Search for biocontrol of invasive American bramble intensifies

Rubus section Arguti plant, Cedara. Photo: Costas Zachariades

Invasive American bramble is a thorn in the side of foresters, farmers and land managers across large swaths of South Africa. It chokes up grasslands, forest fringes and river banks, and is notoriously difficult to eradicate. But there is light at the end of the tunnel as a team of scientists are tracking these elusive invaders to find an effective biological control …

American bramble continues to be a major scourge to agriculture, forestry and biodiversity conservation in many of the temperate areas of KwaZulu-Natal, Mpumalanga and Limpopo provinces. The weed forms impenetrable, thorny thickets which impede the passage and access to water of livestock and other animals, replace grazing, smother young plantation trees and make the maintenance and harvesting of older trees difficult. Bramble infestations replace native vegetation, with negative consequences for natural ecosystems, particularly in temperate grasslands. They can also negatively affect specialist flower-visitors. Native bird species increase the spread and germination rates of invasive alien brambles.

There are a number of indigenous bramble species in South Africa, as well as several invasive alien ones. These all belong to the genus Rubus, which falls under the rose family Rosaceae. There are also many species, hybrids and varieties of cultivated Rubus. The most well-know of these are the blackberries and raspberries, but they also include youngberries, boysenberries, cloudberries, dewberries and loganberries. There is a small berry industry in South Africa, but most of that sold in our shops is imported – Mexico, for example, is currently one of the main exporters of blackberries worldwide.

Examples of fruits on Rubus section Arguti, southern KZN. Photo: Grant Martin.

The biology of Rubus is somewhat unique, in that most or all species, although perennial, have a biennial flowering and fruiting cycle. “Primocanes” grow from the ground in the first year – long, robust stems which bear no flowers or fruit. In the second year, these become “floricanes”, which bear the flowers and fruit, and subsequently die back. The study of Rubus is also quite specialized, and comes with its own moniker – “batology” – while those who work with Rubus are known as batologists!

The genus Rubus is large and complex, and is characterized by its ability to hybridise. The genus is divided into a number of Subgenera, and within each of these, one or more ‘Sections’. It is widely distributed, with native representatives on six continents, and invasive alien species and hybrids are also widely distributed and cause great harm in certain areas. In South Africa, American bramble is the most damaging of the invasive brambles. Several introductions of various brambles into South Africa were made in the late 19th and early 20th century, chiefly with berry production in mind. By the 1930s, however, American bramble was becoming problematic: an early reference to this bramble as “Rubus cuneifolius” was by E.J. Philips and co-authors in 1939, in “Farming in South Africa”. Rubus cuneifolius is native to Florida and the southern states of the USA, with the common name “sand blackberry”. However, it was soon realized that there was more than one form of this bramble; J.P. Marais, in a 1960 report, divided it into the “Hilton Road variety” (which was shorter, more upright, and grew in more open areas) and the “Richmond variety” (taller, with more arching canes, growing more prominently in more sheltered areas with partial shade).

Jacobus Egberink carried out some of the first comprehensive studies on the weed and its control as part of his MSc in Agriculture through the University of Natal (now UKZN), completed in 1965. Dr Danie Erasmus, based at the Cedara campus of the Plant Protection Research Institute of the national Department of Agriculture (now the Agricultural Research Council’s Plant Health and Protection [ARC-PHP] institute) conducted further work in the 1980s, including on chemical control. Various other studies, on the biology and taxonomy of Rubus in South Africa, were also undertaken in the 1980s by Prof. Charles Stirton, Dr Johan Spies and Henriette du Plessis.

Rubus section Cuneifolii invasion in the Drakensberg. Photo: Michal Sochor.

Worldwide, biological control of Rubus species initially achieved low success, mainly because of the complex nature of the genus, in particular its tendency to hybridise, and therefore difficulties in finding natural enemies in the region of origin that were able to develop on the introduced target weeds. With the advent of genetic techniques, success rates have increased. In South Africa, the first attempts towards Rubus biocontrol were undertaken by Dr Mike Morris and colleagues of the Agricultural Research Council’s Plant Protection Research Institute in the 1990s, using plant pathogens (rust fungi). They discovered that one of these (Kuehneola uredines), already widespread in the country, only developed on the upright form of R. cuneifolius, and was not particularly damaging. They then imported another rust fungus (Gymnoconia nitens) from Florida, USA, where it had been collected off R. cuneifolius, into their quarantine laboratory. However, this fungus only infected some specimens of the sprawling form of American bramble, as well as a commercial variety of Rubus and a native species, so it was rejected as a biocontrol agent. Because of the differences in infection patterns between upright and sprawling forms of R. cuneifolius, Dr Morris and his team believed that these might in fact be two separate species; they also realized that the upright form tended to grow at higher altitude than the sprawling form. In the early 2000s, ARC-PHP attempted to initiate genetic work in order to understand origins and identities of the forms of American bramble present in South Africa, in collaboration with Dr Lawrence (Larry) Alice of Western Kentucky University, USA, but this project did not come to fruition.

Recent efforts – from 2018 on

Given the ongoing problems caused by American bramble, interest in undertaking a feasibility study revived in 2018. A small ARC-PHP project (managed by Dr Costas Zachariades) was granted funding by the Department of Environmental Affairs (its Natural Resource Management Programmes directorate, which includes the Working for Water programme). At a similar time, the recently formed Centre for Biological Control, attached to Rhodes University, initiated a project on northern temperate weeds, under the management of Dr Grant Martin. These two units collaborated. An M.Econ. student, Brett Mason, undertook a study looking at some of the costs and benefits of Rubus in South Africa. Coincidentally, in 2017 a young dynamic researcher from the Centre of the Region Haná for Biotechnological and Agricultural Research, Olomouc, in the Czech Republic, Dr Michal Sochor, had started a study on the taxonomy and phylogeny of Rubus in South Africa, in collaboration with Dr John Manning of SANBI. Dr Sochor had previously undertaken research on the taxonomy and phylogeny of European brambles, and was thus highly experienced and knowledgeable; the European approach differs dramatically from the current American approach: European researchers tend to recognise many more species than their North American counterparts – while the latter are “lumpers”, the former are “splitters”, and will describe “microspecies”. This divergent approach has not been consistent; for example, from 1941-1945, Dr L.H. Bailey undertook the most recent comprehensive revision of the genus in North America, and recognized hundreds of species, many of which he described himself. In stark contrast, Dr Larry Alice, in a 2015 article, sank the entire Rubus section Arguti, consisting of about 110 species listed in Bailey’s monograph, under one species, R. pensilvanicus. While the assignment of variable forms to separate species or microspecies may be complex and confusing, the lumping of many variable forms under one species is not helpful for the purposes of identification of invasive forms and determination of their origins.

Close-up of Rubus section Arguti plant, Cedara. Photo: Costas Zachariades.

Dr Sochor and colleagues used several techniques in their work, including extensive field and herbarium studies across South Africa, aimed at clarifying Rubus taxonomy in the region with the help of DNA – ploidy estimation and assessment of reproductive mode. They have subsequently published their findings in two scientific papers: the first, in 2018, deals with Rubus in the Cape Floristic Region, while the second, in 2022, examines the entire country, and is thus more relevant for the purposes of American bramble. They found that the upright form of what had been previously referred to as Rubus cuneifolius is a separate species, and in a separate section of the genus, to the sprawling form. Unfortunately they were unable to put species names to these, and refer to them only as Rubus section Cuneifolii (upright) and Rubus section Arguti (sprawling). Rubus section Cuneifolii is found predominantly in KwaZulu-Natal, while Rubus section Arguti is more widespread, occurring predominantly in KZN, Mpumalanga and Limpopo. Furthermore, the Arguti plants could be divided into two commonly occurring forms. Interestingly, both Rubus section Cuneifolii and Rubus section Arguti are “facultatively apomictic” (meaning that they can reproduce asexually) – apparently this indicates that both of these invasive Rubus are in fact hybrids, not true species; in discussion with Dr Sochor, he felt it was likely that such hybridization had occurred in North America, under natural circumstances, prior to the plants being imported into South Africa. Dr Sochor and colleagues also identified two hybrids of which one parent was Arguti and the other was one of two indigenous Rubus species, but did not find any hybrids of Cuneifolii.

Left: Dr Grant Martin. Centre: Rubus section Arguti invading young pine plantation, southern KZN (photo by Grant Martin). Right: Dr Costas Zachariades.

A year prior to the publication of the 2022 paper discussed above, Dr Bram van de Beek, a Dutch theologian who had devoted many years to the study of Rubus in South Africa, published an article focusing on the Cape, although he examined material from across the country, using only taxonomic features (i.e. no ploidy or reproductive methods). Collaborating with Dr Mark Widrlechner, an expert on Rubus taxonomy at the University of Iowa, they identified one of the two sprawling forms (Rubus section Arguti) as Rubus originalis and described the other as a new species, Rubus revealii. For the upright form, they identified a few (“stronger”) plants from KZN as Rubus pascuus but used Rubus probabilis for most plants. Dr Sochor does not feel confident in these identifications; in general we have aligned our work with Dr Sochor rather than Dr van de Beek, but we are also working with Dr Widrlechner in the USA.

In order to familiarize ourselves with the South African Rubus flora (both alien and indigenous), we joined Dr Sochor on one of his fieldtrips to South Africa, in early 2020. Despite our initial confusion as non-botanists, we soon found it quite easy to distinguish between various species based on characteristics such as leaf shape and flower colour. This trip also gave us an opportunity to look for natural enemies present on both alien and indigenous Rubus species. This proved interesting, as we found many more species of insects and pathogens on the indigenous species than the alien ones – although this is expected, it does give an indication that many natural enemies of Rubus are specialized, and secondly that were we to introduce natural enemies from North America onto these alien Rubus plants in South Africa, they have the potential to reduce the invasiveness of these plants.


Comparison of flowers and leaves of Rubus section Cuneifolii (top) and Rubus section Arguti (bottom). Photos: Michal Sochor.

Current work and the way forward

What is the relevance of the recent taxonomic and phylogenetic studies discussed above to our biocontrol project? The lack of much hybridization, together with the weediness of the plants, led us to restrict our focus to invasive North American Rubus i.e. plants previously falling under “Rubus cuneifolius” in South Africa. In order to progress, we need to firstly understand how many species, and how much genetic variability exists in these species in South Africa. We hope that it allows us to find plants growing in the USA which are close matches to at least some of these invasive Rubus, and thereby find potential biocontrol agents (insects, mites and pathogens) which are compatible with the plants. To achieve the first goal, Dr Sochor agreed to undertake genetic analysis of these species – we therefore undertook a fieldtrip across KZN in early 2023 to collect as much genetic material and herbarium specimens as possible (the latter have been lodged in the Bews Herbarium at UKZN, Pietermaritzburg). This fieldtrip confirmed previous observations that the upright form (Rubus section Cuneifolii) occurs more commonly at high altitude (KZN Drakensberg), while the sprawling form (Rubus section Arguti) occurs more commonly at lower altitude (KZN Midlands). Dr Martin also opportunistically collected some specimens in the USA while on a fieldtrip there for other purposes. Dr Sochor has conducted some analysis of our specimens, and has concluded that while the upright form is genetically quite homogeneous and consistent with a single species/hybrid, the sprawling form in KZN consists of three species/microspecies/hybrids (we still need to sample Rubus section Arguti in other provinces). Furthermore, he did not find a close match between the South African and North American specimens sampled, and the North American samples displayed a high level of variability amongst themselves.


Stem girdles caused by insect larvae on two indigenous Rubus species. Photos: Brett Mason.

So the identification of North American plants which are genetically close to ours remains a critical step. One way to do this is to obtain genetic material from Rubus herbarium specimens in the USA which are morphologically similar to our invasive ones. Dr Sochor has found that leaf material from herbarium specimens, even those over 100 years old, can yield good DNA. Dr Widrlechner has agreed to assist in obtaining such material, and also in re-examining Rubus specimens of species said to be similar to ours. Bearing in mind that both Rubus section Cuneifolii and Rubus section Arguti in South Africa are hybrids, we may not find a perfect match among herbarium specimens in the US, but we hope this exercise gives us some direction. If so, we can transfer our attention to the field in the US – to areas where these herbarium specimens were originally collected. Again, it would be extremely helpful if a local taxonomist such as Dr Widrlechner could assist us to identify these plants in the field. From there, there are two options, viz. (i) to survey these plants for natural enemies, and to import such natural enemies into quarantine in South Africa; (ii) a better option would be to plant out South African material among genetically similar plants in the USA on which natural enemies are present, and allow these natural enemies to colonise our South African plants on their own. In this way we will be more likely to obtain potential biocontrol agents which are compatible with our plants, and thus more likely to be successful in the field in South Africa, should they prove to have a sufficiently narrow host range (i.e. do not attack native or commercial Rubus) to be safe for release in South Africa. Whether the US biosecurity authorities would permit us to plant out South African Rubus is uncertain, but we plan to apply for permission to do so.

There is a final spoke in the works, and that is a lack of current funding. Funding from DFFE: NRMP became more erratic in 2018, and dried up completely in 2023, with no prospect of revival in the short term. CBC itself has funded some of the work since then, but its means are limited. Adequate funding would allow the US work described above to be undertaken properly, and, should natural enemies be found there that show promise as biocontrol agents, to import these into South African quarantine in order to conduct host-range testing.

Funding notwithstanding, what seemed in the 1980s and 1990s as an intractable, complex situation is now resolving itself into a more manageable research project, with some light at the end of the tunnel due to improved understanding of Rubus taxonomy and phylogeny. It is not inconceivable that within the next 10 years, an effective biocontrol agent could be released for one or more of the invasive North American brambles in South Africa, resulting in reduced vigour and competitiveness of these plants, and correspondingly, more cost-effective management using non-biocontrol methods.

Plant of Rubus section Cuneifolii. Photo: Michal Sochor.

Authors:-
C. Zachariades, Agricultural Research Council’s Plant Protection Research Institute
G. Martin, Centre for Biological Control

Notes from the field
Roger Poole, Member Services Co-ordinator for NCT and Agro-Chemical Liaison Officer for the Timber Industry Pesticide Working Group (TIPWG), provided some useful notes on the chemical control of American bramble …

Bramble is a tricky one due to it having two stages of growth, these being the older stems one always sees and then there is secondary (new) growth you'll find inside the thicket. Timing is critical. Best months to spray are between February through to April as the plant is building up reserves for winter so absorption of herbicide is the most efficient.

The best herbicide is metsulfuron methyl, it is slow acting and gives the best results. If there are other invasive species within the area of treatment, then one can look at glyphosate or the picloram/fluroxypyr formulation.

Glyphosate is not the best but does work on bramble that has been cut down. Depending on the size of the thickets one can apply it with a high pressure unit (bakkie sakkie) or tractor boom sprayer. Knapsacks only work on bramble that a person can walk through so you’ll have to cut the thicket with either a brush cutter or tractor-mounted slasher.

Aerial application has been done previously but water volumes need to be checked and applied as per label due to the need for penetration to ensure the mixture gets through the thicket and results in a good coverage.

Grassland dynamics & bush encroachment in forestry plantations

By Lize Joubert-van der Merwe, Veldtology (Pty) Ltd
I really like grasslands. I especially like how they ripple in waves up and down hill and mountain slopes when there is an approaching thunderstorm; how they change to that rich golden color in the final sunspots just before the dark-grey, rolling thunderclouds and lightning chases you indoors; and how their inflorescences hold rain drops from the previous night like a chandelier of diamonds. My fascination with grasslands extends beyond their aesthetic appearance to also include their ecology, management, and why this matters to forestry.

Moisture and temperature shape grasslands
Grasslands are so vast that we often accept their presence as ubiquitous, yet, they are constantly changing in response to natural and anthropogenic drivers of diversity. I interpret grassland diversity from an understanding that moisture and temperature influenced broad vegetation patterns over the past few thousand years, as outlined by Frank Neumann. Did you know that the current wet-and-warm climatic period has only been around for 800-1000 years? A cooler period with less fires (more than 13 000 years ago) caused grassland to have many more fynbos elements, and we still see relics of Protea and Erica communities growing on cooler, south-facing slopes (Figure 1 above). Much later (~4600 to 3500 years ago), there was a drier period when grasslands saw an increase in karroid elements, specifically Pentzia incana (Ankerkaroo) that nowadays dominates sheep farms in the central Karoo.

In the current wet-and-warm climatic period, grasses dominate in grasslands (hence, the name), but they still have to compete with flowering forbs, trees and alien plants to remain numero uno. For this, they use various competitive strategies. Grasses keep flowering forbs at bay by rapidly growing into a dense layer that intercepts heat and sunlight from (s)lower-growing plants. This strategy to monopolize access to sunlight is quite a dicey move, because grasses are themselves not tolerant of shading. In fact, it happens in the absence of fire and grazing that build-up of leaf litter and moribund grass blocks sunlight from reaching live buds and leaves, which causes die-back of grass tussocks – a phenomenon known as ‘self-shading’. This is why fire is such an important part of grassland management. Fire is truly the exfoliating treatment that removes dead and dry cells from grasslands, so that new life can flourish.

Importantly, the ability of grasses to outgrow forbs and intercept limiting resources is directly tied to the current climate. During periodic droughts, when grasses cannot maintain their productivity levels, forbs are quite capable of recruiting successfully from seeds (Figure 2). Similarly, grasslands subjected to severe overgrazing are not able to keep forbs in check, leading to an overabundance of flowering forbs that is sometimes even visible on satellite images (Figure 3).

The role of fire
If we shift our focus to the woodies in our midst, grasses keep shrubs and trees in check by sustaining a ‘fire trap’ from which tree seedlings hardly ever emerge unscathed (Figure 4). A fire trap is essentially the fire flame zone of the grassy layer. Unlike grasses, most indigenous shrubs and trees are sensitive to fire, especially as seedlings. So, fire gives grasses the competitive edge over shrubs and trees, just like climate gives grasses the competitive edge over forbs.

Grassland with a well-developed grass layer that burns at the correct intervals (when biomass ~ 4 tons / ha) should have no problem with invading trees. However, where the grass layer is jeopardized by too frequent burning, overgrazing or shading by large trees, fire intensity will be lower with consequently less killing power to aim at invading tree seedlings. For example, in communal rangeland (with heavier grazing → less grass → cooler fires), it often happens that tree seedlings escape the fire trap and grow into bigger trees that are more fire tolerant. Shading by timber trees also play an important role in advancing bush encroachment into grassland, especially in narrow corridors of forestry plantations. In fact, shading might explain much of the ‘edge effect’ of timber on adjacent vegetation, previously reported by Prof. James Pryke.

The role of atmospheric CO2
Interrogations of the local and global drivers of bush encroachment have led to a growing consensus among researchers that elevated atmospheric CO2 levels is an important global driver of bush encroachment. The exact mechanism is still unknown, but possibilities include the fertilizer effect of atmospheric CO2 on woody shrubs and trees, or an indirect effect on soil water content and its depletion in the surface soil layers where grass roots sit. Encouragingly though, a team of researchers led by Prof. Sally Archibald and Prof. William Bond found that bush do not encroach as rapidly in protected areas with elephants – the big giants of Africa that create their favored grassland habitat by pushing over trees. Although I am not advocating for the introduction of elephants to eradicate bush, this shows that local actions can trump global drivers in shaping vegetation dynamics. This is indeed encouraging.

Practical solutions customized to local context
The trick is to find practical management solutions that can be applied in forestry plantations to help control bush encroachment. Such solutions will probably involve a combination of management actions sustained for longer periods of time, rather than single once-off interventions. For example, it would be pointless to do a once-off clearing of dense stands of Ouhout trees, with no follow-up burning and thinning operations to keep shrubs and trees in check. Moreover, instead of looking for a silver-bullet strategy that works well everywhere, management actions would probably need to be customized to fit local context and challenges.

Key local issues will include the shape and size of conservation areas. By virtue of their close proximity, any management intervention inside a narrow, small or irregularly-shaped conservation area has a greater probability of affecting adjacent timber compartments, than if you had a wider or larger conservation area. Thus, when a decision has to be made to control bush encroachment in one conservation area (but not another), shape and size is a useful starting point. In fact, it is non-negotiable that the conservation area must be the correct shape and of reasonable size to allow for safe burning.

Additional considerations include proximity to important conservation areas (e.g., with Red-Listed species or threatened ecosystems) and level of wetness. If bush encroachment threatens the functionality of a threatened grassy ecosystem, this is a good reason to prioritize bush thinning operations. Even more so when that threatened ecosystem contains threatened species, such as Long-toed Tree Frogs (Leptopelis xenodactylus), Swamp Nightjars (Caprimulgus natalensis) or African Grass Owls (Tyto capensis) that all depend on grassy habitats.

Lastly, level of wetness seems to influence vegetation succession (grassland -> bushy thickets or forest) and / or how the wetland ecosystem responds to bush thinning and burning. This is beautifully shown in the delineated areas of Zululand, where wetter wetlands have a greater tendency to remain grassy, whereas drier wetlands have a greater tendency to become bush encroached. We do not quite understand the mechanism of this phenomenon - it might be that grasses (with shallower roots) respond quicker than shrubs and trees when there is a shallow water table present. If this is the case, it will mean a better ecosystem response to burning, because a healthy grass layer is better able to sustain a fire trap that kills tree seedlings. Personally, I would consider shortlisting wetter wetlands for bush thinning and burning.

The role of roads in shaping fire
It makes logical sense that all management operations should be aligned with clearly-visible, on-the-ground features so that operators know where to work. Such features can be roads, trace belts, streams or fence lines, depending on what is available. Where fire management is concerned, roads (mostly vegetated or dirt tracks) work exceptionally well, because they also provide access to vehicles and fire-fighting equipment, and should have low fuel loads (due to routine road maintenance). This makes it possible for foresters to set alight vegetation along the road, so that the fire burns from the periphery towards the interior of a conservation area, with minimal risk to adjacent timber. This is probably why we find grassland vegetation in larger conservation areas with roads along their edges, but bushy thicket in those without roads (or wrongly placed roads) (Figure 5). Exceptions include narrow or irregularly-shaped conservation areas that will probably not burn, regardless of presence or absence of roads, because of risk to adjacent timber. Another exception is conservation areas on steep slopes, where management (also roads placement and burning) would be adjusted to fit the soil erosion risk profile.

The value of well-placed roads is not new and already embedded in wetland delineation procedures for some forestry companies. Especially in Zululand (where terrain is not a problem), valley-bottom cut-off roads are routinely implemented at the edge of conservation areas (where they join commercial timber) to mark new compartment boundaries, to provide access, and to enable the use of fire in alien plant control (Figure 6). Getting control of alien plants within the first few years after felling timber is a major delineation goal, because it feeds into water security and sets the direction of ecosystem recovery in terms of biodiversity. Most of the roads around conservation areas have vegetated surfaces and are not expensive to maintain (Figure 6), but they make the world’s difference in restoring delineated land to a semi-natural state.

I think many environmentally-minded people (including myself) have been blinded by the negative impacts of roads, notably in connection with soil erosion and sedimentation. Perhaps, it is time to recognize that well-designed road networks (with roads that are well-placed, well-drained and well-maintained) can be conservation assets too.

Using fire in alien plant control
Fire, along with foliar herbicide sprays and cut-stump applications are your cost-effective tools in the fight against alien plant invasions, notably American Bramble. Of these, fire followed by foliar sprays is the most cost-effective treatment option available in grasslands, but it is seldom (if ever) used in dense thickets.

Fire in grassland simplifies access, reduces the size of alien plants and causes a flush of new growth, which is more susceptible to foliar sprays than stems and mature leaves (Figure 7). However, of critical importance is the timing of post-fire follow-up sprays to hit the flush of new vegetative growth just at the right time, i.e., when plants are between knee and hip height. Get the timing wrong, and it is back to square one. No alien plant control operation should start without a viable follow-up plan that can be implemented with available resources - money, manpower and the necessary expertise to guide effective alien plant control.

Different stages of bush encroachment
Bush encroachment is a gradual process of indigenous shrubs and trees replacing grasses often over a period of >10 years. Drought with uncontrolled grazing and shading of the grass layer in narrow corridors can increase the rate of bush encroachment, while expeditious burning can delay or stop the process. At the end of the day, there will be different stages of bush encroachment in a forestry landscape, with at least some alien plants that need to be controlled.

Early stages of bush encroachment (when shrub and tree cover is still sparse) should be prioritized for intervention, because the cost-effective management of alien plants with fire and foliar sprays is still possible. Basically, foliar sprays can be used until alien plants are about shoulder height, but do remember that bigger plants → more herbicide → greater cost. For alien plants above shoulder height, cut-stump applications are your next-best option, but at a far greater expense. Even larger specimens can be frilled or ring-barked, which are labor-intensive and time-consuming operations. Here, you must ensure it is done correctly to ensure maximum effectiveness.

For late stages of bush encroachment (dense thicket or early-successional forest), a different mechanism is used to effectively control alien plants. Here, the tree canopy effectively intercepts sunlight from reaching the soil surface, which prevents alien seeds from germinating. Dense thickets with an intact tree canopy generally do not have a problem with alien plants. It is only where there are gaps in the canopy (due to windfall or along thicket edges) that alien plants can establish, and where they need to be controlled.

A word of caution, though. Chopping down a large, solitary pine or eucalypt tree inside a dense thicket patch will create a gap in the tree canopy that presents an opportunity for alien plant recruitment. It is better to ring bark or frill such trees, so that surrounding indigenous trees are not damaged and so that there is not a sudden flush of sunlight available at the forest floor. The longer ‘time-to-kill’ for ring barked or frilled alien plant specimens also leaves a window of opportunity for indigenous tree species to fill the tree canopy gap, which effectively removes the alien plant recruitment opportunity. The effective control of alien plants in dense thickets considers treatments along with this careful manipulation of shade and sunlight on the forest floor.

In contrast to early and late stages of bush encroachment, there is an intermediary bushy state that presents a problem to management, and also has less biodiversity value than both more-grassy and more-forested states, according to Dr. René Gaigher. Here, fire cannot be used anymore (due to lack of grass cover) and the tree canopy has not yet locked out sunlight. This stage is susceptible to alien invasion, but it is difficult to gauge extent of invasion due to poor visibility and accessibility. For the same reasons, alien plants are difficult to find and treat. Viable treatment options in this context are expensive and time-consuming: cut-stump applications and frilling. Foliar herbicide spraying is an option along edges, but cannot be recommended for the interior of these bushy patches.

Where it makes sense to do so, the long-term strategy to control alien plants would be to reverse the intermediary bushy state back to grassland, so that fire and foliar spraying can again be used as treatment options. This will also benefit biodiversity. However, it will be an expensive and difficult journey of sustained effort for many years, which necessitates careful consideration of the points raised earlier (size and shape, important conservation areas, slope direction, and level of wetness). For all other conservation areas that is still in a predominantly grassy state, it is of utmost importance to maintain that grassy state with appropriate management.

Grassland for water production
Probably the greatest benefit of functional grasslands in conservation corridors involves their ecological function in the sense of water production. Grasslands use less water than bushy thickets, and much less than alien vegetation. According to the National Water Act (Act 36 of 1998), commercial forestry is a stream-flow reduction activity that requires a water use license to safeguard national water security. It is this legal framework that enforces wetland delineation and control of alien plants along waterways, but it does not stipulate desired natural vegetation type (grassland vs. bushy thickets vs. forest) once timber along streams and around wetlands is felled. If the objective of legislation is to safeguard water security for downstream users, it would seem advantageous to have more grassland and less bushy vegetation in riparian and wetland buffer zones.

However, conservation corridors are not just water production areas. They also conserve biodiversity and ecosystem function, specifically ecological values representative of the historic state before timber dominated these landscapes. If the historic state in Zululand is coastal forest along streams, with grassland a bit further way, there is no reason for bush thinning operations in the riparian zone. Burning of grassland adjacent the forested riparian zone will maintain a functional ecotone and ensure that the coastal forest do not expand to dominate the entire drainage line. Maintaining this natural range of habitat types (grassy and woody types in wetter and drier areas) will tick the ‘biodiversity conservation’ block along with the one for water production.

A bit of practical wisdom also goes a long way for the management of rugged, south-facing, bush encroached hillslopes in the KZN Midlands. If the terrain is too rugged to have roads (to help safe burning) and if the microclimate on that hillslope is too cold and wet to sustain a fire, it might be best not to intervene with bush thinning operations. However, conservation areas on warmer hillslopes that jut down to rivers and with bush encroachment that can be traced back to a clear starting point (such as a change in ownership or retirement of an experienced forester) are good candidates for bush thinning operations that will probably also benefit water production.

Lize Joubert-van der Merwe is an independent consultant specialising in sustainable agriculture and forestry through improved management of natural resources.