High intensity fires do not reverse bush encroachment in an African savanna

A decade-long experiment on the use of high-intensity fire to control bush encroachment in South Africa’s Kruger National Park (KNP) has revealed that, despite an initial short-term effect, these fires did not result in a meaningful reversal in bush encroachment in the long-term.

The results of this long-term and large-scale experiment were published in the Journal of Applied Ecology today (16 February 2023), in an article titled “High-intensity fires may have limited medium-term effectiveness for reversing woody plant encroachment in an African savanna”.

Prof. Brian van Wilgen, emeritus professor in invasion biology at Stellenbosch University (SU) and one of the co-authors, says most papers that advocate the use of high-intensity fires are based on observation over a short time: “I think this is the first study to assess this management practice’s effectiveness over a decade,” he adds.

Decade-long experiment

Bush or woody encroachment, which is a process where the density of smaller trees and shrubs are increasing, is taking place across the world. This is worrying because increased woody cover results in less grass available for animals that eat grass or use it as habitat. Although the causes are complex and not that clear, a decrease in the use of fire, coupled with an increase in CO_2, may be to blame in savannas.

In 2010 and again in 2013, adjacent sites in the southern parts of the KNP, covering thousands of hectares, were burned using low, medium and high intensity fires. The objective was to examine whether high-intensity fires can be used to reverse bush encroachment. At the time of the first burn application, KNP officials had to deal with substantial negative publicity after some animals were mortally injured.

A study reporting the effects on woody plant cover one year after the completion of the second set of experimental fires showed that woody plant cover did indeed decline significantly over this short period. The initial study, however, cautioned that the long-term efficacy of these high intensity fires still needed to be confirmed.

In 2020, using a combination of ground surveys and remote sensing data, scientists from South African National Parks (SANParks), Harvard University and the Centre for Invasion Biology at Stellenbosch University compared the 2010 data with their 2020 observations. Despite the initial encouraging results, they found that the reduced woody plant cover was not maintained after ten years. Even though there were large differences between the fire treatment sites one year after the experimental fires, these differences had disappeared after 10 years, strongly suggesting that differences in fire intensity did not have a long-term effect on bush encroachment.

Even more concerning was the trend of tall tree loss in the KNP. Over the past decade, trees taller than ten metres declined by about 65% across all the experimental sites, regardless of the fire treatment.

According to SANParks scientist and first author on the article, Tercia Strydom, this was because of a combination of elephant and fire damage: “Tall trees are normally capable of withstanding frequent savanna fires due to their thick bark. But when elephants debark a tree (because they eat it), that thick protective bark is removed, exposing the inner wood. When this inner wood dries out, it burns and smoulders within the tree until the tree eventually succumbs,” she explains.

Lessons learned

According to Prof. van Wilgen, the paper shows the value of long-term monitoring to establish the real outcome, and not to jump to conclusions: “It also shows how complex things can be, and that there is no easy fix to bush encroachment”.

Dr Izak Smit, senior scientist at SANParks and another co-author, says the park continues to follow a strategic adaptive management approach in which local context is paramount: “Best available knowledge needs to be gathered from the literature, and based on that a careful evaluation must inform what management action you consider appropriate to implement in your area of interest. The outcomes of this action should be carefully monitored to see whether or not you actually obtain the anticipated outcomes”.

“This way, ecology reveals its complexity and nuance in different contexts and allows us to fine-tune the way we manage protected areas – learning as we continue to manage in the face of uncertainty,” he adds.

They warn that, on a global scale, inter-continental and even intra-continental generalisations regarding the most appropriate management actions for bush encroachment remain elusive. According to Strydom, this case study may not even be representative of other parts of southern Africa based on differences in elephant density, soil type and rainfall, amongst a plethora of other factors.

The way forward

KNP has a long history of scientific research and looking forward, Dr Smit says they are now exploring the potential of early wet or early dry season fires to address bush encroachment as an alternative to high-intensity fires.

He explains: “When woody species are dormant during the dry season, they move a lot of their resources to their roots while the top parts are ‘resting’. If a fire consumes the above-ground parts during this period, the plant still has a lot of resources stored below-ground from which it can resprout when conditions become favourable. However, if the plant is in leaf and a fire burns the above-ground parts, it is expected that the damage is more costly to the plant, which is what we want when concerned about bush encroachment. The trick is therefore to try to burn when there is sufficient dry grass to carry fires of reasonable intensity, while at the same time the bush species are in a more vulnerable leaf-on phase. Spring sometimes provides this window of opportunity where grass is still dry enough to burn and the woody species are already in leaf and thus more vulnerable to fire damage”.

In the end, he says, one needs to understand both the local (e.g., fire, grazing) and global drivers of change (such as elevated atmospheric CO₂), and their interactions with each other, in order to tease out if and how different management actions (or inactions) may improve the situation, or even make it worse.