have a very large negative effect on plant diversity, because they graze plants shorter to the ground and incorporate a larger proportion of forbs in their diet (O’Connor et al. 2011). A high proportion of sheep in a herd of livestock, and a high stocking rate are the two main drivers of plant diversity loss in natural grassland.
• The recommended stocking rate for grassland in the KwaZulu-Natal Midlands is 2.5 – 3.5 ha / large animal unit. These rates vary across the country according to climate, dominant soil type, and historical disturbances, as reflected in veld condition.
• Rotational system: It is best to graze grassland on a rotational basis, and not continuously as is typical in communal rangeland.
• Seasonal rest for a full growing season (6-12 months) is recommended every 3-5 years for optimal seed production and to maintain vigor of the grass sward.
• When burning to remove moribund plant biomass, grazing of regrowth should be carefully managed to prevent bare patches from appearing.

Box 4: The case of Themeda triandra
No discussion about grazing in natural grassland is complete without mentioning 
T. triandra, which is an economically important, fire-climax grass species. It is prone to local extinction, because it is palatable and produces a low number of large seeds with poor dispersal ability. Careful management and knowledge of the biology of 
T. triandra is required to make sure it continues to dominate in grasslands.
More information can be found in publications by Tim O’Connor and co-workers (1991; 1992; 1996).

*First published in SA Forestry magazine, June 2017