Scientific Services

Almost twenty years of experimental herbivore exclusion in the Kruger National Park has shown that nutrient-poor savannas are quite resilient to herbivore impacts with fewer changes in trees and shrubs than expected

Sodic site in the full exclosure (top) and outside the exclosures (below). The exclusion of herbivores in the full exclosure resulted in higher densities of woody plants, but little change in species composition. Sickle bush thickened up and then had high mortality during the 2015/15 drought (Photos: Corli Coetsee).

If you walk through the “full” exclosure at Nkuhlu in southern Kruger National Park, which should actually be called the “empty” exclosure because fences mostly keep all animals out, you are experiencing a world where plant-eating animals do not exist in Kruger. The canopy is thick and shady, and at first, you enjoy being in this different world. However, after a while you get tired of bundu bashing through the thick grass. You notice that you cannot see very far, and you start to hope that no animal accidentally managed to find its way into the exclosure or that a snake is not lying in your way, because you will step on it before you see it.

The Nkuhlu exclosure was constructed almost 20 years ago with the aim of understanding what would happen when plants are protected from all herbivores larger than a hare. In addition to the full exclosure, there is another exclosure close by, known as the “partial exclosure”. The design of the partial exclosure ensures that elephants and giraffes (i.e. mega-herbivores) are excluded, but all smaller meso-herbivores and larger herbivores, such as buffalo and rhino, have access. Both exclosures are on the banks of the Sabie River, close to Nkuhlu picnic site, and incorporate different types of vegetation, starting close to the river (i.e. riparian vegetation) and moving across sodic sites to the top of the crests.

We used data collected in 2002 and 2019 on trees and shrubs both inside and outside the exclosures to answer the following questions: 1.) How did the exclusion of herbivores affect woody density and structure across the different types of vegetation? 2.) Did the impact of herbivores result in some species being lost and others increasing, such that woody species composition was significantly different inside and outside of the exclosures? 3.) Did herbivore exclusion result in a higher proportion of species that are good to eat (i.e. palatable) for animals?

We found that 17 years of herbivore exclusion impacted mostly the heights and densities of existing species, with little change in the type of species. Densities of trees and shrubs increased in the full exclosure, but the change was more moderate than expected (increased with only 350 individuals per ha) (Fig. 1). A large number of the dominant encroacher species, Dichrostachys cinerea (sickle bush), died inside the full exclosure during the 2015/16 drought, which may have played a role in keeping densities lower than expected. On the other hand, where elephants and other herbivores were present, they (1) reduced the density of woody plants to the point of counteracting the thickening of trees and shrubs (number of trees and shrubs decreased by 780 ha-1), (2) reduced the number of large trees (> 5m), and (3) reduced the possibility of trees growing taller and larger. Meso-herbivores alone, without elephant and giraffe, did not change the densities of individual trees and shrubs but decreased height and canopy size in certain vegetation types. Surprisingly, there was no significant recruitment of new trees and shrubs when protected from all herbivores. The lack of new seedlings, especially for palatable species, despite protection from herbivory suggests that processes beyond elephants, such as low seed dispersal and seed predation are important.

Figure 1. Counts of individual plants per plot from two time periods (A) and changes in counts (B). Treatments are open (O – control), partial exclosure (P) and full exclosure (F). Although tree densities decreased in the open to herbivory treatments, the changes were much lower than expected.

Exclosures, such as these at Nkuhlu, are important long-term research infrastructure that help us understand how savannas work. Across the world’s savannas, shrubs and trees are in the process of thickening (sometimes called encroaching), possibly due to higher CO2 levels, and many savannas are increasingly resembling the “empty” exclosure. At this nutrient-poor sandy soil site with high herbivore presence due to the closeness of the river, we have learned that elephants, giraffe and even smaller herbivores to a certain extent contribute to maintaining open savannas. We have also learned that palatable species, such as marula, will not recruit in large numbers at this site when herbivores are kept out. Long term monitoring will continue to determine under which conditions recruitment may happen and we look forward to those results.

This article was originally published in the 2022/2023 Research Report.

This article is based on a publication in Austral Ecology

Coetsee C, Botha J, Case MF, Manganyi A & Siebert F. 2023. The hard lives of trees in African savanna—Even without elephants. Austral Ecology 48: 532-551.