by Stuart Hall
Department of Conservation Ecology and Entomology & Centre for Invasion Biology
Stellenbosch University, South Africa
Invasive alien plants negatively impact natural ecosystems through a long term decrease in biodiversity and altered ecosystem functioning. Alien plant removal may mitigate some of these impacts, but in many cases this intervention alone does not facilitate effective restoration of an ecosystem. In a critically endangered vegetation type within the mega-diverse fynbos biome in the Cape Region of South Africa, passive and active interventions were compared in terms of restoration success following removal of invasive Acacia saligna, since this ecosystem is a good example of where clearing alone has often failed to facilitate recovery of the native ecosystem.
Lowland Fynbos typical vegetation structure and taller invasive Acacia saligna behind (Photo Pat Holmes)
It is of vital importance to understand the ecosystem attributes before attempting to initiate restoration interventions. In the case of fynbos, this vegetation is generally restricted to nutrient poor sandy soils, with a Mediterranean climate of cool wet winters and hot dry summers. This is a fire-dependent ecosystem in which most plant species are killed by fire and have seeds that are adapted to germinate in response to a fire through a fire-related germination cue such as smoke residue or heat. Other species are able to survive fires by resprouting from a persistent rootstock or are geophytic. Furthermore, the lowlands of the Cape Region are ecologically distinct from mountain habitats, in terms of temperature and moisture regimes, soil depths, native species composition and dominant invasive species, and therefore restoration interventions which have been successful in the mountains have been found to be ineffective in the lowlands. Furthermore, the lowlands are overall more highly degraded by habitat loss to urban and agricultural development and hence remnant natural habitats are highly fragmented and disturbed, with more frequent fires and increased opportunities for alien plants to invade these areas.
A typical remnant lowland fynbos habitat within race-tracks, isolated and disturbed (Photo Anonymous)
Within invaded fynbos ecosystems, certain elements of the vegetation persist better than others. Resprouting species can persist longer due to accumulated resources in their rootstock, while non-sprouting species put all resources into seeds and so cannot tolerate being shaded by taller growing woody invaders such as Acacia saligna. However, eventually even resprouters get outcompeted and disappear from the ecosystem. These species usually produce very little or low viability seed, and so are tricky to restore once lost from the ecosystem.
Fynbos recently cleared of dense invasive acacia, with only resprouting shrubs persisting (Photo Stuart Hall)
Fynbos generally has a persistent soil seedbank, and so even if the standing vegetation cover is lost, much of the plant diversity should be present in the soil until a fire stimulates germination, as has been found in mountain fynbos. However, this does not appear to be the case in lowland fynbos, where only a few elements – pioneer shrubs and herbaceous annuals or short-lived perennials - appear to persist in the seedbank after more than one fire cycle of dense aliens. Here a biotic threshold has been crossed, whereby the ecosystem will revert to a different stable state to its historical pristine state after removing the alien species, without active restoration intervention. Furthermore, Acacia saligna being a legume is a nitrogen fixer, and increases total soil nitrogen as well as altering other soil chemistry variables, and these changes can persist after removal of the species. Some fynbos species are sensitive to certain nutrients e.g. high Phosphorus levels are toxic to native Protea repens, while high Nitrogen in the soil favours growth of competitive weedy forbs and graminoids.
Habitat cleared of Acacia saligna two years prior but not burnt; now dominated by nonnative annual grasses (Photo Stuart Hall)
In our study site, 32 hectares were cleared of Acacia saligna and the felled slash subsequently burnt to try and stimulate germination of any remaining fynbos seedbank. This was only successful in a few patches where higher quality vegetation had persisted under the acacia canopy, while across the rest of the area there was very little or no fynbos recovery evident. Therefore in these ecosystems active seeding intervention is usually necessary in order to restore the diversity of vegetation structural components. However, seed germination in Acacia saligna is also fire-stimulated and burning after clearing caused mass acacia recruitment, while the unburnt comparison had very limited and more manageable reestablishment, but more limited autogenic fynbos recovery.
Habitat cleared of Acacia saligna two years prior and then burnt, now dominated by acacia seedlings with very limited autogenic fynbos shrub recovery apparent after follow-up clearing (mostly persisting resprouting shrubs) (Photo Stuart Hall)
In order to investigate seeding as an active intervention, we collected seed from over 20 native species within the immediate vicinity of the restoration site representing different structural components - overstorey proteas, resprouting and non-sprouting ericoid shrubs, restios (graminoid-like shrubs), geophytes, forbs – as well as representing a range of seed morphologies and therefore likely requiring different germination cues to break dormancy. Initially seed was sown into plots shortly after the site was burnt following initial acacia clearing with the rationale that sufficient smoke residue was still present to stimulate germination, but the majority of species showed no sign of germination or establishment – only around five species were successfully restored in most plots. While much work has been done showing the positive effect of smoke residue on germination of many fynbos species, other cues such as heat exposure have only been tested in a few studies involving only a few species. Therefore, we performed germination tests on the species we collected, across a range of heat treatments in combination with smoke to determine optimal species-specific pre-treatment protocol for applying before follow-up sowing treatments were done a year later in our field site.
Plots seeded after clearing and burning two years prior; some diversity of shrub cover establishing (Photo Stuart Hall)
We found that most species fell into one of two categories based on germination response to heat: either species did not benefit or were negatively affected by heat exposure of more than 60 °C and longer duration of heat exposure, or species required a heat treatment of between 60 and 100 °C, while temperatures of 150 °C and above destroyed all seed of the species tested. Species in the first group were mostly those which germinated successfully in the initial sowing treatment, while some of the species in the second group are keystone structural species and therefore important for reestablishing ecosystem structure and function, and these also did not germinate successfully in the initial sowing treatment.
Grouping of species based on seed response to heat pre-treatment before sowing: two species that did not respond positively to heat treatment (red stars: Metalasia densa; Anthospermum aethiopicum) germinated successfully in field sowing plots; heat pulse treatment increases germination success in certain keystone species (green stars: Restio – Thamnochortus punctatus; Respouters – Phylica cephalantha; Trichocephalus stipularis) (Photos Stuart Hall)
Follow-up sowing was then conducted to compare both untreated and pre-treated seed, with the latter including seed either treated just with smoke or smoke and 100 °C heat exposure for five minutes, based on species specific results from the seed pre-treatment study. Pre-treated seed resulted in better and more rapid germination success and establishment of certain species with both smoke and heat-stimulated seed, so for minimal additional effort, pre-treating seed before sowing can greatly improve active restoration success. However, some species which germinated well under lab conditions following pre-treatment of seeds still did not establish successfully in the field, so for these species there exists an additional barrier to establishment which still needs to be determined – perhaps drought sensitivity of newly germinated seedlings, disease or predation by rodents. For some of these species seed availability was very limited due to a lack of appropriate seed sources and this was insufficient for sowing on the scale required.
Plot sown with pre-treated seed showing a diversity of species and vegetation structure (Photo Stuart Hall)
Although unlikely to resemble a pristine reference condition in terms of species diversity and ecological function within a short to medium timeframe, restoring intact shrub cover consisting of at least a few species representing all the key vegetation structural components of the ecosystem is a crucial step towards restoration success. Once intact shrub cover is restored this provides competition against reestablishment of the primary invader, in this case Acacia saligna, or secondary invaders including nonnative annual grasses and forbs which will otherwise tend to dominate fynbos habitats previously invaded by alien acacias. Furthermore, having a diversity of structural component species present will help to restore a more natural fire regime; habitat invaded by woody acacias will burn too intensely and could destroy the native soil seedbank, while forb or graminoid dominated vegetation without sufficient shrub cover will burn too frequently or at too low intensity to break dormancy of native species with heat-stimulated seeds. Once intact shrub cover is restored, more of the plant diversity of the ecosystem can be returned over time either through supplementary seeding or planting out nursery grown stock for species such as those which still did not establish in spite of seed pre-treatment.
Good quality lowland fynbos vegetation in foreground with ongoing restoration of degraded habitat beyond (Photo Stuart Hall)
Key references and Further reading:
Hall, S., Newton, R., Rebelo, P., Gaertner, M., Esler, K.J. (2017) Heat and smoke pre‐treatment of seeds to improve restoration of an endangered Mediterranean climate vegetation type. Austral Ecology, 42(3): 354-366.
Hall, S. (2018) Restoration potential of alien-invaded Lowland Fynbos. PhD Dissertation, Stellenbosch University. Supervisor: Prof. K.J. Esler; Co-supervisors: Prof Patricia M. Holmes; Dr M. Gaertner. 143p.
Holmes P.M, Esler K.J., Gaertner M., Geerts S., Hall S.A., Nsikani M.M., Richardson D.M., Ruwanza S. (2019) Biological invasions and ecological restoration in South Africa In: van Wilgen B.W., Measey G.J., Richardson D.M., Wilson J.R., Zengeya, T. (eds) Biological invasions in South Africa (pp. 665-700). Springer, Berlin.
Nsikani, M.M., Gaertner, M., Kritzinger-Klopper, S., Ngubane, N.P., Esler, K.J. (2019) Secondary invasion after clearing invasive Acacia saligna in the South African Fynbos. South African Journal of Botany 125: 280-289.
Partners:
The City of Cape Town, Key Contact: Dr Charmaine Oxtoby; Biodiversity Management Branch I Environmental Management Department I Spatial Planning and Environment Directorate; charmaine.oxtoby@capetown.gov.za
Stellenbosch University, Centre for Invasion Biology and Dept. Conservation Ecology & Entomology. Key Contacts: Professor Karen Esler (ConsEnt; kje@sun.ac.za), Professor Pat Holmes (ConsEnt & C.I.B, holmes.patricia.m@gmail.com).
Millennium Seed Bank Partnership, Royal Botanic Gardens, Kew, United Kingdom. Key Contact: Dr Rosemary Newton; r.newton@kew.org